google ads
Search

Everything you need to know about Computed Tomography (CT) & CT Scanning

August 2022 Imaging Pearls - Educational Tools | CT Scanning | CT Imaging | CT Scan Protocols - CTisus
Imaging Pearls ❯ August 2022

-- OR --

3D and Workflow

  • “The Surgeon General has identified medical misinformation as a major public health threat, and many professional societies, including the American Medical Association, have called for action to combat it.”
    Physicians Spreading Misinformation on Social Media — Do Right and Wrong Answers Still Exist in Medicine?
    Richard J. Baron, Yul D. Ejnes
    n engl j med 387;1 nejm.org July 7, 2022
  • "There aren’t always right answers, but some answers are clearly wrong.”
    Physicians Spreading Misinformation on Social Media — Do Right and Wrong Answers Still Exist in Medicine?
    Richard J. Baron, Yul D. Ejnes
    n engl j med 387;1 nejm.org July 7, 2022
  • "Medicine has a truth problem. In the era of social media and heavily politicized science, “truth” is increasingly crowdsourced: if enough people like, share, or choose to believe something, others will accept it as true. This way of determining “truth” doesn’t involve scientific methods; it relies instead on “the wisdom of crowds,” which has particular power in a democratic society in which leaders and policies are chosen by the will of the group. Such choices anchor concepts like freedom and liberty. But they may not be helpful in determining whether a building will collapse, whether your brakes will stop your car — or whether a medication or vaccine works.”
    Physicians Spreading Misinformation on Social Media — Do Right and Wrong Answers Still Exist in Medicine?
    Richard J. Baron, Yul D. Ejnes
    n engl j med 387;1 nejm.org July 7, 2022
Cardiac

  •  • Serial cardiac troponin (cTn) biomarkers, preferably high-sensitivity cardiac troponin (hs-cTn), are useful for rapid detection and exclusion of myocardial injury (class 1 strength of recommendation; level B-NR quality of evidence [nonrandomized]).
    • Structured risk assessment and evidence-based clinical decision pathways(CDPs) should be used to facilitate disposition and guide diagnostic evaluation (class 1 strength; level B-NR quality).
    • Low-risk patients with acute or stable chest pain may be discharged home without urgent cardiac testing (class 2a strength for acute chest pain, class 1 strength for stable chest pain; level B-R quality [randomized]).
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022
  • • For intermediate-risk patients with acute chest pain and no known coronary artery disease (CAD), coronary computed tomographic angiography (CCTA) is useful for exclusion of atherosclerotic plaque and obstructive CAD (class I strength; level A quality).
    • For intermediate-risk patients with acute chest pain and no known CAD, functional testing (eg, exercise electrocardiography, stress echocardiography, stress positron emission tomography/single-photon emission computed tomography myocardial perfusion imaging, or stress cardiac magnetic resonance) is useful for diagnosis of myocardial ischemia (class I strength; level B-NR quality [nonrandomized]).
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022
  • • For intermediate-risk patients with acute chest pain and no known CAD, functional testing (eg, exercise electrocardiography, stress echocardiography, stress positron emission tomography/single-photon emission computed tomography myocardial perfusion imaging, or stress cardiac magnetic resonance) is useful for diagnosis of myocardial ischemia (class I strength; level B-NR quality[nonrandomized]).
    • Clinically stable patients presenting with chest pain should be included in decision-making. Information about risk of adverse events, radiation exposure, costs, and alternative options should be provided to facilitate the discussion.
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022
Chest

  • “The true mortality associated with undiagnosed pulmonary embolism is estimated to be less than 5%, but recovery from pulmonary embolism is associated with complications such as bleeding due to anticoagulant treatment, recurrent venous thromboembolism, chronic thromboembolic pulmonary hypertension, and long-term psychological distress. Approximately half the patients who receive a diagnosis of pulmonary embolism have functional and exercise limitations 1 year later (known as post–pulmonary-embolism syndrome), and the health-related quality of life for patients with a history of pulmonary embolism is diminished as compared with that of matched controls. Therefore, the timely diagnosis and expert management of pulmonary embolism are important.”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.

  • Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • • Pulmonary embolism is a common diagnosis and can be associated with recurrent venous thromboembolism, bleeding due to anticoagulant therapy, chronic thromboembolic pulmonary hypertension, and long-term psychological distress.
    • A minority of patients who are evaluated for possible pulmonary embolism benefit from chest imaging (e.g., computed tomography).
    • Initial treatment is guided by classification of the pulmonary embolism as high-risk, intermediate-risk, or low-risk. Most patients have low-risk pulmonary embolism, and their care can be managed at home with a direct oral anticoagulant.
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • • Patients with acute pulmonary embolism should receive anticoagulant therapy for at least 3 months. The decision to continue treatment indefinitely depends on whether the associated reduction in the risk of recurrent venous thromboembolism outweighs the increased risk of bleeding and should take into account patient preferences.
    • Patients should be followed longitudinally after an acute pulmonary embolism to assess for dyspnea or functional limitation, which may indicate the development of post–pulmonary-embolism syndrome or chronic thromboembolic pulmonary hypertension.
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • “Occult cancer is detected in 5.2% of patients within 1 year after a diagnosis of unprovoked pulmonary embolism. An extensive screening strategy may detect more cancers than limited screening, but data are limited as to whether such screening is associated with better patient outcomes. Experts recommend limited cancer screening guided by medical history, physical examination, basic laboratory tests and chest radiographs, and age-specific and sex-specific cancer screening.”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • "Appropriate management of subsegmental pulmonary embolism (a single isolated subsegmental pulmonary embolus or multiple emboli, without the presence of pulmonary embolism in segmental or more proximal pulmonary vessels and without deep-vein thrombosis in the legs) is uncertain. Although some guidelines suggest clinical surveillance instead of anticoagulation in patients with low-risk subsegmental pulmonary embolism, a recent prospective cohort study involving such patients who were treated without anticoagulation therapy showed a higher-than expected  incidence of recurrent venous thromboembolism during 90-day follow-up A randomized, placebo-controlled trial of clinical surveillance as compared with anticoagulation in this patient population is ongoing (ClinicalTrials.gov number, NCT04263038).”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
Colon

  • "Rosuvastatin (Crestor, AstraZeneca) is a commonly used drug for managing hypercholesterolaemia. It is a very safe medication with mostly acceptable side effects. Rare but serious side effects are not well known. A 64-year-old woman presented with bloody diarrhoea after starting rosuvastatin for hypercholesterolaemia. Stool microscopy and culture ruled out infective causes. Abdominal CT scan revealed normal calibre celiac axis and superior mesenteric artery. Colonoscopic biopsy revealed ischaemic colitis as the final histological diagnosis. The patient is in complete remission after ceasing the medication. Rosuvastatin causing ischaemic colitis should be considered a rare but serious adverse drug reaction.”
    Adverse drug reaction: Rosuvastatin as a cause for ischaemic colitis in a 64-year-old woman
    Tan J et al.
    BMJ Case Reports 2012
  • “In conclusion, statin use was associated with a lower risk of CD, consistent with one previous observational analysis. Notwithstanding their inherent limitations, additional adequately-powered observational studies examining statin use and IBD risk would be of value. Although there is evidence to suggest the existence of a preclinical phase in CD,our ability to identify those at risk of IBD, in whom statin use could avert disease progression, remains limited. Further studies focusing on the role of statins in IBD progression are also therefore warranted.”
    Association Between Statin Use and Inflammatory Bowel Diseases: Results from a Swedish, Nationwide, Population-based Case-control Study,  
    Paul Lochhead et al.  
    Journal of Crohn's and Colitis, Volume 15, Issue 5, May 2021, Pages 757–765
  • Introduction: Nonocclusive Ischemic Colitis has become more prevalent and frequently of uncertain etiology. Patients infrequently experience recurrence. The author cared for a 63 year old female who had 4 episodes over a 14 month period. She was hypertensive but otherwise in good health. Review of her medications was remarkable for pravastatin use for more than 10 years. Because isolated cases have been reported while on rosuvastatin, it was elected to discontinue the pravastatin and perform a case control .
    Conclusion: This study demonstrates that statin use is a risk factor for ischemic colitis. The pathway of causality is unclear. Statins have been shown to enhance barro receptor sensitivity and therefore could lead to hypotension. The medications could induce hypercoaguability.  
    Increased Risk of Non-Occlusive Ischemic Colitis With Statin Use: A Case Control Study
    Goldberg, Neil
    American Journal of Gastroenterology: October 2018 - Volume 113 - Issue - p S1527
Deep Learning

