Imaging Pearls ❯ October 2025
|
-- OR -- |
|
3D and Workflow
- AI capabilities have rapidly transformed from a primarily perceptive modality, like speech recognition and computer vision, to its ubiquitous generative form, as seen through many generative pretrained transformer platforms today. However, the scalability of AI goes beyond the production of original tokens and usage as a tool and is now characterized by its capacity for autonomous tasks, paving the way for a complementary digital workforce .Agentic AI redefines the conventional deterministic or rule-based paradigm, as, it is now capable of handling complex workflows and business processes.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - “Agentic AI leverages generative AI to create entities that mimic how humans perform tasks by breaking down complex workflows into modular, specialized components. For example, one health care agent could include subtasks services such as health record retrieval and review, literature and reference table analysis, medical specialist services, checklist management, and privacy and compliance guardrails— all delivered through a friendly and intuitive user interface. When integrated, those components form a cohesive unit capable of handling intricate, multidomain workflows. Each distinct subtask assumes varying methodology, each with its own inherent strengths and weaknesses, to investigate scenarios from a broad range of perspectives and execute complex functions with minimal human oversight.”
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - For instance, Abridge listens during patient visits and automatically generates clinical notes, reducing documentation time and enabling natural conversations between the patient and provider without requiring each party to remember every detail. Similarly, tools such as Deloitte’s Frontline AI, Hippocratic.ai , and Intrivo use digital intake agents to automate pre-operative checklists, postoperative care, and medical concierge—freeing health care workers to focus on direct care.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - In summary, the emergence of agentic AI represents an existential shift in the standard functionality of the health care industry—one that transcends the traditional boundaries of perceptive or generative AI, merging intelligence with service in transformative ways. Its adaptability and customizable subtasks enable specified designs,empowering both providers and patients to reclaim their time and autonomy. More than just a technological advancement, agentic AI introduces a new operational model that is scalable, modular, and cohesively integrated into the existing digital infrastructure of health care.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - Radiology, like other medical specialties, is at the forefront of change in what can only be described as the “industrial revolution of intelligence.” We are not seeing just incremental adjustments to simply squeeze out a bit more efficiency—but, instead we are witnessing the start of a radical change in how work will be done across industries and professions. In time, we will learn how to optimize our use agentic AI as a powerful tool to reshape our entire workflow. Whether in our interactions with patients or collaborations with referring clinicians, AI presents opportunities to optimize the mundane aspects of what we do and to rediscover the joy of practicing medicine.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - How we process our tasks, especially those that are multimodal, will benefit most. Particularly in complex, collaborative environments like tumor boards, AI integration is becoming tangible.The fundamental underpinnings of our workflow today have evolved minimally in the past decade—but we now stand on the brink of fully contextualized, augmented, real-time decision making when imaging, genomics, pathology, clinical trials, and longitudinal patient histories can converge seamlessly, whether it is in the conference room or in clinic.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440
Deep Learning
- But something about this moment feels different. Generative artificial intelligence (AI), with tools like ChatGPT, offers information in ways that feel uniquely conversational and tailored. Their tone invites dialogue. Their confidence implies competence. Increasingly, patients are bringing AI-generated insights into my clinic and are sometimes confident enough to challenge my assessment and plan.
When Patients Arrive With Answers.
Sundar KR.
JAMA. 2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print. - Yet comparing LLMs to clinicians feels inherently unfair. Clinicians often work under pressure with rushed visits and overflowing inboxes as health care systems demand productivity and performance. My concern is that research is comparing clinicians who are not given the luxury of performing their best under strained systems with the inexhaustible resources of generative AI. That is not a fair fight, but it is reality. I find patients may seek clear answers, but even more, they want to feel recognized, reassured, and truly heard. Unfortunately, under the weight of competing demands, that is what often slips for me—not just because of systemic constraints but also because I am merely human.
When Patients Arrive With Answers.
Sundar KR. JAMA.
2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print. - Clinicians remain the gatekeepers. In practice, this means navigating patient requests like a tilt-table test for intermittent dizziness— tests that are not unusual but may not be appropriate at a specific stage of care. I find myself explaining concepts like overdiagnosis, false-positives, or other risks of unnecessary testing. At best, the patient understands the ideas, which may not resonate when one is the person experiencing symptoms. At worst, I sound dismissive. There is no function that tells ChatGPT that clinicians lack routine access to tilt-table testing or that echocardiogram appointments are delayed due to staffing shortages. I have to carry those constraints into the examination room while still trying to preserve trust.
When Patients Arrive With Answers.
Sundar KR.
JAMA. 2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print. - From a patient’s perspective, I recognized that all the cognitive explaining can sound like: “I still know more than you, no matter what tool you used, and I’m going to overwhelm you with things you don’t understand.” What working in that moment was much different. I said, “We’ll talk through the testing options. But first, I’m so sorry for your loss. I can’t imagine how you’re feeling. I want to figure this out with you and make a plan together.”
That moment of acknowledgment was what really opened the door. When Patients Arrive With Answers.
Sundar KR.
JAMA. 2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print. - Seeing AI-informed visits as opportunities for deeper dialogue rather than threats to clinical authority may sound aspirational, but it reflects a necessary shift. The practical challenges are real. Patients may come with unrealistic expectations or cite recommendations that do not align with evidence-based guidelines or are impractical for a given resource setting. These moments are not new. We have long had to explain why magnetic resonance imaging scan is not always needed for back pain or why antibiotics will not help a viral infection. We know the solution is not to shut these conversations down but to meet them with patience and curiosity. Medicine has always depended on relationships. What is changing is how those relationships begin, and what patients bring to the table. If patients are arming themselves with information to be heard, our task as clinicians is to meet them with recognition, not resistance. In doing so, we preserve what has always made medicine human: the willingness to share meaning, uncertainty, and hope, together.
When Patients Arrive With Answers.
Sundar KR.
JAMA. 2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print. - We know the solution is not to shut these conversations down but to meet them with patience and curiosity. Medicine has always depended on relationships. What is changing is how those relationships begin, and what patients bring to the table. If patients are arming themselves with information to be heard, our task as clinicians is to meet them with recognition, not resistance. In doing so, we preserve what has always made medicine human: the willingness to share meaning, uncertainty, and hope, together.
When Patients Arrive With Answers.
Sundar KR.
JAMA. 2025 Jul 24. doi: 10.1001/jama.2025.10678. Epub ahead of print.
- Expectation
Healthcare professionals often have unreasonably high expectations for AI/ML systems, expecting them to be flawless before clinical use.
Medicine operates in a nondeterministic world; AI/ML should augment oncologists' judgment rather than replace it - Probability
Many AI/ML studies present results in a binary manner, oversimplifying the nuanced nature of medical decision-making.
Embracing probabilistic outcomes allows for better integration of AI/ML insights with clinician expertise. - Calibration
AI/ML models often produce outputs that are not properly calibrated, leading to inaccurate probability estimates.
Techniques like Platt scaling and isotonic regression can help align model outputs with reality - Reproducibility
Consistency in AI/ML outputs is crucial for clinical decision-making; stochastic algorithms can lead to different answers on repeated runs.
Irreproducibility is unacceptable in clinical settings, unlike in research. - Validity
The effectiveness of AI/ML algorithms is tied to the diversity of the training population.
Models must be trained on representative datasets to ensure validity across different demographic groups - Transparency
"Glass-box" models provide transparency in AI/ML decision-making processes, building trust and improving interpretability.
Techniques like SHAP and LIME can clarify the factors influencing model outputs - Positive Predictive Value
Many studies report performance metrics based on balanced datasets, which do not reflect real-world distributions.
In low-incidence settings, high sensitivity and specificity may yield low positive predictive value, leading to unnecessary follow-ups - Automation Bias
Automation bias can lead to overreliance on AI/ML outputs, undermining human critical evaluation.
Clinicians must remain engaged in the decision-making process to mitigate risks associated with automation bias. - AI capabilities have rapidly transformed from a primarily perceptive modality, like speech recognition and computer vision, to its ubiquitous generative form, as seen through many generative pretrained transformer platforms today. However, the scalability of AI goes beyond the production of original tokens and usage as a tool and is now characterized by its capacity for autonomous tasks, paving the way for a complementary digital workforce .Agentic AI redefines the conventional deterministic or rule-based paradigm, as, it is now capable of handling complex workflows and business processes.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - “Agentic AI leverages generative AI to create entities that mimic how humans perform tasks by breaking down complex workflows into modular, specialized components. For example, one health care agent could include subtasks services such as health record retrieval and review, literature and reference table analysis, medical specialist services, checklist management, and privacy and compliance guardrails— all delivered through a friendly and intuitive user interface. When integrated, those components form a cohesive unit capable of handling intricate, multidomain workflows. Each distinct subtask assumes varying methodology, each with its own inherent strengths and weaknesses, to investigate scenarios from a broad range of perspectives and execute complex functions with minimal human oversight.”