  • Background Emerging blood-based multi-cancer early detection (MCED) tests can detect a variety of cancer types across stages with a range of sensitivity, specificity, and ability to predict the origin of the cancer signal. However, little is known about the general US population’s preferences for MCED tests.
    Objective To quantify preferences for MCED tests among US adults aged 50–80 years using a discrete choice experiment (DCE).
    Conclusions While there is significant heterogeneity in cancer screening preferences, the majority of participants preferred MCED screening and the accuracy of these tests is important. While the majority of participants preferred adding an MCED test to complement current cancer screenings, the latent class analyses identified a small (16%) and specific subset of individuals who value attributes differently, with particular concern regarding false-negative and false-positive test results, who are significantly less likely to opt-in.  
    Patient Preferences for Multi‑Cancer Early Detection (MCED) Screening Tests
    Heather Gelhorn et al.
    The Patient - Patient-Centered Outcomes Research https://doi.org/10.1007/s40271-022-00589-5
  • “Offering an MCED screening test as part of the standard of care to individuals between the ages of 50 and 80 years is likely to be well received by the majority of this population. Based on the results of the current study, this could represent a viable approach to population-based cancer screening.”  
    Patient Preferences for Multi‑Cancer Early Detection (MCED) Screening Tests
    Heather Gelhorn et al.
    The Patient - Patient-Centered Outcomes Research https://doi.org/10.1007/s40271-022-00589-5
  • "Recent advances in Artificial Intelligence (AI) indicate that AI has the potential to enhance how cancer is studied, diagnosed, and treated. In the near future, AI may be able to predict certain clinical outcomes, such as a patient’s response to anti-cancer drugs or combinations of such drugs. Analysis of large datasets using AI may also help discover novel cancer mechanisms, novel biomarkers of therapy response, or uncover novel therapeutic targets in cancer models and cancer patients . For example, tumor cells can be imaged directly in tissue or after being cultured and treated with pharmacological agents, then analyzed using deep-learning tools to unearth features associated with drug response or disease processes such as metastasis. Automatically integrating several disparate data types obtained from patient data, such as radiology images and molecular profiles of blood, may allow AI to improve patient diagnoses and detect cancer earlier than currently possible.”
    Artificial intelligence in oncology: From bench to clinic  
    Jamal Elkhader, Olivier Elemento
    Seminars in Cancer Biology 84 (2022) 113–128 
  • "There are however numerous challenges in developing accurate AI models and implementing them in the research and/or clinical setting. The datasets used to train AI models have started falling under more scrutiny than before as implicit biases in some training datasets have become apparent, the most important of which being lack of ethnic diversity and under-representation of certain groups, e.g. African Americans. Another related challenge is the relative scarcity of datasets that can be used as external validation of AI models, owing to privacy concerns and competition between medical centers that limit data sharing; this in turn hampers the vetting necessary to adopt AI models in clinical environments. Here we review recent advances and opportunities in AI and data science applied to cancer and the challenges AI will need to overcome to thrive in the laboratory and the clinic.”  
    Artificial intelligence in oncology: From bench to clinic  
    Jamal Elkhader, Olivier Elemento
    Seminars in Cancer Biology 84 (2022) 113–128 
  • “Machine learning, and especially deep learning, have been used to dramatically enhance research in field of digital pathology. One application of deep learning is the detection and classification of specific cell types in histopathology slides. For example, Sirinukunwattana et al. proposed a deep learning method based on spatially constrained convolutional neural network for detecting and classifying cell nuclei in colon cancer tissue. Qupath, an open-sourced software for digital pathology and whole slide image analysis is capable of detecting nuclei but also comes with a user interface to label histopathology slides and create datasets for training new machine learning models. Within histological slides, deep learning can also be used to detect certain areas of interest and classify specific types of lesions. ”
    Artificial intelligence in oncology: From bench to clinic  
    Jamal Elkhader, Olivier Elemento
    Seminars in Cancer Biology 84 (2022) 113–128  
  • Just like pathology, deep learning is dramatically enhancing the field of radiology. To address the detection and treatment of lung cancer, Hua et al. used a deep learning framework to perform pulmonary nodule classification using CT images from the Lung Image Database Consortium dataset In the case of head and neck cancer, the identification of tumor extranodal extension is known to be difficult to diagnose radiographically, and has previously been diagnosed by postoperative pathology. Kann et al. used a CNN to identify tumor extranodal extension and nodal metastasis, trained on >2000 CT-segmented lymph node samples from patients at the Yale School of Medicine, and tested on 131 samples, achieving an AUC of 0.91 (CI: 0.85􀀀 0.97) on both extranodal extension prediction and nodal metastasis prediction. With a >85 % accuracy (extranodal extension: PPV: 0.66, NPV: 0.95; nodal metastasis: PPV: 0.88; 0.82), the use of this model as an identification tool shows favorable potential . ”
    Artificial intelligence in oncology: From bench to clinic  
    Jamal Elkhader, Olivier Elemento
    Seminars in Cancer Biology 84 (2022) 113–128 
  • “Artificial intelligence will continue to play a key role in various facets of the medical field, including drug discovery and clinical decision-making. Implementing artificial intelligence techniques into clinical practice is a promising area, allowing for progress to be made while remaining both vigorous and transparent. With improved imagine diagnostics, the efficient utilization of imaging, molecular, and cellular cancer data to predict clinical outcomes, and providing a catalyst for the development of oncologic drugs, AI has the potential for a powerful transformation. The influx of medical data is likely to continue to blossom as precision medicine continues to be implemented."
    Artificial intelligence in oncology: From bench to clinic  
    Jamal Elkhader, Olivier Elemento
    Seminars in Cancer Biology 84 (2022) 113–128
  • "Several suggestions may be helpful for consideration by clinicians and decision makers who are designing and using AI tools. First, clinicians should not assume that traditional metrics, such as the area under the receiver operating characteristic curve, translate to clinical effects because such performance metrics are usually not optimized or evaluated for specific clinical contexts. Second, clinicians should be involved in guiding the design of metrics to ensure that the algorithms produce outputs that are clinically useful and patient-centered to minimize unintended harms.”
    Rethinking Algorithm Performance Metrics for Artificial Intelligence in Diagnostic Medicine
    Reyna MA et al.
    JAMA Published online July 8, 2022 
  • "Third, clinicians should prioritize the use of AI tools with well documented and understandable explanations of performance metrics because doing so could enable informed decisions on whether and how best to use the algorithm. Fourth, clinicians should expect the prospective evaluation of algorithms in clinical settings. Evaluation in varied settings demonstrates the potential utility of an algorithm for actual clinical outcomes. Fifth, adopters of AI tools should require that Ai developers make available the full code for an algorithm, including the training data and code, so that the metrics used to develop the algorithms are explicit and modifiable.”
    Rethinking Algorithm Performance Metrics for Artificial Intelligence in Diagnostic Medicine
    Reyna MA et al.
    JAMA Published online July 8, 2022
  • "Clinicians and other health care decision makers have the responsibility to choose algorithms that are transparent, clinically useful, and effective across diverse patient populations. To facilitate an informed decision, algorithm development teams should also be diverse and work closely with clinicians to develop and implement AI performance metrics that incorporate clinical context. This process should also recognize and reflect the diversity of objectives and stakeholders in diagnostic medicine to improve the relevance and representation of AI tools in clinical practice.”  
    Rethinking Algorithm Performance Metrics for Artificial Intelligence in Diagnostic Medicine
    Reyna MA et al.
    JAMA Published online July 8, 2022
  • BACKGROUND & AIMS: Our purpose was to detect pancreatic ductal adenocarcinoma(PDAC) at the prediagnostic stage (3-36 months prior to clinical diagnosis) using radiomics based machine learning (ML) models, and to compare performance against radiologists in a case control study.
    METHODS: Volumetric pancreas segmentation was performed on prediagnostic CTs (median interval between CT and PDAC diagnosis: 398 days) of 155 patients and an age-matched cohort of 265 subjects with normal pancreas. Total 88 first order and gray level radiomic features were extracted and 34 features were selected through LASSO-based feature selection method. Dataset was randomly divided into training (292 CTs: 110 prediagnostic and 182 controls) and test subsets (128 CTs: 45 prediagnostic and 83 controls). Four ML classifiers - K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF) and XGBoost (XGB) - were evaluated. Specificity of model with highest accuracy was further validated on an independent internal dataset (n=176) and the public NIH dataset (n=80). Two radiologists (R4 and R5) independently evaluated the pancreas on a five-point diagnostic scale.
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • “Our study has limitations. The retrospective nature of the study is generally prone to selection bias. As with other radiomics studies, the precise pathologic correlates of the radiomic features that constitute the ML classifiers are not entirely known. We did not investigate the impact of differences in all the acquisition or post-processing parameters (e.g., voxel width, bin width, etc.) on the classifiers, which will be subject of the next phase of our ongoing investigation. Although we validated the high specificity of the SVM classifier on an independent internal cohort of control CTs as well as on the public NIH-PCT dataset, the sample size of these cohorts was small and the subjects in these cohorts were relatively younger. Thus, prospective larger cohorts with both cases and controls are warranted for further validation. Such prospective studies would also help determine the optimal operating point for the models to avoid a high false positive rate in context of a screening paradigm.”  
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • “In conclusion, we detected and quantified the imaging signature of early pancreatic  carcinogenesis from volumetrically segmented normal pancreas on standard-of-care CTs. The radiomics-based ML classifiers had high discrimination accuracy for classification of pancreas into prediagnostic for PDAC versus normal. The high accuracy of the SVM model was validated on CTs from external institutions. Its high specificity was generalizable on an independent internal cohort and on an external public dataset. In contrast, radiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, which shows that novel AI-based approaches can detect PDAC at a subclinical stage when it is beyond the scope of the human interrogation. Prospective validation of these ML models and their integration with complementary blood and other fluid-based biomarkers has the potential to further improve cancer prediction capabilities at the prediagnostic or symptom-free stage. Such models also have the potential to elucidate the longitudinal changes of carcinogenesis that precede the clinical diagnosis of PDAC. Finally, such models can be deployed to detect early cancer in ongoing clinical trials such as the Early Detection Initiative that seeks to evaluate outcomes of a screening strategy utilizing clinical risk-prediction models and CT in cohorts at high-risk for PDAC.”
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • Objective: Quality gaps in medical imaging datasets lead to profound errors in experiments. Our objective was to characterize such quality gaps in public pancreas imaging datasets (PPIDs), to evaluate their impact on previously published studies, and to provide post-hoc labels and segmentations as a value-add for these PPIDs.
    Conclusion: Substantial quality gaps, sources of bias, and high proportion of CTs unsuitable for AI characterize the available limited PPIDs. Published studies on these PPIDs do not account for these quality gaps. We complement these PPIDs through post-hoc labels and segmentations for public release on the TCIA portal. Collaborative efforts leading to large, well-curated PPIDs supported by adequate documentation are critically needed to translate the promise of AI to clinical practice.
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • "Public medical imaging datasets have stimulated widespread interest to explore AI to address unmet healthcare needs. In order to fully leverage these public datasets, there is a critical need to understand their strengths and limitations. Our study of public datasets in the pancreas imaging domain identified only three public datasets. The MSD dataset is the largest one with 420 CTs. Both the NIH-PCT and the TCIA PDA datasets have less than 100 CTs each. These datasets are insufficient for deep learning applications, which require very large volumes of data. There is a general hesitation to share digital assets due to concerns related to data ownership and patient privacy. Ongoing developments in federated learning architecture and privacy-preserving AI could promote wider sharing of such datasets.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • “Finally, presence of medical devices such as stents is another critical confounding factor. In the context of PDA, a tumor classification, or detection model can learn to associate the presence of a biliary stent with the diagnosis of PDA, which can lead to inadvertent overestimation of the model’s performance. Secondly, the course of such stents through the pancreatic head results in streak, artifacts and can obscure delineation of tumors in the pancreatic head. These challenges can increase the variability in tumor segmentation or result in the stent being included in segmentation mask with consequent errors in AI models. Therefore, if CTs with stents form a part of PPIDs, these should be explicitly specified in the metadata to ensure that users can make an Informed decision regarding their potential use for AI experiments.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • "In  summary, there is a need for carefully curated public imaging datasets supported by adequate documentation in the pancreas imaging domain. The available datasets for pancreatic pathologies have substantial quality gaps, sources of bias, and high proportion of CTs unsuitable for AI experiments. In our assessment, the factors  responsible for such quality gaps include general hesitation to share highly curated digital assets due to concerns related to data ownership and patient privacy, absence of tangible incentives fordata sharing, limited guidance on the dataset preparation process, inadequate involvement of domain experts in dataset curation process, and lack of awareness of the impact of insufficient documentation on the AI development pipeline.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • BACKGROUND & AIMS: Our purpose was to detect pancreatic ductal adenocarcinoma (PDAC) at the prediagnostic stage (3-36 months prior to clinical diagnosis) using radiomics based machine learning (ML) models, and to compare performance against radiologists in a case control study.  
    CONCLUSIONS: Radiomics-based ML models can detect PDAC from normal pancreas when it is beyond human interrogation capability at a substantial lead time prior to clinical diagnosis. Journal Pre-proof 6 Prospective validation and integration of such models with complementary fluid-based biomarkers has the potential for PDAC detection at a stage when surgical cure is a possibility
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • METHODS
    Volumetric pancreas segmentation was performed on prediagnostic CTs (median interval between CT and PDAC diagnosis: 398 days) of 155 patients and an age-matched cohort of 265 subjects with normal pancreas. Total 88 first order and gray level radiomic features were extracted and 34 features were selected through LASSO-based feature selection method. Dataset was randomly divided into training (292 CTs: 110 prediagnostic and 182 controls) and test subsets (128 CTs: 45 prediagnostic and 83 controls). Four ML classifiers - K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF) and XGBoost (XGB) - were evaluated. Specificity of model with highest accuracy was further validated on an independent internal dataset (n=176) and the public NIH dataset (n=80). Two radiologists (R4 and R5) independently evaluated the pancreas on a five-point diagnostic scale.
  •  RESULTS
    Median (range) time between prediagnostic CTs of the test subset and PDAC diagnosis was 386 (97-1092) days. SVM had the highest sensitivity (mean; 95% CI) (95.5; 85.5-100.0), specificity (90.3; 84.3-91.5), F1-score (89.5; 82.3-91.7), AUC (0.98; 0.94-0.98) and accuracy (92.2%; 86.7-93.7) for classification of CTs into prediagnostic versus normal. All three other ML models KNN, RF, and XGB had comparable AUCs (0.95, 0.95, and 0.96, respectively). The high specificity of SVM was generalizable to both the independent internal (92.6%) and the NIH dataset (96.2%). In contrast, inter-reader radiologist agreement was only fair (Cohen’s kappa 0.3) and their mean AUC (0.66; 0.46-0.86) was lower than each of the four ML models (AUCs: 0.95-0.98) (p < 0.001). Radiologists also recorded false positive indirect findings of PDAC in control subjects (n=83) (7% R4, 18% R5). 
  • CONCLUSIONS
    Radiomics-based ML models can detect PDAC from normal pancreas when it is beyond human interrogation capability at a substantial lead time prior to clinical diagnosis. Prospective validation and integration of such models with complementary fluid-based biomarkers has the potential for PDAC detection at a stage when surgical cure is a possibility. 
  • “These observations support the biologic insights from prior studies that the prediagnostic stage of PDAC is marked by substantial cellular activity and infiltration, which results in marked tissue heterogeneity . Our study suggests that this tissue heterogeneity is beyond the human perceptive ability but can be captured and leveraged for actionable insights through computational postprocessing techniques such as radiomics.”  
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • “The radiomics-based ML classifiers had high discrimination accuracy for classification of pancreas into prediagnostic for PDAC versus normal. The high accuracy of the SVM model was validated on CTs from external institutions. Its high specificity was generalizable on an independent internal cohort and on an external public dataset. In contrast, radiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, which shows that novel AI-based approaches can detect PDAC at a subclinical stage when it is beyond the scope of the human interrogation. Prospective validation of these ML models and their integration with complementary blood and other fluid-based biomarkers has the potential to further improve cancer prediction capabilities at the prediagnostic or symptom-free stage. Such models also have the potential to elucidate the longitudinal changes of carcinogenesis that precede the clinical diagnosis of PDAC.”
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • Background The diagnostic performance of CT for pancreatic cancer is interpreter-dependent, and approximately 40% of tumours smaller than 2 cm evade detection. Convolutional neural networks (CNNs) have shown promise in image analysis, but the networks’ potential for pancreatic cancer detection and diagnosis is unclear. We aimed to investigate whether CNN could distinguish individuals with and without pancreatic cancer on CT, compared with radiologist interpretation.
    Interpretation CNN could accurately distinguish pancreatic cancer on CT, with acceptable generalisability to images of patients from various races and ethnicities. CNN could supplement radiologist interpretation.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • “The promise of artificial intelligence (AI) to improve and reduce inequities in access, quality, and appropriateness of high-quality diagnosis remains largely unfulfilled. Vast clinical data sets, extensive computational capacity, and highly developed and accessible machine learning tools have resulted in numerous publications that describe high-performing algorithmic approachesfor a variety of diagnostic tasks. However, such approaches remain largely unadopted in clinical practice.”
    Rethinking Algorithm Performance Metrics for Artificial Intelligence in Diagnostic Medicine
    Reyna MA et al.
    JAMA Published online July 8, 2022 
  • Methods In this retrospective, diagnostic study, contrast-enhanced CT images of 370 patients with pancreatic cancer and 320 controls from a Taiwanese centre were manually labelled and randomly divided for training and validation (295 patients with pancreatic cancer and 256 controls) and testing (75 patients with pancreatic cancer and 64 controls; local test set 1). Images were preprocessed into patches, and a CNN was trained to classify patches as cancerous or non-cancerous. Individuals were classified as with or without pancreatic cancer on the basis of the proportion of patches diagnosed as cancerous by the CNN, using a cutoff determined using the training and validation set. The CNN was further tested with another local test set (101 patients with pancreatic cancers and 88 controls; local test set 2) and a US dataset (281 pancreatic cancers and 82 controls). Radiologist reports of pancreatic cancer images in the local test sets were retrieved for comparison.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • Findings Between Jan 1, 2006, and Dec 31, 2018, we obtained CT images. In local test set 1, CNN-based analysis hada sensitivity of 0·973, specificity of 1·000, and accuracy of 0·986 (area under the curve [AUC] 0·997 (95% CI 0·992–1·000). In local test set 2, CNN-based analysis had a sensitivity of 0·990, specificity of 0·989, and accuracy of 0·989 (AUC 0·999 [0·998–1·000]). In the US test set, CNN-based analysis had a sensitivity of 0·790, specificity of 0·976, and accuracy of 0·832 (AUC 0·920 [0·891–0·948)]. CNN-based analysis achieved higher sensitivity than radiologists did (0·983 vs 0·929, difference 0·054 [95% CI 0·011–0·098]; p=0·014) in the two local test sets combined. CNN missed three (1·7%) of 176 pancreatic cancers (1·1–1·2 cm). Radiologists missed 12 (7%) of 168 pancreatic cancers (1·0–3·3 cm), of which 11 (92%) were correctly classified using CNN. The sensitivity of CNN for tumours smaller than 2 cm was 92·1% in the local test sets and 63·1% in the US test set.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • Added value of this study We trained a CNN using contrast enhanced-CT images of Asian patients to distinguish pancreatic cancer from healthy pancreases. CNN achieved excellent accuracy and improved sensitivity compared with radiologist interpretation in independent Asian test sets, with acceptable performance in a North American test set obtained from patients of various races and ethnicities using diverse scanners and settings. These results provide the first solid proof of concept that CNN can capture the elusive CT features of pancreatic cancer to assist and supplement radiologists in the detection and diagnosis of pancreatic cancer.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • “Implications of all the available evidence CNN can accurately differentiate pancreatic cancer from non-cancerous pancreas, and with improvements mightaccommodate variations in patient race and ethnicity and imaging parameters that are inevitable in real-world clinical practice. CNN holds promise for developing computer-aided detection and diagnosis tools for pancreatic cancer to supplement radiologist interpretation.”
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
Gallbladder

  • “Gallstone formation plays an important role in gallbladder disease pathophysiology and affects 10% to 15% of the adult population. Although up to 80% of patients with gallstones will never be symptomatic or require treatment, complications related to gallstones result in a significant health care burden of approximately $6.2 billion annually, an increase of more than 20% over the past 3 decades. Despite the indolent course of typical gallstone disease, patients at high risk for biliary complications may require intervention.”
    Gallbladder Imaging Interpretation Pearls and Pitfalls
    Sergio P. Klimkowski et al.
    Radiol Clin N Am - (2022) https://doi.org/10.1016/j.rcl.2022.05.002
  • "Despite the indolent course of typical gallstone disease, patients at high risk for biliary complications may require intervention. These groups include patients with large gallstones greater than 3 cm or gallbladders packed with gallstones, patients with sickle cell disease,solid organ transplant recipients, and/or high-risk patients undergoing abdominal surgery for other reasons (eg, morbidly obese patients undergoing bariatric surgery).”
    Gallbladder Imaging Interpretation Pearls and Pitfalls
    Sergio P. Klimkowski et al.
    Radiol Clin N Am - (2022) https://doi.org/10.1016/j.rcl.2022.05.002
  • "Gallbladder disease is a common clinical problem in both emergent and nonemergent settings and includes both malignant and benign etiologies. Imaging evaluation of the right upper quadrant plays a key role in establishing the correct diagnosis. Clinical history, physical examination, and laboratory values are important in the accurate imaging assessment of the right upper quadrant. Significant overlap in imaging features of both benign and malignant etiologies can pose diagnostic dilemmas. Additionally, inherent limitations of ultrasound and cross-sectional techniques may lead to diagnostic pitfalls preventing accurate diagnosis. Knowledge of these pitfalls and their solutions help in the accurate assessment of the gallbladder.”
    Gallbladder Imaging Interpretation Pearls and Pitfalls
    Sergio P. Klimkowski et al.
    Radiol Clin N Am - (2022) https://doi.org/10.1016/j.rcl.2022.05.002 
Kidney