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - For instance, Abridge listens during patient visits and automatically generates clinical notes, reducing documentation time and enabling natural conversations between the patient and provider without requiring each party to remember every detail. Similarly, tools such as Deloitte’s Frontline AI, Hippocratic.ai , and Intrivo use digital intake agents to automate pre-operative checklists, postoperative care, and medical concierge—freeing health care workers to focus on direct care.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - In summary, the emergence of agentic AI represents an existential shift in the standard functionality of the health care industry—one that transcends the traditional boundaries of perceptive or generative AI, merging intelligence with service in transformative ways. Its adaptability and customizable subtasks enable specified designs,empowering both providers and patients to reclaim their time and autonomy. More than just a technological advancement, agentic AI introduces a new operational model that is scalable, modular, and cohesively integrated into the existing digital infrastructure of health care.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - Radiology, like other medical specialties, is at the forefront of change in what can only be described as the “industrial revolution of intelligence.” We are not seeing just incremental adjustments to simply squeeze out a bit more efficiency—but, instead we are witnessing the start of a radical change in how work will be done across industries and professions. In time, we will learn how to optimize our use agentic AI as a powerful tool to reshape our entire workflow. Whether in our interactions with patients or collaborations with referring clinicians, AI presents opportunities to optimize the mundane aspects of what we do and to rediscover the joy of practicing medicine.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440 - How we process our tasks, especially those that are multimodal, will benefit most. Particularly in complex, collaborative environments like tumor boards, AI integration is becoming tangible.The fundamental underpinnings of our workflow today have evolved minimally in the past decade—but we now stand on the brink of fully contextualized, augmented, real-time decision making when imaging, genomics, pathology, clinical trials, and longitudinal patient histories can converge seamlessly, whether it is in the conference room or in clinic.
Agentic Artificial Intelligence: The Power to Change Medicine and Our World.
Powell K, Fishman EK, Chu LC, Rowe SP, Crawford CK.
J Am Coll Radiol. 2025 Jun 25:S1546-1440
- “Every week, numerous articles are published that showcase the potential of artificial intelligence (AI) and machine learning (ML) to revolutionize medical diagnostics. These studies often report impressive results, yet a significant gap persists between these conceptual and theoretical advances and the practical deployment of AI/ML in clinical settings. This disconnect raises critical questions about why AI/ML tools, despite their promise, have yet to become commonplace in oncology.”
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041. Epub ahead of print PMID: 40833461. - Although the application of AI/ML in health care is relatively new, the underlying principle of deriving actionable Insights from data in diagnostics are not. For centuries, Physicians have observed clinical signs and symptoms to recognize patterns and make informed decisions. In more In recent decades, the use of biomarkers combined with statistical Analysis has refined this process, enabling more precise, data-driven diagnostics. Although these processes may not have been labeled “artificial intelligence,” their core principles They are strikingly similar. Health care providers can use AI/ML systems to help analyze data and identify patterns once they They are trained to do so. AI/ML’s key advantage is its ability to be trained on massive amounts of data—too vast for any single individual or committee—thus highlighting anomalies that might otherwise be missed.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041. - Expectation. Many healthcare professionals hold high— sometimes unreasonably high—expectations for AI/ML systems, believing these tools must be flawless before clinical integration. Although high standards are essential, it is important to recognize that medicine operates in a nondeterministic world. Just as the discovery of the CA-125 biomarker for ovarian cancer provided valuable insights without guaranteeing perfect predictions, AI/ML algorithms can improve clinical decision-making without being perfect. These tools should be seen as augmenting an oncologist’s judgment rather than replacing it.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041. - Probability. A major issue in many AI/ML studies is presenting results as definitive outcomes—labeling diagnoses simply as “cancer” or “no cancer.” This binary approach assumes a deterministic world, oversimplifying the nuanced nature of medical decision-making. For example, a patient with a 99% probability of having cancer is markedly different from one with a 51% probability, yet deterministic models might treat these cases as equivalent. Medicine and human biology are inherently probabilistic; most outcomes cannot be perfectly predicted, and variability is a fundamental characteristic of all data. Even two experienced pathologists, radiologists, or oncologists may differ in their diagnoses when analyzing the same data. Embracing probabilistic outcomes facilitates the fusion of insights generated from AI/ML with those generated by clinicians.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041. - Calibration. AI/ML models often generate outputs on a scale from 0 to 1, which are sometimes interpreted as probabilities. However, such outputs are often not properly calibrated, meaning they do not accurately estimate the probability of a diagnosis (1). For example, an accurate algorithm (one that has an acceptable sensitivity and specificity) may predict—via an output score of 0.9—that a patient has a 90% chance of developing cancer when the actual probability of developing cancer is only 10%. Though techniques such as Platt scaling (2) and isotonic regression (3) can help align a model’s probabilistic outputs with reality, these are often not employed and even when employed do not always function effectively. The development of more effective calibration techniques is an important area for future research.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041. - Reproducibility. For clinical decision-making, it is important that the answer to a question, when the same data are used to derive the answer, is identical. For example, if a pathologist were to change a diagnosis based on examining the same microscopic slide on two sequential days, that pathologist’s diagnostic skill would be put into question. This issue poses a major problem for many tools commonly used in AI/ML. For example, recent transformer-based neural network architectures sample the next token from an estimate of prior data, rendering them inherently stochastic. Even classical machine learning approaches typically use stochastic algorithms to train their models, including stochastic gradient descent, bagging, and boosting. This means that the identical ML/AI algorithm could provide different answers if they were run multiple times on the same data, and in practice, they sometimes do. Though somewhat different answers are acceptable for many applications of ML/AI in research settings, such irreproducibility is not acceptable for clinical decision-making.
- Validity. The effectiveness of an AI/ML algorithm is closely tied to the population on which it was trained. For example, a skin cancer detection model developed primarily using images from Caucasian patients may struggle to accurately diagnose individuals with darker skin tones (4). This limitation underscores the critical need for training AI/ML models on diverse, representative datasets to ensure validity across different demographic groups. Even with diverse training data, situations may arise where a patient’s data fall outside the distribution of the training set—a scenario known as “external validity” or “out-of-distribution” (5). In addition to differences among patients, technical differences in data acquisition can lead to misleading predictions. This issue is especially relevant to fields such as medical imaging, in which even minor changes in imaging devices can significantly affect model performance.
- Transparency. Unlike “black-box” models that provide outputs without revealing their decision-making processes, “glass-box” AI/ML algorithms offer transparency by showing clinicians how predictions are generated. Whether data come from medical imaging or tabular records, transparency builds trust in AI/ML systems, aids in identifying potential errors, and improves interpretability. By understanding the factors driving predictions, clinicians can make more informed decisions and confidently integrate AI/ML tools into practice. Although many AI/ML algorithms—such as deep learning models—are often considered black-boxes, interpretability techniques like SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME)can help clarify their inner workings. These methods offer insights into which factors most influence model outputs, making it easier for clinicians to understand and trust the decisions made by complex models. .
- A compelling example of the need for glass-box models is a study that introduced a deep neural network designed to diagnose skin cancer with accuracy comparable with board-certified dermatologists. Later, it was discovered that the AI model had inadvertently learned to identify rulers present in the images—tools often found in images labeled as cancerous—rather than focusing solely on lesions. Such a major problem would have been averted if the model had been more transparent.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041 - Positive Predictive Value. Many AI/ML studies report impressive performance metrics based on balanced datasets, in which the prevalence of a condition—such as cancer—is set to equal proportions (e.g., 50% cancer and 50% noncancer). In these datasets, positive and negative cases are often selected in ways that do not mirror real-world data distributions.Whereas metrics such as sensitivity and specificity are robust to class imbalances, other metrics—particularly the positive predictive value (PPV)—are not. PPV is defined as the ratio of true positives to the total positive calls (true positives plus false positives). In low-incidence settings—such as pancreatic cancer screening, annual incidence of which is <0.1%—simply knowing that a test has high sensitivity and specificity provides limited insights into its practical performance. A test with 80% sensitivity and 80% specificity yields a PPV of <1% in such circumstances, resulting in unnecessary follow-up tests, increased healthcare costs,and heightened patient anxiety.
- Though recent advances have shown that it is possible to design AI/ML algorithms with predefined PPVs, much work remains to be done, particularly in applications involving screening of large numbers of healthy individuals. An excess of false-positive results can make a test more likely to do harm than good when applied to a population that has an average risk for cancer.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041 - Automation Bias. Automation bias refers to the tendency to place excessive trust in technology or automated systems solely because their output is computer-generated. Overreliance on automated outputs can undermine human performance when humans fail to perform their own critical evaluations. Recent discussions on clinical decision support highlight the danger of uncritically accepting AI/ML suggestions and help mitigate anchoring bias—initial information of which disproportionately influences subsequent decisions. Overcoming this challenge aligns with calls for explainable AI/ML, emphasizing that clinicians must remain “in the loop” and verify automated recommendations independently.
Translating Artificial Intelligence Breakthroughs into Cancer Diagnostic Breakthroughs.
Lavista Ferres JM, Fishman EK, Catmull E, Vogelstein B, Vogelstein JT.