  • “Spontaneous renal hemorrhage (SRH) or hematoma is an intraparenchymal renal hemorrhage of unknown origin in a patient without trauma or anticoagulation . SRH is most commonly related to occult vascular renal tumors (angiomyolipoma or renal cell carcinoma), vasculitides (polyarteritis nodosa), or vascular malformations. A few cases are idiopathic or have been attributed to infection, uncontrolled hypertension, ruptured hemorrhagic cysts, or erosion from large renal stones.”
    Spontaneous Renal Hemorrhage: A Case Report and Clinical Protocol
    Olivia Antonescu et al.
    Cureus 13(6): e15547. doi:10.7759/cureus.15547 
  • "Patients with SRH will classically present with Lenk’s triad of flank pain, tenderness, and “symptoms of blood loss,” including generalized fatigue, tachycardia, dizziness, and hypotension depending on the volume of blood loss. However, it can mimic many acute abdominal pathologies. As such, SRH is usually discovered incidentally on ultrasound or contrast-enhanced abdominal CT. Abdominal CTA however is the preferred imaging modality for definitive diagnosis and preprocedural planning. Laboratory findings often include anemia and hematuria.”
    Spontaneous Renal Hemorrhage: A Case Report and Clinical Protocol
    Olivia Antonescu et al.
    Cureus 13(6): e15547. doi:10.7759/cureus.15547
  • “Michaelis and Gutmann first described malakoplakia in 1902. The term malakoplakia is derived from the Greek words malakos, which means soft, and plakos, which means plague. The age at diagnosis ranges from 6 to 85 years, with an average age of 50 years at presentation. There is a female predominance, with a female to male ratio of 4:1. Malakoplakia can affect any organ system including the gastrointestinal system, bones, lungs, lymph nodes and skin, but the collecting system of the urinary tract is most frequently involved. The renal lesion is most often multifocal .”
    Renal malakoplakia presenting as a renal mass in a 55-year-old man: a case report  
    Maryam Abolhasani et al.
    Journal of Medical Case Reports 2012, 6:379
  • “Malakoplakia is an uncommon chronic inflammatory condition that has a gross and microscopic appearance resembling that of xanthogranulomatous pyelonephritis. It is characterized by distinctive Michaelis-Gutmann bodies. Malakoplakia can affect any organ system but genitourinary tract involvement is the most common, particularly iimmunocompromised individuals. Very rare cases have been reported to present as a unifocal lesion mimicking a renal tumor.”
    Renal malakoplakia presenting as a renal mass in a 55-year-old man: a case report  
    Maryam Abolhasani et al.
    Journal of Medical Case Reports 2012, 6:379
  • “Renal malakoplakia must be kept in mind for patients presenting with a renal mass and a history of long-term recurrent renal infections or renal failure. The large, rapidly growing nodules of malakoplakia may mimic renal cell carcinoma in imaging studies. In these cases, a true cut needle biopsy may help the correct diagnosis and prevent unnecessary surgery.”
    Renal malakoplakia presenting as a renal mass in a 55-year-old man: a case report  
    Maryam Abolhasani et al.
    Journal of Medical Case Reports 2012, 6:379
  • “In imaging studies, the appearance of the affected kidney ranges from that of a normal kidney to an enlarged, nonfunctioning kidney. Commonly, multiple poorly defined renal lesions enlarging the kidney, and often involving both kidneys, are present. The renal lesions can distort the pelvis and calices but seldom cause obstruction. Perirenal extension and renal vein thrombosis have been reported. Focal renal lesions are usually poorly defined and hypoechoic on ultrasound study. Parenchymal calcification is rare. A unifocal renal lesion is uncommon and can resemble a necrotic renal cell carcinoma. Differential diagnoses in radiologic studies include local abscess, granuloma, xanthogranulomatous pyelonephritis, lymphoma.”
    Renal malakoplakia presenting as a renal mass in a 55-year-old man: a case report  
    Maryam Abolhasani et al.
    Journal of Medical Case Reports 2012, 6:379
  • “The diagnosis of malakoplakia must be kept in mind for patients presenting with a renal mass and a history of long-term recurrent renal infections or renal failure [4]. Renal malakoplakia may mimic renal tumors and lead to unnecessary surgery. The patient in our report had renal malakoplakia but underwent a nephrectomy with the clinical diagnosis of a renal tumor. A nephrectomy can be a choice for unifocal malakoplakia, but the preoperative diagnosis of renal malakoplakia in appropriate clinical settings can prevent unnecessary surgery.”
    Renal malakoplakia presenting as a renal mass in a 55-year-old man: a case report  
    Maryam Abolhasani et al.
    Journal of Medical Case Reports 2012, 6:379
  • “The diagnosis of malakoplakia must be kept in mind for patients presenting with a renal mass and a history of long-term recurrent renal infections or renal failure. Renal malakoplakia may mimic renal tumors and lead to unnecessary surgery. The patient in our report had renal malakoplakia but underwent a nephrectomy with the clinical diagnosis of a renal tumor. A nephrectomy can be a choice for unifocal malakoplakia, but the preoperative diagnosis of renal malakoplakia in appropriate clinical settings can prevent unnecessary surgery.”
    Malacoplakia Presenting as a Solitary Renal Mass
    Aaron J. Wielenberg et al.
    AJR. 2004;183: 1703-1705
  • “The appearance of the affected kidney on im-aging studies ranges from that of a normal kidney   to   an   enlarged,   nonfunctioning   kidney. Commonly,  multiple  poorly  defined  renal  lesions  enlarging  the  kidney  and  often  involving both kidneys are present. The renal lesions candistort  the  pelvis  and  calices  but  seldom  cause obstruction.  Perinephric  extension  and  renal vein thrombosis have been reported. Focal renal  lesions  on  sonography  usually  are  poorly defined and hypoechoic. CT reflects the pathologic  range  of  renal  involvement  of  multifocal disease involving one or both kidneys. Most of-ten,  renal  enlargement  exists  with  poorly  de-fined,  variably  sized  solid  masses.  Perinephric extension  and  renal  vein  thrombosis  are  well. shown. Parenchymal calcification is rare [2, 6].A  unifocal  renal  lesion  is  uncommon  and  can resemble  a  necrotic  renal  cell  carcinoma,  as  in the  case  presented  here.  Neovascularity  is  generally  absent  but  has  been  reported  with  unifo-cal disease.”
    Malacoplakia Presenting as a Solitary Renal Mass
    Aaron J. Wielenberg et al.
    AJR. 2004;183: 1703-1705
Musculoskeletal

  • “Osteopetrosis, translated as “stone bone,” is a rare inherited bone disease. It is also known as marble bone disease where bones harden and become abnormally dense, opposite to osteoporosis where bones become less dense and more brittle, or osteomalacia where bones soften. Various modalities of imaging have been shown to be useful in detecting and diagnosing osteopetrosis.”
    Osteopetrosis: radiological & radionuclide imaging.  
    Sit C, Agrawal K, Fogelman I, Gnanasegaran G.  
    Indian J Nucl Med. 2015 Jan-Mar;30(1):55-8. 
  • “Autosomal dominant osteopetrosis (ADO) is also known as Albers-Schönberg disease after first being described in 1904. Typically in the mildest form of the disorder, affected individuals may show no symptoms. The incidence is 1:20000, and a mutation in the CLCN7 gene is responsible for 75% of ADO. Due to its benign symptoms, ADO is usually discovered by accident when X-ray is done for another reason. Clinical manifestations become apparent in late childhood or adolescence; symptoms include: Multiple bone fractures, scoliosis, arthritis, and osteomyelitis.”
    Osteopetrosis: radiological & radionuclide imaging.  
    Sit C, Agrawal K, Fogelman I, Gnanasegaran G.  
    Indian J Nucl Med. 2015 Jan-Mar;30(1):55-8. 
  • "Computed tomography scan often shows increased area of bone density “bone-in-bone”, appearance in the vertebrae and phalanges, and sometimes focal sclerosis of skull base, pelvis and vertebral end plates, giving rise to “sandwich” vertebrae, and “rugger-jersey” spine .”
    Osteopetrosis: radiological & radionuclide imaging.  
    Sit C, Agrawal K, Fogelman I, Gnanasegaran G.  
    Indian J Nucl Med. 2015 Jan-Mar;30(1):55-8. 
  • “Osteopetrosis is a group of rare bone disorders, within the family of sclerosing bone dysplasias, characterized by reduced osteoclastic bone resorption that results in a high bone mass. Rather than conferring strength, the overly dense bone architecture belies a structural brittleness that predisposes to fracture. The disruption of normal bone modeling and remodeling can give rise to skeletal deformity and dental abnormalities and can interfere with mineral homeostasis. In addition, expansion of bone into marrow cavities and cranial nerve foramina can compromise hematologic and neurologic function, respectively; the former may manifest as profound anemia, bleeding, frequent infections, and hepatosplenomegaly from extramedullary hematopoiesis. The latter can lead to blindness, deafness, and nerve palsies.”
    Diagnosis and Management of Osteopetrosis: Consensus Guidelines From the Osteopetrosis Working Group,
    Calvin C Wu, et al.
    The Journal of Clinical Endocrinology & Metabolism, Volume 102, Issue 9, 1 September 2017, Pages 3111–3123
  • “Osteopetrosis, derived from the Greek words for "bone" ("osteo") and "stone" ("petros"), is a fitting name for a disease in which generalized osteosclerosis identifiable on standard radiographs is pathognomonic. Parallel bands of dense bone can give the appearance of “bone-within-bone” or “endobones.” This finding is often prominent in the pelvis, long bones, phalanges, and vertebrae. Vertebrae can also be uniformly dense or take on a “sandwich vertebrae” or “rugger-jersey” appearance (when a normal-appearing vertebral midbody is sandwiched between dense bands along the superior and inferior endplates).”
    Diagnosis and Management of Osteopetrosis: Consensus Guidelines From the Osteopetrosis Working Group,
    Calvin C Wu, et al.
    The Journal of Clinical Endocrinology & Metabolism, Volume 102, Issue 9, 1 September 2017, Pages 3111–3123
  • BACKGROUND. Bone scintigraphy (BS) using 99mTc-labeled methylene diphosphonate(99mTc-MDP) remains the recommended imaging modality for the detection of bone metastases in patients with prostate cancer (PCa). However, PET/CT using prostate-specific membrane antigen (PSMA) ligands is increasingly recognized as a means of evaluating disease extent in patients with PCa, including use as a possible stand-alone test in high-risk patients.
    OBJECTIVE. The purpose of this study is to compare the diagnostic performance of68Ga-PSMA-11 PET/CT with that of 99mTc-MDP BS for the detection of bone metastases inpatients with PCa.
    Head-To-Head Comparison of 68Ga-PSMA-11 PET/CT and 99mTc-MDP Bone Scintigraphy for the Detection of Bone Metastases in Patients With Prostate Cancer: A Meta-Analysis
    Gege Zhao, Bin Ji
    AJR 2022; 219:1–11
  • EVIDENCE SYNTHESIS. Six studies with 546 patients were included. Pooled sensitivity and specificity, respectively, were 98% (95% CI, 94–99%) and 97% (95% CI, 91–99%)for 68Ga-PSMA-11 PET/CT versus 83% (95% CI, 69–91%) and 68% (95% CI, 41–87%) for99mTc-MDP BS. The AUC was 0.99 (95% CI, 0.96–1.00) for 68Ga-PSMA-11 PET/CT and 0.85(95% CI, 0.81–0.87) for 99mTc-MDP BS. Among 408 patients from five included studies,68Ga-PSMA-11 PET/CT correctly identified bone metastases in 43 of 193 patients (22.3%)with negative 99mTc-MDP BS results, whereas 99mTc-MDP BS correctly identified bone metastases in four of 210 patients (1.9%) with negative 68Ga-PSMA-11 PET/CT results.
    CONCLUSION. On a per-patient basis, the diagnostic performance of 68Ga-PSMA-11PET/CT is superior to that of 99mTc-MDP BS for the detection of PCa bone metastases. Furthermore,99mTc-MDP BS offers limited additional information in patients with negative68Ga-PSMA-11 PET/CT results.
    CLINICAL IMPACT. According to current evidence, 99mTc-MDP BS is highly unlikely to be additive to 68Ga-PSMA-11 PET/CT in identifying bone metastases in patients with PCa.
    Head-To-Head Comparison of 68Ga-PSMA-11 PET/CT and 99mTc-MDP Bone Scintigraphy for the Detection of Bone Metastases in Patients With Prostate Cancer: A Meta-Analysis
    Gege Zhao, Bin Ji
    AJR 2022; 219:1–11
  • CONCLUSION. On a per-patient basis, the diagnostic performance of 68Ga-PSMA-11PET/CT is superior to that of 99mTc-MDP BS for the detection of PCa bone metastases. Furthermore,99mTc-MDP BS offers limited additional information in patients with negative68Ga-PSMA-11 PET/CT results.
    CLINICAL IMPACT. According to current evidence, 99mTc-MDP BS is highly unlikely to be additive to 68Ga-PSMA-11 PET/CT in identifying bone metastases in patients with PCa.
    Head-To-Head Comparison of 68Ga-PSMA-11 PET/CT and 99mTc-MDP Bone Scintigraphy for the Detection of Bone Metastases in Patients With Prostate Cancer: A Meta-Analysis
    Gege Zhao, Bin Ji
    AJR 2022; 219:1–11
OB GYN