Cancer Discov. 2025 Aug 20:OF1-OF3. doi: 10.1158/2159-8290.CD-25-1041 - Like any powerful tool, the value of AI/ML depends on how well its strengths and limitations are understood. We suggest that current data generated through AI/ML be designated as “AI-informed data” to distinguish those from other types of patient data, such as those relating to clinical symptoms, signs, or conventional laboratory tests. Clinicians, patients, researchers, and regulators should view these technologies as valuable tools that inform and support diagnosis rather than replacements for clinical decision-making. In oncology, there are established safeguards to protect patients from new forms of diagnosis or treatment until they have been unequivocally established to be clinically useful: prospective, interventional clinical trials. Such trials often require substantial investments of time, resources, and effort by the private and public sectors. By addressing the challenges discussed above prior to performing definitive prospective clinical trials, we believe it is much more likely that such clinical trials will be successful. Then, and only then, will it be possible to fully integrate AI/ML into clinical practice, ultimately improving the care of patients with cancer.
- Pancreatic Lesion Detection
We proposed a multitask model: a model that combines a local segmentation head using voxel cross entropy loss, with a global classification head that captures secondary cues such in the gland, the pancreatic duct and common bile duct.
Our model can aggregate local and global information for more accurate detection. Also our model provides a set of classification probabilities as it analyzes each lesion. 
Kidney
- “Wilms tumor (or nephroblastoma) is the most common renal malignancy in children. The majority of cases are solitary tumors, but 4%–13% of children have bilateral tumors and about 10% have multifocal disease within a single kidney. About 95% of Wilms tumors occur in children younger than 15 years of age, with the vast majority presenting in the first 5 years of life. Ninety-one percent are sporadic diseases; the other 9% have a genetic predisposition and typically present at a younger age. Clinical presentation ranges from asymptomatic abdominal masses to hypertension, abdominal pain, hematuria, or retroperitoneal hemorrhage manifesting with anemia.”
Staging and Restaging Pediatric Abdominal and Pelvic Tumors: A Practical Guide.
de Faria LL, et al.
Radiographics. 2024 Jun;44(6):e230175. - “Wilms tumor often exhibits internal necrosis and bleeding, resulting in a varied imaging appearance, but it is uncommon to find calcifications (~10% of cases). Vessels are directly invaded or displaced rather than encased and extension into the spinal canal is never seen, which can aid in distinguishing it from neuroblastoma.”
Staging and Restaging Pediatric Abdominal and Pelvic Tumors: A Practical Guide.
de Faria LL, et al.
Radiographics. 2024 Jun;44(6):e230175. - “The most challenging aspect in diagnosing Wilms tumor is differentiating it from neuroblastoma, since both can manifest as a large retroperitoneal mass. Identifying that the mass has a renal origin is one of the most important criteria in the differential diagnosis. Imaging findings of renal masses typically include distortions of the normal renal parenchyma, creating a claw sign, whereas extrarenal masses displace the kidney.”
Staging and Restaging Pediatric Abdominal and Pelvic Tumors: A Practical Guide.
de Faria LL, et al.
Radiographics. 2024 Jun;44(6):e230175. - Tumors (mean diameter, 13.1 ± 4.5 cm) exhibited an expansible appearance and disrupted the reniform contour (16/16), cystic components (16/16), curvilinear calcification (1/16), poorly marginated (2/16), hemorrhage (16/16), displacement of renal pelvis or calyx (13/16), and had lymph node or distal metastases (5/16). Attenuation of WT was less or equal compared to renal parenchyma on unenhanced CT (P > 0.05), while tumor enhancement after administration of a contrast agent was lower than that of normal renal parenchyma (P < 0.05).
CT and MRI imaging features and long-term follow-up of adult Wilms' tumor.
Wu J, Zhu Q, Zhu W, Chen W.
Acta Radiol. 2016 Jul;57(7):894-900. 
Multimodality imaging review of retroperitoneal fibrosis.
Czerniak S, Mathur M.
Abdom Radiol (NY). 2025 Mar 4. doi: 10.1007/s00261-025-04847-6. Epub ahead of print. PMID: 4035807.
- Renal Oncocytoma: Facts
The most common benign mimic of clear cell RCC is oncocytoma, which has a similar shape and tends to have a more homogeneous enhancement with unenhanced HU between 25 and 40, corticomedullary phase peak absolute HU below 120 and slower washout with nephrographic phase HU and excretory phase enhancement below 100. A subset of both oncocytomas and clear cell RCC have a central scar with delayed enhancement. Another mimic of clear cell RCC is fat-poor AML, which tends to have a lobular, mushroom or ovoid shape lesion often with acute angles to the renal cortex. - Renal oncocytomas are common benign kidney neoplasms that account for 3 to 7 percent of renal neoplasms. They usually occur in adults, most frequently in the seventh decade of life. Renal oncocytomas may be discovered incidentally or may be diagnosed with biopsy or excision. Renal oncocytomas usually have an excellent prognosis and are not associated with an aggressive clinical course; however, there can occasionally be metastasis to liver and bone and fatalities have occurred. Surgical excision is curative in the absence of metastases.
Renal Oncocytoma. [Updated 2023 Aug 28].
In: StatPearls
Williams GM, Lynch DT. - Oncocytomas make up 3% to 7% of all renal epithelial tumors. They are most commonly seen in adults older than 50 years, are more common in males, and have a peak frequency in the seventh decade of life. However, these tumors rarely occur in the pediatric population. This lesion is commonly seen in patients with Birt-Hogg-Dube syndrome. These patients develop a mean of 5.3 renal tumors in their lifetime. These lesions consist mostly of chromophobe renal carcinoma and oncocytomas.
Renal Oncocytoma. [Updated 2023 Aug 28].
In: StatPearls
Williams GM, Lynch DT. - Evaluation is usually done on imaging initially. Oncocytomas may be discovered incidentally. There are no precise radiologic criteria for distinguishing a benign oncocytoma from a malignant lesion like renal cell carcinoma. However, central scar and hypervascularity can often be appreciated on imaging Evaluation of a kidney mass can be done with core biopsy, fine needle aspiration, or resection. Oncocytomas are often small, and since 20% to 45% of small renal masses are ultimately found to be benign, active surveillance is often an option offered to some patients. If the mass is biopsied, there is up to an 80% diagnostic rate and subtype and nuclear grade can be provided.
Renal Oncocytoma. [Updated 2023 Aug 28].
In: StatPearls
Williams GM, Lynch DT.
- Renal Medullary Carcinoma: Facts
Renal medullary carcinoma (RMC) is a rare and highly aggressive form of kidney cancer. It mainly affects young people of African descent, especially people who carry the sickle cell trait or have sickle cell disease or other sickle hemoglobinopathies
RMC was originally described in 1995 and is one of the most aggressive kidney cancers. Half of the patients with RMC described in the original 1995 study did not survive longer than 4 months after diagnosis. With current therapies this has improved to 13 months. - Solitary Fibrous Tumor of the Kidney
The origin of most cases of SFT of the kidney is difficult to determine. Some reported cases of solitary fibrous tumor of the kidney were reported to have originated from the renal capsule.
Although most cases are benign, the behavior of SFTs is unpredictable. Roughly 10% to 15% of these tumors behave aggressively; thus, long-term follow-up is mandatory.
- IgG4-related disease (IgG4-RD) is an immune-mediated disorder marked by fibro-inflammatory masses that can infiltrate multiple organ systems. Due to its relatively recent discovery and limited understanding of its pathophysiology, IgG4-related disease may be difficult to recognize and is consequently potentially underdiagnosed. Renal involvement is becoming regarded as one of the key features of this disease. To date, the most well-recognized renal complication of IgG4-related disease is tubulointerstitial nephritis, but membranous glomerulonephritis, renal masses, and retroperitoneal fibrosis have also been reported.
Renal Manifestations of IgG4-Related Disease: A Concise Review.
Towheed ST, Zanjir W, Ren KYM, et al..
Int J Nephrol. 2024 Jun 24;2024:4421589. - Renal involvement is now regarded as one of the key features of this disease. In 2004, the first reports of an association between Type 1 autoimmune pancreatitis and renal dysfunction were identified. Since that time, renal dysfunction has also been associated with Mikulicz's syndrome , IgG4-related hepatic involvement , and other extra-renal IgG4-RD syndromes. In these initial reports, renal involvement is manifested as tubulointerstitial nephritis, which to date is still the most well-recognized renal manifestation of IgG4-RD. However, glomerular involvement has also been described—membranous glomerulonephritis is the primary glomerular injury pattern noted in the literature. Additional described manifestations include renal masses and retroperitoneal fibrosis that can secondarily affect the renal system The diversity of potential manifestations in the kidney has led to more encompassing terminology entitled IgG4-related kidney disease (IgG4-RKD).
Renal Manifestations of IgG4-Related Disease: A Concise Review.
Towheed ST, Zanjir W, Ren KYM, et al..
Int J Nephrol. 2024 Jun 24;2024:4421589. - IgG4 Renal Disease
Abnormal renal radiologic findings:
Multiple low-density lesions on enhanced computed tomography
Diffuse kidney enlargement
Hypovascular solitary mass in the kidney
Hypertrophic lesion of the renal pelvic wall without irregularity of the renal pelvic surface - Radiologically, one of the most reliable findings is the presence of multiple low-attenuation renal lesions on contrast-enhanced CT, as depicted. Additional findings may include diffuse kidney enlargement and solitary renal masses mimicking neoplasms, among others . Clinically, the onset and course of renal involvement can be acute but are generally slowly progressive.
Renal Manifestations of IgG4-Related Disease: A Concise Review.
Towheed ST, Zanjir W, Ren KYM, et al..