  • “Uterine leiomyomas, sometimes incorrectly colloquially referred to as uterine fibroids, are the most frequently encountered benign myomatous tumors of the uterus, being observed in up to 20—40% of reproductive-age women and 70—80% of perimenopausal women. In addition, these benign tumors may become symptomatic in 20—50% of patients and subsequently produce pelvic pain, subfertility or abnormal uterine bleeding, requiring gynecologic hospitalization in about 30% of affected women.”
    How to differentiate uterine leiomyosarcoma from leiomyoma with imaging
    Sun S et al.
    Diagnostic and Interventional Imaging (2019) 100, 619—634
  • “On the malignant spectrum, uterine sarcomas tend to occur in an older patient population when compared to leiomyomas, and only account for 3—7% of all uterine malignancies. They often present with the same symptoms as leiomyomas and thus cannot reliably be distinguished clinically .Leiomyosarcomas (LMSs) are the most common uterine sarcomas, with an estimated annual incidence of 0.5—7/100,000 per women, followed by endometrial stromal sarcomas with an annual incidence of 1—2/million per woman.”
    How to differentiate uterine leiomyosarcoma from leiomyoma with imaging
    Sun S et al.
    Diagnostic and Interventional Imaging (2019) 100, 619—634
  • “CT plays a limited role in the initial diagnosis and local staging of myometrial lesions. CT is excellent for demonstrating calcifications; they are often found in leiomyomas but may also be present in LMSs. CT can also be useful in the initial evaluation of patients presenting with acute abdominal pain, especially those with torsed subserosal leiomyomas, which can then undergo hemorrhagic necrosis and confound the diagnosis. In women with LMS, CT is primarily used for staging purposes and to exclude distant recurrence post therapy (LMS tends to metastasize to the lungs and liver). CT is also optimal for visualizing the postoperative pelvic anatomy, allowing for proper evaluation of surgical complications including bowel obstruction or injury, ureteral or bladder injuries and urinary fistulas .”
    How to differentiate uterine leiomyosarcoma from leiomyoma with imaging
    Sun S et al.
    Diagnostic and Interventional Imaging (2019) 100, 619—634
  • "While CT alone is not particularly helpful for the differentiation of leiomyomas from LMSs, it has shown utility in combination with 18F-Fludeoxyglucose (18F-FDG)-PET in the context of indeterminate myometrial lesions on MRI. It is thought that malignant tumors experience upregulation of glucose transporter genes (GLUTs) either due to increase of normal cellular enzymes or synthesis of new transporters after transformation due to oncogenes. Based on this hypothesis, greater 18F-FDG uptake of LMS would be expected in comparison to leiomyomas due to their increased metabolic rate and glycolysis. There are few studies on this specific topic, however a study by Umesaki et al. has shown 100% sensitivity of 18F-FDG PET in the diagnosis of uterine sarcoma vs. leiomyoma with a positive standardized uptake value (SUV) cut-off value of 2.5.”
    How to differentiate uterine leiomyosarcoma from leiomyoma with imaging
    Sun S et al.
    Diagnostic and Interventional Imaging (2019) 100, 619—634
Pancreas