Int J Nephrol. 2024 Jun 24;2024:4421589. - Immunoglobulin G4–related disease (IgG4-RD) is a systemic fibroinflammatory disease characterized by focal or diffuse organ infiltration of IgG4-bearing plasma cells. The diagnosis of IgG4-RD is based on a combination of clinical, serologic, radiologic, and histopathologic findings. IgG4-RD has been reported to affect almost all organ systems. The kidney is the most frequently involved of the genitourinary organs. The most common renal manifestation of IgG4-RD is IgG4-RD tubulointerstitial nephritis, followed by membranous glomerulonephropathy and, less frequently, obstructive nephropathy involving the renal pelvis, ureter, or retroperitoneum.
Immunoglobulin G4–related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, Soon Nam Oh, Seo Yeon Youn, and Joon-Il Choi
RadioGraphics 2020 40:5, 1265-1283 - - The kidneys are the genitourinary organs most commonly involved with IgG4-RD. Kidney involvement is found in approximately one-fourth to one-third of patients with IgG4-RDautoimmune pancreatitis, but can also occur without the involvement of other organs.
- IgG4-RD of the kidney predominantly involves the renal cortex, but the renal pelvis, renal sinus, and perirenal space can be involved. The most frequent renal manifestation of IgG4-RD is IgG4-RD tubulointerstitial nephritis, followed by IgG4-RD membranous glomerulonephropathy. Less frequently, obstructive nephropathy can be caused by postrenal obstruction secondary to renal pelvis, ureter, or retroperitoneal involvement.
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - - Renal parenchymal IgG4-RD may show several imaging pat-terns, including multiple nodular lesions, diffuse patchy infiltrative lesions, and a single nodular lesion. Of these, the multiple nodular type is the most common imaging finding for IgG4-RD of the kidney.
- Ureteral IgG4-RD can be classified into three types on the basis of gross morphologic features: polypoid mass-forming lesions, segmental ureteral wall thickening, and periureteral fibrosis.
- IgG4-RD involving the kidneys, ureter, bladder, urethra, prostate, testes, and female reproductive organs can show a broad spectrum of imaging findings, such as a localized mass in or surrounding the involved organ or diffuse enlargement of the involved organ, which may mimic a variety of benign and malignant diseases.
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283- The characteristic imaging findings of autoimmune pancreatitis, which include sausage-like enlargement of the pancreas and a peripancreatic halo, can be strongly suggestive of IgG4-RD if they are detected in the proper clinical context that includes (a) mild abdominal symptoms, usually without acute attacks of pancreatitis;(b) occasional occurrence of obstructive jaundice; (c) increased serum gamma globulin, IgG, and/or IgG4 concentrations; and (d) occasional association with other organ involvement .The presence of a peripancreatic halo corresponding to a fibroinflammatory process extending into the peripancreatic adipose tissue is a useful imaging finding for the diagnosis of IgG4-RD and for differentiating it from pancreatic cancer or other pancreatitis.
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - Patients with IgG4-RD of the kidney present at an average age of 65 years (range, 14–85 years) with a male-to-female ratio of 4:1 to 3:1, which is similar to that in patients with IgG4-RD involving other organs. Patients with IgG4-RD of the kidney usually present with mild symptoms or are asymptomatic. The usual symptoms include hematuria and an elevated serum creatinine level caused by acute renal injury, obstructive uropathy, and flank pain similar to that with renal malignancy. However, two well-recognized major clinical manifestations are incidental abnormal imaging findings during systemic screening workup for IgG4-RD and unexplained renal dysfunction. Approximately 80% of patients with IgG4-RD of the kidney show elevation of serum IgG4 levels.
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - IgG4-RD of the kidney may show a broad spectrum of imaging manifestations according to the involved anatomic location and stage of disease. Because the imaging findings of IgG4-RD of the kidney are diverse and nonspecific, radiologists should be familiar with them when considering IgG4-RD in the differential diagnosis. According to its location, IgG4-RD of the kidney can be divided into renal parenchymal lesions, renal pelvic and/or sinus lesions, and perinephric lesions. Of these, renal parenchymal lesions are the most common manifestations of IgG4-RD of the kidney.
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - “Retroperitoneal fibrosis is a fibroinflammatory disease that develops around the abdominal aorta and the iliac arteries and spreads into the adjacent retroperitoneum. It can be either idiopathic(>75% of cases) or secondary to infections, malig-nancies, drugs, or other conditions. IgG4-RD is the cause of up to two-thirds of cases of idiopathic retroperitoneal fibrosis. Ureteral involvement is the most common complication of retroperitoneal fibrosis, which usually causes medial deviation of the ureter and/or obstruction by extrinsic compression. Ureteral encasement can be unilateral or bilateral. At imaging, it is difficult to differentiate IgG4-RD retroperitoneal fibrosis from other causes. Typical imaging findings of idiopathic retroperitoneal fibrosis are a well-demarcated but irregular soft-tissue mass or plaque surrounding the anterolateral sides of the abdominal aorta and its major branches .”
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - “IgG4-RD can involve almost every organ in the genitourinary system. Of these, the kidneys are the most frequently involved organs. IgG4-RDinvolving the kidneys, ureter, bladder, urethra, prostate, testes, and female reproductive organs, can show a broad spectrum of imaging findings, such as a localized mass in or surrounding the involved organ or diffuse enlargement of the involved organ, which may mimic a variety of both benign and malignant diseases. The diagnosis of IgG4-RD is based on a combination of clinical history, imaging findings, serologic markers, and characteristic histopathologic features. Although imaging findings are nonspecific for genitourinary system involvement of IgG4-RD, imaging has a key role in the detection of disease and monitoring of treatment response.”
Immunoglobulin G4–Related Disease of the Genitourinary System: Spectrum of Imaging Findings and Clinical-Pathologic Features
Ji Woon Oh, Sung Eun Rha, Moon Hyung Choi, et al.
RadioGraphics 2020 40:5, 1265-1283 - The classic triad of flank pain, a palpable abdominal mass, and hematuria occurs in less than 10% of patients with newly diagnosed RCC.23 Because the retroperitoneal space can accommodate substantial tumor growth prior to symptom onset, only large RCCs are detected by palpation. Currently, the widespread use of abdominal imaging leads to incidental RCC detection in 37% to 61% of cases. With increased incidental detection, gross hematuria is currently reported in less than 25% of patients and occurs more often in advanced disease. Approximately 1.3% of patients with gross hematuria are diagnosed with RCC.
Renal Cell Carcinoma: A Review
Tracy L. Rose, William Y. Kim
JAMA. 2024;332(12):1001-1010
Pancreas
- Pancreatic Lesion Detection
We proposed a multitask model: a model that combines a local segmentation head using voxel cross entropy loss, with a global classification head that captures secondary cues such in the gland, the pancreatic duct and common bile duct.
Our model can aggregate local and global information for more accurate detection. Also our model provides a set of classification probabilities as it analyzes each lesion.
Spleen
- Applying an approach frequently used in imaging to the splenic mass—based on the number and consistency of lesions and refined by supplementary imaging features—allows formulation of a useful differential diagnosis. Solitary cystic masses include true cysts, pseudocysts, and parasitic cysts. When multiple cystic lesions are present, the differential diagnosis expands to include infectious lesions (abscess or microabscesses) and lymphangioma (a benign cystic neoplasm). Hemangioma is the most common solitary solid mass, although other vascular lesions (hamartoma, sclerosing angiomatoid nodular transformation) and nonvascular lesions (inflammatory pseudotumor, lymphoma) manifest as solitary and solid.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - When multiple solid masses are present, diffuse inflammatory disease (sarcoidosis), littoral cell angioma, and lymphoma should be considered. Malignancies, such as angiosarcoma or metastasis, can manifest as solitary or multiple and solid or cystic masses but are typically associated with symptoms or widespread primary malignancy. Careful assessment of the multimodality imaging characteristics of splenic lesions based on this approach aids the radiologist faced with the incidental splenic lesion.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - A solitary splenic cyst is almost always benign and usually classified as parasitic or nonparasitic; nonparasitic cysts are then subclassified as primary true or secondary false cysts, as determined by the presence or absence of an epithelial lining. Secondary cysts, also known as nonpancreatic pseudocysts, are more common and are associated with trauma, including hemorrhage, infarction, or inflammation.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. 