  • BACKGROUND & AIMS: Our purpose was to detect pancreatic ductal adenocarcinoma(PDAC) at the prediagnostic stage (3-36 months prior to clinical diagnosis) using radiomics based machine learning (ML) models, and to compare performance against radiologists in a case control study.
    METHODS: Volumetric pancreas segmentation was performed on prediagnostic CTs (median interval between CT and PDAC diagnosis: 398 days) of 155 patients and an age-matched cohort of 265 subjects with normal pancreas. Total 88 first order and gray level radiomic features were extracted and 34 features were selected through LASSO-based feature selection method. Dataset was randomly divided into training (292 CTs: 110 prediagnostic and 182 controls) and test subsets (128 CTs: 45 prediagnostic and 83 controls). Four ML classifiers - K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF) and XGBoost (XGB) - were evaluated. Specificity of model with highest accuracy was further validated on an independent internal dataset (n=176) and the public NIH dataset (n=80). Two radiologists (R4 and R5) independently evaluated the pancreas on a five-point diagnostic scale.
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • “Our study has limitations. The retrospective nature of the study is generally prone to selection bias. As with other radiomics studies, the precise pathologic correlates of the radiomic features that constitute the ML classifiers are not entirely known. We did not investigate the impact of differences in all the acquisition or post-processing parameters (e.g., voxel width, bin width, etc.) on the classifiers, which will be subject of the next phase of our ongoing investigation. Although we validated the high specificity of the SVM classifier on an independent internal cohort of control CTs as well as on the public NIH-PCT dataset, the sample size of these cohorts was small and the subjects in these cohorts were relatively younger. Thus, prospective larger cohorts with both cases and controls are warranted for further validation. Such prospective studies would also help determine the optimal operating point for the models to avoid a high false positive rate in context of a screening paradigm.”  
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • “In conclusion, we detected and quantified the imaging signature of early pancreatic  carcinogenesis from volumetrically segmented normal pancreas on standard-of-care CTs. The radiomics-based ML classifiers had high discrimination accuracy for classification of pancreas into prediagnostic for PDAC versus normal. The high accuracy of the SVM model was validated on CTs from external institutions. Its high specificity was generalizable on an independent internal cohort and on an external public dataset. In contrast, radiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, which shows that novel AI-based approaches can detect PDAC at a subclinical stage when it is beyond the scope of the human interrogation. Prospective validation of these ML models and their integration with complementary blood and other fluid-based biomarkers has the potential to further improve cancer prediction capabilities at the prediagnostic or symptom-free stage. Such models also have the potential to elucidate the longitudinal changes of carcinogenesis that precede the clinical diagnosis of PDAC. Finally, such models can be deployed to detect early cancer in ongoing clinical trials such as the Early Detection Initiative that seeks to evaluate outcomes of a screening strategy utilizing clinical risk-prediction models and CT in cohorts at high-risk for PDAC.”
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis
    Sovanlal Mukherjee,et al.
    Gastroenterology (2022),doi:https://doi.org/10.1053/j.gastro.2022.06.066.
  • Objective: Quality gaps in medical imaging datasets lead to profound errors in experiments. Our objective was to characterize such quality gaps in public pancreas imaging datasets (PPIDs), to evaluate their impact on previously published studies, and to provide post-hoc labels and segmentations as a value-add for these PPIDs.
    Conclusion: Substantial quality gaps, sources of bias, and high proportion of CTs unsuitable for AI characterize the available limited PPIDs. Published studies on these PPIDs do not account for these quality gaps. We complement these PPIDs through post-hoc labels and segmentations for public release on the TCIA portal. Collaborative efforts leading to large, well-curated PPIDs supported by adequate documentation are critically needed to translate the promise of AI to clinical practice.
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • "Public medical imaging datasets have stimulated widespread interest to explore AI to address unmet healthcare needs. In order to fully leverage these public datasets, there is a critical need to understand their strengths and limitations. Our study of public datasets in the pancreas imaging domain identified only three public datasets. The MSD dataset is the largest one with 420 CTs. Both the NIH-PCT and the TCIA PDA datasets have less than 100 CTs each. These datasets are insufficient for deep learning applications, which require very large volumes of data. There is a general hesitation to share digital assets due to concerns related to data ownership and patient privacy. Ongoing developments in federated learning architecture and privacy-preserving AI could promote wider sharing of such datasets.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • “Finally, presence of medical devices such as stents is another critical confounding factor. In the context of PDA, a tumor classification, or detection model can learn to associate the presence of a biliary stent with the diagnosis of PDA, which can lead to inadvertent overestimation of the model’s performance. Secondly, the course of such stents through the pancreatic head results in streak, artifacts and can obscure delineation of tumors in the pancreatic head. These challenges can increase the variability in tumor segmentation or result in the stent being included in segmentation mask with consequent errors in AI models. Therefore, if CTs with stents form a part of PPIDs, these should be explicitly specified in the metadata to ensure that users can make an Informed decision regarding their potential use for AI experiments.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • "In  summary, there is a need for carefully curated public imaging datasets supported by adequate documentation in the pancreas imaging domain. The available datasets for pancreatic pathologies have substantial quality gaps, sources of bias, and high proportion of CTs unsuitable for AI experiments. In our assessment, the factors  responsible for such quality gaps include general hesitation to share highly curated digital assets due to concerns related to data ownership and patient privacy, absence of tangible incentives fordata sharing, limited guidance on the dataset preparation process, inadequate involvement of domain experts in dataset curation process, and lack of awareness of the impact of insufficient documentation on the AI development pipeline.”
    Quality gaps in public pancreas imaging datasets: Implications & challenges for AI applications
    Garima Suman et al.
    Pancreatology 21 (2021) 1001-1008
  • BACKGROUND & AIMS: Our purpose was to detect pancreatic ductal adenocarcinoma (PDAC) at the prediagnostic stage (3-36 months prior to clinical diagnosis) using radiomics based machine learning (ML) models, and to compare performance against radiologists in a case control study.  
    CONCLUSIONS: Radiomics-based ML models can detect PDAC from normal pancreas when it is beyond human interrogation capability at a substantial lead time prior to clinical diagnosis. Journal Pre-proof 6 Prospective validation and integration of such models with complementary fluid-based biomarkers has the potential for PDAC detection at a stage when surgical cure is a possibility
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • METHODS
    Volumetric pancreas segmentation was performed on prediagnostic CTs (median interval between CT and PDAC diagnosis: 398 days) of 155 patients and an age-matched cohort of 265 subjects with normal pancreas. Total 88 first order and gray level radiomic features were extracted and 34 features were selected through LASSO-based feature selection method. Dataset was randomly divided into training (292 CTs: 110 prediagnostic and 182 controls) and test subsets (128 CTs: 45 prediagnostic and 83 controls). Four ML classifiers - K-Nearest Neighbor (KNN), Support Vector Machine (SVM), Random Forest (RF) and XGBoost (XGB) - were evaluated. Specificity of model with highest accuracy was further validated on an independent internal dataset (n=176) and the public NIH dataset (n=80). Two radiologists (R4 and R5) independently evaluated the pancreas on a five-point diagnostic scale.
  •  RESULTS
    Median (range) time between prediagnostic CTs of the test subset and PDAC diagnosis was 386 (97-1092) days. SVM had the highest sensitivity (mean; 95% CI) (95.5; 85.5-100.0), specificity (90.3; 84.3-91.5), F1-score (89.5; 82.3-91.7), AUC (0.98; 0.94-0.98) and accuracy (92.2%; 86.7-93.7) for classification of CTs into prediagnostic versus normal. All three other ML models KNN, RF, and XGB had comparable AUCs (0.95, 0.95, and 0.96, respectively). The high specificity of SVM was generalizable to both the independent internal (92.6%) and the NIH dataset (96.2%). In contrast, inter-reader radiologist agreement was only fair (Cohen’s kappa 0.3) and their mean AUC (0.66; 0.46-0.86) was lower than each of the four ML models (AUCs: 0.95-0.98) (p < 0.001). Radiologists also recorded false positive indirect findings of PDAC in control subjects (n=83) (7% R4, 18% R5). 
  • CONCLUSIONS
    Radiomics-based ML models can detect PDAC from normal pancreas when it is beyond human interrogation capability at a substantial lead time prior to clinical diagnosis. Prospective validation and integration of such models with complementary fluid-based biomarkers has the potential for PDAC detection at a stage when surgical cure is a possibility. 
  • “These observations support the biologic insights from prior studies that the prediagnostic stage of PDAC is marked by substantial cellular activity and infiltration, which results in marked tissue heterogeneity . Our study suggests that this tissue heterogeneity is beyond the human perceptive ability but can be captured and leveraged for actionable insights through computational postprocessing techniques such as radiomics.”  
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • “The radiomics-based ML classifiers had high discrimination accuracy for classification of pancreas into prediagnostic for PDAC versus normal. The high accuracy of the SVM model was validated on CTs from external institutions. Its high specificity was generalizable on an independent internal cohort and on an external public dataset. In contrast, radiologist readers had low interreader agreement, sensitivity, and discrimination accuracy, which shows that novel AI-based approaches can detect PDAC at a subclinical stage when it is beyond the scope of the human interrogation. Prospective validation of these ML models and their integration with complementary blood and other fluid-based biomarkers has the potential to further improve cancer prediction capabilities at the prediagnostic or symptom-free stage. Such models also have the potential to elucidate the longitudinal changes of carcinogenesis that precede the clinical diagnosis of PDAC.”
    Radiomics-Based Machine Learning Models Can Detect Pancreatic Cancer on Prediagnostic CTs at a Substantial Lead Time Prior to Clinical Diagnosis,
    Sovanlal Mukherjee et al.
    Gastroenterology 2022 (in press)
  • Background The diagnostic performance of CT for pancreatic cancer is interpreter-dependent, and approximately 40% of tumours smaller than 2 cm evade detection. Convolutional neural networks (CNNs) have shown promise in image analysis, but the networks’ potential for pancreatic cancer detection and diagnosis is unclear. We aimed to investigate whether CNN could distinguish individuals with and without pancreatic cancer on CT, compared with radiologist interpretation.
    Interpretation CNN could accurately distinguish pancreatic cancer on CT, with acceptable generalisability to images of patients from various races and ethnicities. CNN could supplement radiologist interpretation.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • Objectives: To investigate the risk factors for early recurrence after curative pancreatoduodenectomy for resectablepancreatic ductal adenocarcinoma.
    Conclusions: Tumor size > 4 cm from the preoperative imaging study was a poor prognostic factor for early recurrence after curative pancreatoduodenectomy for resectable pancreatic adenocarcinoma indicated that they may have radiological occult metastasis, thus, staging laparoscopy may reduce the number of unnecessary laparotomies and avoid missing radiologically negative metastases.
    Analysis of preoperative risk factors for early recurrence after curative pancreatoduodenectomy for resectable pancreatic adenocarcinoma
    Pipit Burasakarn et al.
    Innov Surg Sci 2022; (in press)
  • “The independent preoperative risk factor associated with adverse disease-free survival was tumor size > 4 cm (hazard ratio [HR], 14.34, p=0.022). The perioperative risk factors associated with adverse disease-free survival were pathological lymphovascular invasion (HR,4.31; p=0.048) and non-hepatopancreatobiliary surgeon (HR, 5.9; p=0.022). Risk factors associated with poor overall survival were microscopical margin positive (R1) resection(HR, 3.68; p=0.019) and non-hepatopancreatobiliary surgeon(HR, 3.45; p=0.031).”
    Analysis of preoperative risk factors for early recurrence after curative pancreatoduodenectomy for resectable pancreatic adenocarcinoma
    Pipit Burasakarn et al.
    Innov Surg Sci 2022; (in press)
  • "Tumor size >4 cm from the preoperative imaging study was a poor prognostic factor for early recurrence after curative pancreatoduodenectomy for resectable pancreatic adenocarcinoma indicated that they may have radiological occult metastasis, thus, staging laparoscopy may reduce thenumber of unnecessary laparotomies and avoid missing radiologically negative metastases.”
    Analysis of preoperative risk factors for early recurrence after curative pancreatoduodenectomy for resectable pancreatic adenocarcinoma
    Pipit Burasakarn et al.
    Innov Surg Sci 2022; (in press)
  • Methods In this retrospective, diagnostic study, contrast-enhanced CT images of 370 patients with pancreatic cancer and 320 controls from a Taiwanese centre were manually labelled and randomly divided for training and validation (295 patients with pancreatic cancer and 256 controls) and testing (75 patients with pancreatic cancer and 64 controls; local test set 1). Images were preprocessed into patches, and a CNN was trained to classify patches as cancerous or non-cancerous. Individuals were classified as with or without pancreatic cancer on the basis of the proportion of patches diagnosed as cancerous by the CNN, using a cutoff determined using the training and validation set. The CNN was further tested with another local test set (101 patients with pancreatic cancers and 88 controls; local test set 2) and a US dataset (281 pancreatic cancers and 82 controls). Radiologist reports of pancreatic cancer images in the local test sets were retrieved for comparison.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • Findings Between Jan 1, 2006, and Dec 31, 2018, we obtained CT images. In local test set 1, CNN-based analysis hada sensitivity of 0·973, specificity of 1·000, and accuracy of 0·986 (area under the curve [AUC] 0·997 (95% CI 0·992–1·000). In local test set 2, CNN-based analysis had a sensitivity of 0·990, specificity of 0·989, and accuracy of 0·989 (AUC 0·999 [0·998–1·000]). In the US test set, CNN-based analysis had a sensitivity of 0·790, specificity of 0·976, and accuracy of 0·832 (AUC 0·920 [0·891–0·948)]. CNN-based analysis achieved higher sensitivity than radiologists did (0·983 vs 0·929, difference 0·054 [95% CI 0·011–0·098]; p=0·014) in the two local test sets combined. CNN missed three (1·7%) of 176 pancreatic cancers (1·1–1·2 cm). Radiologists missed 12 (7%) of 168 pancreatic cancers (1·0–3·3 cm), of which 11 (92%) were correctly classified using CNN. The sensitivity of CNN for tumours smaller than 2 cm was 92·1% in the local test sets and 63·1% in the US test set.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • Added value of this study We trained a CNN using contrast enhanced-CT images of Asian patients to distinguish pancreatic cancer from healthy pancreases. CNN achieved excellent accuracy and improved sensitivity compared with radiologist interpretation in independent Asian test sets, with acceptable performance in a North American test set obtained from patients of various races and ethnicities using diverse scanners and settings. These results provide the first solid proof of concept that CNN can capture the elusive CT features of pancreatic cancer to assist and supplement radiologists in the detection and diagnosis of pancreatic cancer.
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
  • “Implications of all the available evidence CNN can accurately differentiate pancreatic cancer from non-cancerous pancreas, and with improvements mightaccommodate variations in patient race and ethnicity and imaging parameters that are inevitable in real-world clinical practice. CNN holds promise for developing computer-aided detection and diagnosis tools for pancreatic cancer to supplement radiologist interpretation.”
    Deep learning to distinguish pancreatic cancer tissue from non-cancerous pancreatic tissue: a retrospective study with cross-racial external validation
    Kao-Lang Liu et al.
    Lancet Digital Health 2020; 2: e303–13
Small Bowel