- Splenic abscesses can be bacterial, mycobacterial, or fungal and can result from hematogenous seeding, direct extension, sequelae of trauma, or prior infarcts. Common causes of bacterial abscess include endocarditis, pneumonia, gastrointestinal perforation, or arteriovenous malformation. The most common bacterial microbes include Escherichia coli, Staphylococcus, Streptococcus,and Salmonella. Although uncommon in the United States, granulomatous infection with Mycobacterium tuberculosis can result in abscesses due to hematogenous disseminated miliary spread.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Fungal microabscesses usually occur in patients with prolonged neutropenia. Predisposing factors include immunocompromise due to HIV infection, chemotherapy, immunosuppression for organ transplantation, or immunodeficiency. The most common fungal pathogens are Candida, Aspergillus, and Cryptococcus. Clinically, patients present with fever, abdominal pain, chills, constitutional symptoms, and sometimes septic shock.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Hemangioma is the most common benign lesion in the spleen. Most are found incidentally in asymptomatic patients. Usually solitary, hemangiomas can be multiple or diffuse; hemangiomatosis can reflect a manifestation of systemic angiomatosis in Klippel-Trenaunay syndrome. Kasabach-Merritt syndrome (anemia, thrombocytopenia, and coagulopathy) has been associated with large hemangiomas.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Splenic hemangioma can show the well-described imaging characteristics of typical hepatic hemangioma: (a) well-defined hyperechoic mass without posterior acoustic shadowing at US; (b) well-circumscribed mass with peripheral nodular discontinuous early enhancement and homogeneous progressive enhancement at CT and MRI; and (c) homogeneous T2 hyperintensity at MRI. However, splenic hemangioma complicated by fibrosis, hemorrhage, or cystic degeneration can have a variable multimodality imaging appearance.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Women are more likely to present with symptomatic larger lesions, suggesting a role for hormonal stimulation. Larger hamartomas can manifest with splenomegaly. Rarely, hematologic disorders including pancytopenia, anemia, or thrombocytopenia can occur from sequestration of hematopoietic cells. Splenic hamartoma has been associated with other hamartomatous entities, including tuberous sclerosis.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - The multimodality imaging features of SANT have been described in the literature in several case reports and generally parallel the gross appearance as a solitary well-defined mass . A characteristic appearance at contrast-enhanced US, CT, and MRI is early peripheral rim enhancement with radiating bands of progressive centripetal enhancement, described as a spokewheel appearance. At MRI, radiating T2-hypointense bands extending toward the center of the mass are thought to reflect the fibrous stroma. Susceptibility artifact can indicate the presence of hemosiderin.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Concomitant with the variable pathologic appearance of angiosarcoma, imaging features are heterogeneous and nonspecific. The spleen is often enlarged, and when infiltrated by diffuse angiosarcoma, will enhance heterogeneously. Discrete angiosarcomas appear as a large dominant mass or multiple masses. As hemorrhage and cystic necrosis are common, angiosarcoma can appear cystic. Increased flow in the bizarre and dilated vascular channels can be shown at color Doppler US. At CT and MRI, angiosarcomas can be solitary or multiple, poorly defined, heterogeneous nodular masses—due to the presence of both solid components and areas of hemorrhage and necrosis—with heterogeneous nodular enhancement of solid components and vascular channels.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - The term inflammatory pseudotumor (IPT) has been used to describe a reactive tumorlike lesion found throughout the body, rarely in the spleen. Originally described in 1984, splenic IPT is most often seen in middle-aged or older patients . Splenic IPT is strongly associated with Epstein-Barr virus (EBV) infection. IPT-like follicular dendritic cell (FDC) tumor is the most frequent subgroup of EBV-associated IPT. Patients can have a wide range of clinical presentations, from asymptomatic to left upper quadrant or epigastric pain, fever, weight loss, and splenomegaly. Anemia, thrombocytopenia, hypergammaglobulinemia, and elevated levels of inflammatory markers have also been described.
Algorithmic Approach to the Splenic Lesion Based on Radiologic-Pathologic Correlation.
Kim N, Auerbach A, Manning MA.
Radiographics. 2022 May-Jun;42(3):683-701. - Most patients with splenules present have a single splenule varying in size from a few millimeters to centimeters (mean measurement = 1.16 cm). Multiple splenules clustered together are less commonly seen (rarely greater than six per cluster). Splenules appear as ovoid or spherical masses with identical imaging characteristics to those of the orthotopic spleen, and one may be able to identify a branch of the splenic artery supplying the splenule. Splenules are generally asymptomatic and found incidentally. Splenules can also harbor any disease that can involve the spleen itself.”
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - Splenosis is an acquired condition of autotransplanted benign splenic tissue caused by spillage of cells after traumatic or iatrogenic splenic injury. The splenic deposits may be found anywhere throughout the body but most commonly occur in the peritoneal cavity. The splenic deposits can vary in number, size, and shape and can easily be confused with neoplasms, such as lymphoma, mesenchymal tumors, melanoma, and metastatic disease. Unlike congenital splenules, splenosis lesions obtain their blood supply from surrounding tissues and have a poorly formed capsule. Otherwise, splenosis deposits generally have the same imaging characteristics as normal splenic tissue: well-circumscribed homogeneously enhancing lesions at CT and MRI (with increased diffusion restriction), solid round or oval homogeneous hypoechoic masses at US, and lacking high focal fluorodeoxyglucose (FDG) uptake at PET.
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - Torsion of a wandering spleen is a rare but severe complication that can occur when the displaced organ twists on its lax vascular pedicle, causing vascular compromise and infarction. Patients with acute splenic torsion causing ischemia may present with acute abdomen and require emergent surgical intervention. At imaging, a wandering spleen appears as an enlarged spleniform mass in an ectopic location with partial or complete absence of parenchymal enhancement; in some cases, the twisted vascular supply may be evident as a whirl sign, with alternating opaque and lucent bands
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - Heterotaxy syndromes represent a spectrum of abnormal disruption of the typical asymmetric visceral organ positionings in the chest and abdomen along the right and left axes. There are two major subtypes, asplenia and polysplenia, and the degree of organ derangement is variable. It is important to recognize, as these patients can also have associated complex cardiac defects (most severe in asplenia), immune deficiency, and bowel malrotation.
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - The cardiac anomalies associated with polysplenia are typically less severe than those associated with asplenia; they can include endocardial cushion defects, atrial septal defect, ventricular septal defect, and transposition of the great arteries. These patients may also present with immune deficiency, as they are usually functionally hyposplenic despite having multiple spleens. Owing to abnormal splenic development, the accessory splenic tissue is also more susceptible to torsion.
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - Asplenia is right isomerism or bilateral right sidedness. This syndrome is more common in males and is usually diagnosed in childhood, as there is a poor prognosis due to associated severe cardiac anomalies, including large ventricular septal defect, single ventricle, and pulmonic stenosis or atresia). These patients can present with severe cyanosis and respiratory distress due to associated cardiac malformations. Other findings include trilobed lungs, bilateral eparterial bronchi (the main bronchus originates above the ipsilateral pulmonary artery), midline liver, absent gallbladder, intestinal malrotation, and duplication of the inferior vena cava without normal splenic tissue. Since there is congenital absence of the spleen, these patients are prone to life-threatening sepsis
Spectrum of Heterotopic and Ectopic Splenic Conditions.
Nelson LW, Bugenhagen SM, Lubner MG, Bhalla S, Pickhardt PJ.
Radiographics. 2024 Nov;44(11) - “Sarcoidosis is a systemic inflammatory condition characterized by noncaseating granulomas. Its annual incidence ranges from 1 to 15 per 100,000 individuals and is more common in women. While pulmonary and mediastinal lymph node involvement is common, affecting 90% of patients, splenic involvement is reported in close to 24% of cases. Splenic sarcoidosis at CT scans reveals multiple solid hypodense nodules of varying sizes, typically hypoenhancing after contrast administration, often coexisting with hepatic nodules and abdominal lymphadenopathies. Traditionally, these splenic lesions exhibit characteristic MRI features based on disease activity. Inflammatory lesions show high signals on T2 and DWI sequences, whereas fibrous lesions present with low signals on both T1 and T2 sequences.”
Spleen anomalies and lesions in CT and MRI: essentials for radiologists and clinicians—a pictorial review
Herrera-Ortiz AF, Del Castillo V, Aguirre D, et al
Abdom Radiol (NY). 2025 Feb;50(2):860-874. - Splenic metastases are rare and often associated with advanced stages of widespread metastatic disease in melanoma,breast, ovarian, lung, and colon cancers . Splenic metastases are generally multiple, although isolated metastasis has also been reported as an even rarer occurrence. At CT, splenic metastases commonly present as hypodense lesions; nevertheless, their appearance can vary based on the primary tumor, occasionally manifesting as cystic lesions with diverse enhancement patterns. At MRI, splenic metastases present low signal intensity on T1 and high signal intensity on T2 sequences, with a variable degree of contrast enhancement, making their diagnosis challenging.
Spleen anomalies and lesions in CT and MRI: essentials for radiologists and clinicians—a pictorial review
Herrera-Ortiz AF, Del Castillo V, Aguirre D, et al
Abdom Radiol (NY). 2025 Feb;50(2):860-874. - “ Splenic metastases can occur with widespread disease, and parenchymal disease is caused by hematogenous dissemination. The most common primary cancers with splenic metastases include melanoma and cancers of the breast, lung, ovary, stomach, and prostate.”
Nonneoplastic, Benign, and Malignant Splenic Diseases: Cross-Sectional Imaging Findings and Rare Disease Entities
Thipphavong S et al.
AJR 2014;203: 315-322 - Primary splenic angiosarcoma predominantly affects patients in the 6th–7th decade of life, with an annual incidence of one case per 4 million individuals. This aggressive neoplasm displays a predilection for males and has a poor prognosis, marked by a high mortality rate and significant risk of rupture in up to 30% of cases. At CT imaging, splenic angiosarcoma often presents as solitary or multiple poorly defined nodular masses distorting the normal anatomy of the spleen and producing enlargement. The contrast enhancement pattern on CT varies depending on the extent of necrosis within the lesion. Its enhancement is usually centripetal and heterogeneous, and 69–100% of the cases tend to present with metastases, most commonly disseminating to the liver, lungs, adrenals, bones, and lymphatics .
Spleen anomalies and lesions in CT and MRI: essentials for radiologists and clinicians—a pictorial review Andr.s Felipe Herrera‑Ortiz et al.
Herrera-Ortiz AF, Del Castillo V, Aguirre D, et al
Abdom Radiol (NY). 2025 Feb;50(2):860-874. - “Practice Recommendation: Splenic artery aneurysms ≥2 cm, or any aneurysm with features suspicious for a pseudoaneurysm should be referred to interventional radiology (or other endovascular specialist) for consideration of treatment. Aneurysms <2 cm can be followed for growth annually with CT or MR angiography, with discontinuation of follow-up made in consultation with a vascular specialist after a period of ongoing stability.”
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - “Practice Recommendation: A single measurement of >13 cm in maximal diameter is recommended to screen for splenomegaly in adults, recognizing that the positive predictive value for disease has not been determined. Volume calculations can be reserved for when more accuracy is required.”
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - Although medical calculators can diagnose splenomegaly by correcting for body size, they are cumbersome, requiring knowledge of the patient’s height, weight, and gender.6 Although the literature suggests that splenic volume calculation may represent the future of spleen measurement, other studies showing a close correlation between a single largest measurement and total spleen volume favour continuing with the current status quo of providing a single value to represent spleen size. Volume calculations can be referenced to body size when more accuracy is required, particularly to avoid overdiagnosing splenomegaly in larger patients.
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - “Lymphoma is the most common malignancy of the spleen, either primary or part of diffuse systemic disease.12,15 Splenic involvement occurs in approximately 33% of patients with Hodgkins and 30% to 40% in patients with non-Hodgkins Lymphoma.16 Lymphoma can present in many forms including splenomegaly, diffuse nodules (either in a miliary pattern or larger nodules), or a solitary mass.12 Primary splenic lymphoma confined only to the spleen ± perisplenic nodes is very rare, comprising less than 1% of cases, and most patients will present with constitutional symptoms.”
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - Splenic incidental findings are defined as lesions detected on imaging in the spleen not related to the clinical history. Incidental splenic lesions are less common than in other organs such as liver or kidneys, but increased demand for imaging means that their frequency is rising. Incidental focal splenic lesions have a wide range of etiologies, ranging from common benign diagnoses (cysts, granulomas, and hemangiomas) to lymphoma or metastases to exceedingly rare primary malignancies such as angiosarcoma. In one study, 1.5% of trauma patients with CT had an incidental splenic lesion and the vast majority are benign. Benign lesions are almost always asymptomatic, whereas malignant lesions are very rarely entirely incidental or a solitary isolated finding.
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - Practice Recommendation: If an incidental isolated indeterminate splenic mass is found on CT or MR in a patient with no history of malignancy or symptoms, it is unlikely to be clinically significant, and no further evaluation or follow-up is necessary.
Practice Recommendation: In patients with constitutional symptoms (fever, weight loss, night sweats), epigastric or left upper quadrant pain, or a history of prior malignancy, the risk of malignancy is low but not negligible. An incidental indeterminate splenic lesion should be further evaluated with MRI, PET/CT, or biopsy, especially if it may affect patient management.
CAR Recommendations for the Management of Incidental Findings of the Spleen and Nodes in Adults
Jeffery R. Bird, Gary L. Brahm, Christopher I. Fung et al.
Canadian Association of Radiologists Journal1–10 (2025) - “Fortunately, most incidental splenic lesions are benign and clinically insignificant. Although supporting literature is sparse, the white paper on incidental findings by the American College of Radiology (ACR) suggests that splenic lesions smaller than 1 cm in a patient without malignancy are likely benign and do not warrant further imaging. Primary malignant splenic lesions are extremely rare and tend to grow rapidly, whereas benign lesions grow at a rate of less than 3 mm per year. Therefore, small size and stability over 1 year are considered excellent markers for benignity in splenic lesions.”
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - “The primary objective of imaging and classification is not necessarily to pinpoint an exact diagnosis but to differentiate between indolent and potentially aggressive lesions, guiding appropriate follow-up or intervention on the basis of malignancy risk.”
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - Before a lesion is classified as incidental, clinical factors and concomitant nonsplenic imaging findings must exclude the possibility of acute hematoma, abscess, or malignancy. Acute splenic hematomas are almost always the result of recent blunt trauma and should be considered in patients with a compatible history. Abscesses typically present with fever, abdominal pain, chills, and leukocytosis. Microabscesses, often fungal in origin, are encountered in patients who are immunocompromised because of chemotherapy, chronic steroid use, organ transplant, or HIV. Imaging reveals numerous similarly sized small lesions, usually accompanied by hepatic involvement.
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - Lymphoma is the most common malignancy involving the spleen. Symptoms include left upper quadrant pain and constitutional systemic symptoms like weight loss, malaise, and fever . As the majority of patients have disseminated lymphoma at presentation, concomitant lymphadenopathy aids in the diagnosis . Nevertheless, primary splenic lymphoma can rarely occur, making up less than 1% of all lymphomas, and is usually non-Hodgkin type.
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - Isolated splenic lymphoma presents as solitary or multiple nodules. The nodules typically appear hypoattenuating on CT, hypointense on T1-weighted MRI, and are often hypointense or isointense on T2-weighted MRI. They show hypoenhancement, with restricted diffusion relative to surrounding parenchyma. Given the normally heterogeneous parenchymal enhancement in the arterial phase, hypoenhancing lesions are frequently better detected in the late venous or equilibrium phase of postcontrast imaging. In the absence of suspicious imaging features, these lesions may be indistinguishable from other solid lesions. Lymphoma will show avid FDG uptake on PET/CT, which may be useful in staging. When FDG avidity is absent, lymphoma is unlikely.
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - Although angiosarcoma is the most common primary splenic malignancy, it remains extremely rare with an annual incidence of 0.14–0.25 cases per million. Predominantly affecting middle-aged to elderly patients, angiosarcomas typically present with abdominal pain and weight loss; splenomegaly is common, with splenic rupture reported in up to 30%. Imaging typically shows aggressive features such as irregular shape and ill-defined margins, and it can present as a large dominant lesion or multiple nodules. On MRI, lesions are typically T1 hypointense and T2 hyperintense; however, the presence of any of necrosis, hemorrhage, hemosiderin, or calcifications can lead to signal heterogeneity. Areas of hemorrhage may cause T1 hyperintensity, whereas hemosiderin or calcifications may result in T2 hypointensity. After IV contrast administration, masses are usually heterogeneously enhancing. This heterogeneityis greater than seen in hemangiomas and may be so pronounced that the mass appears cystic. Tumors can infiltrate beyond the splenic capsule and present with local or distant metastases.
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - “The proposed Sp system presented here is based on expert consensus and has not yet undergone formal validation. Future studies are needed to assess its interreader reliability, diagnostic accuracy, and clinical utility. Despite these limitations, the framework provides a starting point to guide radiologists in evaluating and managing incidental splenic lesions, with the ultimate goal of improving patient care.”
Incidental Splenic Lesions: A Proposed Algorithm for Assessment and Management
Myles T. Taffel, MD1, Julie Y. An, MD1, Frank H. Miller et al.
Roentgen Ray Rev 2025; 1:e2401039 - BACKGROUND. Splenomegaly historically has been assessed on imaging by use of potentially inaccurate linear measurements. Prior work tested a deep learning artificial intelligence (AI) tool that automatically segments the spleen to determine splenic volume.
OBJECTIVE. The purpose of this study is to apply the deep learning AI tool in a large screening population to establish volume-based splenomegaly thresholds.
METHODS. This retrospective study included a primary (screening) sample of 8901 patients (4235 men, 4666 women; mean age, 56 Å} 10 [SD] years) who underwent CT colonoscopy (n = 7736) or renal donor CT (n = 1165) from April 2004 to January 2017 and a secondary sample of 104 patients (62 men, 42 women; mean age, 56 Å} 8 years) with endstage liver disease who underwent contrast-enhanced CT performed as part of evaluation for potential liver transplant from January 2011 to May 2013. The automated deep learning AI tool was used for spleen segmentation, to determine splenic volumes.
Automated Deep Learning Artificial Intelligence Tool for Spleen Segmentation on CT: Defining Volume-Based Thresholds for Splenomegaly
Alberto A. Perez, Victoria Noe-Kim, Meghan G. Lubner, et al.
AJR 2023; 221:1–9 - RESULTS. In 8853 patients included in analysis of splenic volumes (i.e., excluding a value of 0 mL or error values), the mean automated splenicvolume was 216 Å} 100 [SD] mL. The weight-based volumetric threshold (expressed in milliliters) for splenomegaly was calculated as (3.01 Å~ weight [expressed as kilograms]) + 127; for weight greater than 125 kg, the splenomegaly threshold was constant (503 mL). Sensitivity and specificity for volume-defined splenomegaly were 13% and 100%, respectively, at a true craniocaudal length of 13 cm, and 78% and 88% for a maximum 3D length of 13 cm. In the secondary sample, both observers identified segmentation failure in one patient. The mean automated splenic volume in the 103 remaining patients was 796 Å} 457 mL; 84% (87/103) of patients met the weight-based volume-defined splenomegalythreshold. CONCLUSION. We derived a weight-based volumetric threshold for splenomegaly using an automated AI-based tool. CLINICAL IMPACT. The AI tool could facilitate large-scale opportunistic screening for splenomegaly.
Automated Deep Learning Artificial Intelligence Tool for Spleen Segmentation on CT: Defining Volume-Based Thresholds for Splenomegaly
Alberto A. Perez, Victoria Noe-Kim, Meghan G. Lubner, et al.
AJR 2023; 221:1–9 - Key Finding
A previously tested automated deep learning AI tool was used to calculate splenic volumes from the CT examinations of 8853 patients from an outpatient screening population. Splenic volume was most strongly associated with weight among a range of patient factors, and a weight-based volume-defined threshold for splenomegaly was derived.
Importance
Use of the automated deep learning AI tool and weightbased volumetric thresholds could allow large-scale evaluation for splenomegaly on CT examinations performed for any indication.
Automated Deep Learning Artificial Intelligence Tool for Spleen Segmentation on CT: Defining Volume-Based Thresholds for Splenomegaly
Alberto A. Perez, Victoria Noe-Kim, Meghan G. Lubner, et al.
AJR 2023; 221:1–9 - “In conclusion, we derived a simple weight-based volumetric threshold for determining the presence of splenomegaly using an automated AI-based tool for determining splenic volume from CT examinations. Standard linear splenic measurements (which historically have been used as a surrogate for splenic volume) had suboptimal performance in detecting volume-based splenomegaly, and the weight-based volumetric thresholds indicated the presence of splenomegaly in most patients who underwent pre–liver transplant CT. The AI tool could be applied for more robust evaluation for splenomegaly in comparison with linear measurements as well as for large-scale opportunistic screening for splenomegaly.”
Automated Deep Learning Artificial Intelligence Tool for Spleen Segmentation on CT: Defining Volume-Based Thresholds for Splenomegaly
Alberto A. Perez, Victoria Noe-Kim, Meghan G. Lubner, et al.
AJR 2023; 221:1–9
Trauma
- “CT angiography (CTA) of the aortoiliofemoral (AIF) arteries in the abdomen, pelvis, and lower extremities has become an invaluable tool in assessment of patients with peripheral arterial disease (PAD) and lower extremity trauma. AIF CTA provides rapid and comprehensive assessment of arterial inflow and outflow, guiding management of patients with chronic claudication and those with more acute manifestations, including atherothrombotic occlusion, embolic disease, or thrombosis of prior interventions such as bypass graft or stent placement.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Emboli should be considered with multiple occlusions at vascular branch points where emboli tend to lodge, particularly in the absence of other atherosclerotic disease or in younger patients. Systemic infarcts of the spleen, kidneys, and less commonly liver and bowel provide another clue to the presence of embolic disease.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Clinically significant extravascular findings are reportedly present in up to 15% of AIF CTA examinations, as patients are often elderly or have multiple comorbidities that have led to their PAD. Performing a standard abdomen and pelvis (or chest, abdomen, and pelvis) search pattern before dedicated vascular evaluation lessens the likelihood of satisfaction of search in answering the clinical question and may provide clues to unify a vascular diagnosis, such as splenic or renal infarcts in the setting of embolic disease.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Apart from PAD and traumatic injury, AIF CTA also serves a role in infectious and inflammatory disorders, such as vascular graft–related infections, Buerger disease, and cystic adventitial disease. Preoperative planning for reconstructive oral surgery using a fibular free flap or breast reconstruction using an upper thigh flap can be accomplished using lower extremity CTA. The AIF CTA protocol can be used as a building block for vascular imaging protocols that also imagethe chest or are designed to evaluate endovascular aortic repair (EVAR) for even more comprehensive evaluation.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “The abdominal aorta bifurcates into the common iliac arteries, which then bifurcate into the internal and external iliac arteries. Anatomic detail of the internal iliac arteries is beyond the scope of this article, and a detailed review can be found by Raniga et al. The transition from the external iliac artery to the common femoral artery occurs at the inguinal ligament. A more useful anatomic landmark at CTA are distal branches of the external iliac artery, such as the deep inferior epigastric and circumflex iliac arteries, which can be used as an adjunct to demarcate the start of the common femoral artery.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Protocols may routinely call for performing an additional delayed acquisition from the level of the knees through the toes, about 15–30 seconds after the conclusion of the initial acquisition, adjusting the start position higher if aware of prior interventions such as bypass graft or stent placement. When imaging occurs too late relative to arrival of the contrast material bolus or in patients with impaired venous outflow, venous contamination may occur, whereby the veins are already opacified, making identification of small arteries in the calves more difficult and potentially resulting in diagnostic uncertainty.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Material decomposition can also be used to create virtual noncalcium images and is commercially available . For example, PureLumen (Siemens Healthineers) allows subtraction of calcifications from the walls, which may be helpful in evaluating heavily calcified vessels, particularly smaller vessels such as the calf arteries. It should be noted that virtual noncalcium or virtual calcium-only images should be interpreted in the context of the source images, as the material decomposition is prone to artifact; for instance, in the setting of virtual noncalcium imaging, a portion of the contrast-enhanced lumen may be subtracted, leading to apparent occlusion .
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Virtual monoenergetic imaging (VMI) uses material decomposition and computation of the iodine attenuation to model data as if a single user-selected photon energy had been used. Low kiloelectron voltage reconstructed VMI (in the range of 40–55 keV) offers higher contrast-tonoise ratio than polychromatic images and can facilitate lower intravenous contrast material dose protocols, which may be of benefit in patients with chronic kidney disease, or salvage examinations, in which contrast-to-noise ratio within the vasculature may have been inadequate at polychromatic imaging. In addition, PCD CT is able to equally weight lower-energy photons, allowing higher image contrast for iodine contrast-enhanced examinations.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - For stenosis, the degree of narrowing should be identified and can be graded as follows: normal (no stenosis), mild (1%– 49% stenosis), moderate (50%–74% stenosis), severe (75%– 99% stenosis), and occluded (100%) (38). Delayed imaging is also paramount to interrogate, as apparent “thrombosis” may be correctly identified as outrunning of the contrast material bolus in the delayed acquisition. Clues to outrunning of the contrast material bolus as opposed to thrombosis include gradual diminution of the contrast enhancement, similar findings in adjacent vessels, or similar findings in the contralateral leg.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Embolic disease is a less common cause of thrombosis than atherosclerotic disease (up to 15% of cases) but remains critical to consider when the patient history or imaging findings are suggestive of the diagnosis. Emboli should be considered with multiple occlusions at vascular branch points where emboli tend to lodge, particularly in the absence of other atherosclerotic disease or in younger patients. Systemic infarcts of the spleen, kidneys, and less commonlyliver and bowel provide another clue to the presence of embolic disease. In comparison with atherosclerotic occlusion, embolic occlusion is more likely to result in acute limb ischemia, given the absence of well-formed collateral circulation.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Peripheral arterial aneurysms are defined as aneurysms outside the aortoiliac system or brain, with similar risk factors as PAD, and can be considered another manifestation of PAD. Popliteal artery aneurysms (PAAs) are the most common peripheral aneurysm (up to 70% of cases) and are defined as a greater than 50% increase in diameter from normal, which ranges from 5 to 10 mm. Screening of the contralateral leg and abdominal aorta is indicated for patients with an incidentally detected PAA. Repair is advocated for PAAs greater than 20 mm to mitigate the risks of an untreated PAA, which include expansion, acute limb ischemia from thrombosis with distal or proximal propagation of clot, and rupture.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - A survey for active bleeding should include not only the major vessels and their branches but the musculature and around fracture sites, as smaller branch vessels and perforators may also be injured. Care should be taken to avoid overcalling active bleeding when early venous enhancement from muscular venous branches adjacent to an artery can mimic the appearance of a vascular injury. The summation of CTA findings helps contribute to medical or surgical decision making in regard to limb salvage.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Pseudoaneurysms are false aneurysms resulting from a vascular insult such as trauma, surgery, or infection, where extravascular blood is contained typically only by adventitia or surrounding soft tissue or hematoma, rather than the normal three vascular layers (intima, media, and adventitia). At AIF CTA, pseudoaneurysms are most accurately distinguished from active bleeding by maintaining the same size and shape between arterial phase and delayed imaging, in distinction to active bleeding, in which the extravascular contrast material will enlarge or diffuse within adjacent tissues. Pseudoaneurysms are typically globular or rounded but may be complex and multicompartmental, with the potential for partial thrombosis from turbulent flow.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - An arteriovenous fistula (AVF) is an abnormal communication between an artery and a vein outside of the capillary bed. Causes of lower extremity AVF are primarily traumatic and iatrogenic, with developmental AVF or arteriovenous malformations being less common. CTA features include early venous opacification to a relatively similar degree as that of the adjacent artery and expansion of the draining vein. Detection of these signs should lead to careful scrutiny for identifying the actual communication between the artery and vein.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Clinically significant extravascular findings are reportedly present in up to 15% of AIF CTA examinations, as patients are often elderly or have multiple comorbidities that have led to their PAD. Performing a standard abdomen and pelvis (or chest, abdomen, and pelvis) search pattern before dedicated vascular evaluation lessens the likelihood of satisfaction of search in answering the clinical question and may provide clues to unify a vascular diagnosis, such as splenic or renal infarcts in the setting of embolic disease. The pulmonary arteries are usually well opacified at AIF CTA and should be scrutinized for pulmonary emboli at the lung bases.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272
Vascular
- “CT angiography (CTA) of the aortoiliofemoral (AIF) arteries in the abdomen, pelvis, and lower extremities has become an invaluable tool in assessment of patients with peripheral arterial disease (PAD) and lower extremity trauma. AIF CTA provides rapid and comprehensive assessment of arterial inflow and outflow, guiding management of patients with chronic claudication and those with more acute manifestations, including atherothrombotic occlusion, embolic disease, or thrombosis of prior interventions such as bypass graft or stent placement.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Emboli should be considered with multiple occlusions at vascular branch points where emboli tend to lodge, particularly in the absence of other atherosclerotic disease or in younger patients. Systemic infarcts of the spleen, kidneys, and less commonly liver and bowel provide another clue to the presence of embolic disease.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Clinically significant extravascular findings are reportedly present in up to 15% of AIF CTA examinations, as patients are often elderly or have multiple comorbidities that have led to their PAD. Performing a standard abdomen and pelvis (or chest, abdomen, and pelvis) search pattern before dedicated vascular evaluation lessens the likelihood of satisfaction of search in answering the clinical question and may provide clues to unify a vascular diagnosis, such as splenic or renal infarcts in the setting of embolic disease.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Apart from PAD and traumatic injury, AIF CTA also serves a role in infectious and inflammatory disorders, such as vascular graft–related infections, Buerger disease, and cystic adventitial disease. Preoperative planning for reconstructive oral surgery using a fibular free flap or breast reconstruction using an upper thigh flap can be accomplished using lower extremity CTA. The AIF CTA protocol can be used as a building block for vascular imaging protocols that also imagethe chest or are designed to evaluate endovascular aortic repair (EVAR) for even more comprehensive evaluation.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “The abdominal aorta bifurcates into the common iliac arteries, which then bifurcate into the internal and external iliac arteries. Anatomic detail of the internal iliac arteries is beyond the scope of this article, and a detailed review can be found by Raniga et al. The transition from the external iliac artery to the common femoral artery occurs at the inguinal ligament. A more useful anatomic landmark at CTA are distal branches of the external iliac artery, such as the deep inferior epigastric and circumflex iliac arteries, which can be used as an adjunct to demarcate the start of the common femoral artery.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - “Protocols may routinely call for performing an additional delayed acquisition from the level of the knees through the toes, about 15–30 seconds after the conclusion of the initial acquisition, adjusting the start position higher if aware of prior interventions such as bypass graft or stent placement. When imaging occurs too late relative to arrival of the contrast material bolus or in patients with impaired venous outflow, venous contamination may occur, whereby the veins are already opacified, making identification of small arteries in the calves more difficult and potentially resulting in diagnostic uncertainty.”
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Material decomposition can also be used to create virtual noncalcium images and is commercially available . For example, PureLumen (Siemens Healthineers) allows subtraction of calcifications from the walls, which may be helpful in evaluating heavily calcified vessels, particularly smaller vessels such as the calf arteries. It should be noted that virtual noncalcium or virtual calcium-only images should be interpreted in the context of the source images, as the material decomposition is prone to artifact; for instance, in the setting of virtual noncalcium imaging, a portion of the contrast-enhanced lumen may be subtracted, leading to apparent occlusion .
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Virtual monoenergetic imaging (VMI) uses material decomposition and computation of the iodine attenuation to model data as if a single user-selected photon energy had been used. Low kiloelectron voltage reconstructed VMI (in the range of 40–55 keV) offers higher contrast-tonoise ratio than polychromatic images and can facilitate lower intravenous contrast material dose protocols, which may be of benefit in patients with chronic kidney disease, or salvage examinations, in which contrast-to-noise ratio within the vasculature may have been inadequate at polychromatic imaging. In addition, PCD CT is able to equally weight lower-energy photons, allowing higher image contrast for iodine contrast-enhanced examinations.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - For stenosis, the degree of narrowing should be identified and can be graded as follows: normal (no stenosis), mild (1%– 49% stenosis), moderate (50%–74% stenosis), severe (75%– 99% stenosis), and occluded (100%) (38). Delayed imaging is also paramount to interrogate, as apparent “thrombosis” may be correctly identified as outrunning of the contrast material bolus in the delayed acquisition. Clues to outrunning of the contrast material bolus as opposed to thrombosis include gradual diminution of the contrast enhancement, similar findings in adjacent vessels, or similar findings in the contralateral leg.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Embolic disease is a less common cause of thrombosis than atherosclerotic disease (up to 15% of cases) but remains critical to consider when the patient history or imaging findings are suggestive of the diagnosis. Emboli should be considered with multiple occlusions at vascular branch points where emboli tend to lodge, particularly in the absence of other atherosclerotic disease or in younger patients. Systemic infarcts of the spleen, kidneys, and less commonlyliver and bowel provide another clue to the presence of embolic disease. In comparison with atherosclerotic occlusion, embolic occlusion is more likely to result in acute limb ischemia, given the absence of well-formed collateral circulation.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Peripheral arterial aneurysms are defined as aneurysms outside the aortoiliac system or brain, with similar risk factors as PAD, and can be considered another manifestation of PAD. Popliteal artery aneurysms (PAAs) are the most common peripheral aneurysm (up to 70% of cases) and are defined as a greater than 50% increase in diameter from normal, which ranges from 5 to 10 mm. Screening of the contralateral leg and abdominal aorta is indicated for patients with an incidentally detected PAA. Repair is advocated for PAAs greater than 20 mm to mitigate the risks of an untreated PAA, which include expansion, acute limb ischemia from thrombosis with distal or proximal propagation of clot, and rupture.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - A survey for active bleeding should include not only the major vessels and their branches but the musculature and around fracture sites, as smaller branch vessels and perforators may also be injured. Care should be taken to avoid overcalling active bleeding when early venous enhancement from muscular venous branches adjacent to an artery can mimic the appearance of a vascular injury. The summation of CTA findings helps contribute to medical or surgical decision making in regard to limb salvage.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Pseudoaneurysms are false aneurysms resulting from a vascular insult such as trauma, surgery, or infection, where extravascular blood is contained typically only by adventitia or surrounding soft tissue or hematoma, rather than the normal three vascular layers (intima, media, and adventitia). At AIF CTA, pseudoaneurysms are most accurately distinguished from active bleeding by maintaining the same size and shape between arterial phase and delayed imaging, in distinction to active bleeding, in which the extravascular contrast material will enlarge or diffuse within adjacent tissues. Pseudoaneurysms are typically globular or rounded but may be complex and multicompartmental, with the potential for partial thrombosis from turbulent flow.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Vasospasm is characterized by transient arterial constriction reducing blood flow. Although this may be seen in the setting of Raynaud disease, vasculitis, or toxic exposures, trauma may also induce vasospasm in the lower extremities, resulting in a thready beaded appearance of the vessels with multifocal short or medium-length segmental narrowing. Vasospasm can be distinguished from vascular injury by a return to normal caliber at repeat short-term imaging in 48–72 hours. Although vasospasm may result in high-grade narrowing of the vessel lumen, it should not be completely occlusive, in contrast to thrombosis or emboli.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - An arteriovenous fistula (AVF) is an abnormal communication between an artery and a vein outside of the capillary bed. Causes of lower extremity AVF are primarily traumatic and iatrogenic, with developmental AVF or arteriovenous malformations being less common. CTA features include early venous opacification to a relatively similar degree as that of the adjacent artery and expansion of the draining vein. Detection of these signs should lead to careful scrutiny for identifying the actual communication between the artery and vein.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Buerger disease, also known as thromboangiitis obliterans, is a nonnecrotizing inflammatory vasculitis affecting small and medium-sized vessels, most commonly of the extremities. Diagnostic criteria include smoking history (very highly associated), onset before age 50 years, infrapopliteal arterial involvement, absence of atherosclerotic risk factors apart from smoking, and either upper limb involvement or migratory thrombophlebitis . Although there are no pathognomonic imaging features, features that are suggestive include segmental occlusion of distal vascular segments with intervening normal segments and the presence of corkscrew-appearing collateral vessels around the occluded segments. Other entities that could have a similar appearance include PAD and connective tissue diseases, including systemic lupus erythematosus, and scleroderma.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272 - Clinically significant extravascular findings are reportedly present in up to 15% of AIF CTA examinations, as patients are often elderly or have multiple comorbidities that have led to their PAD. Performing a standard abdomen and pelvis (or chest, abdomen, and pelvis) search pattern before dedicated vascular evaluation lessens the likelihood of satisfaction of search in answering the clinical question and may provide clues to unify a vascular diagnosis, such as splenic or renal infarcts in the setting of embolic disease. The pulmonary arteries are usually well opacified at AIF CTA and should be scrutinized for pulmonary emboli at the lung bases.
Aortoiliofemoral Lower Extremity CT Angiography
Anup S. Shetty • Mark J. Hoegger • Mohamed Z. Rajput
RadioGraphics 2025; 45(10):e240272