  • Different small bowel (SB) imaging patterns have been associated with metastatic melanoma (MM). More frequently, it presents with polypoid nodules causing intussusception, and less frequently as ulcerating mural nodules, exo-enteric lesions, infiltrating masses or serosal deposits. Aneurysmal dilatation is a rare presentationwith only a single case previously reported in the literature to the best of our knowledge. This appearance is defined as a cavitary dilatation of the intestinal lumen (> 4 cm) with a nodular, irregular luminal contour and peripheral bowel wall thickening and was first described as an imaging finding characteristic of lymphoma. However, it has subsequently been associated with other cancer types, including primary SB adenocarcinoma,leiomyosarcoma, gastrointestinal stromal tumours, and metastatic disease from non-small cell bronchogenic carcinoma, adenocarcinoma of the rectum and endometrial stromal sarcoma , as well as with amyloidosis .
    The forgotten appearance of metastatic melanoma in the small bowel
    Eva Mendes Serrao et al.
    Cancer Imaging (2022) 22:27 https://doi.org/10.1186/s40644-022-00463-5
  • “Gastrointestinal spread from MM is relatively common, with the SB representing the most common site of involvement. However, SB involvement is still vastly underappreciated clinically, with studies reporting the presence of lesions in 43.5–60% of cases postmortem, but only 1.5–4.4% antemortem. SB melanoma metastases are frequently multiple, due to haematogenous dissemination, and preferentially affect the terminal ileum and jejunum As in our reported cases, SB metastases are more common with cutaneous as opposed to non-cutaneous melanomas. Primary mucosal melanomas arising in the SB are rare, remaining a controversial diagnosis as the possibility of a MM from an unidentified or regressed primary cutaneous melanoma should always be considered.”
    The forgotten appearance of metastatic melanoma in the small bowel
    Eva Mendes Serrao et al.
    Cancer Imaging (2022) 22:27 https://doi.org/10.1186/s40644-022-00463-5
  • "Interpretation of staging CT studies of patients with MM can be challenging given the need for a thorough review of the scans and unpredictable metastatic patterns with this disease. Nevertheless, the bowel is an important review area in these patients, particularly in the case of primary cutaneous melanoma arising in the head and neck region, trunk and lower extremity. On the other hand, when a SB lesion is incidentally identified on imaging, one should always consider the possibility of MM, and careful medical history should be taken regarding prior history of primary skin lesions.”
    The forgotten appearance of metastatic melanoma in the small bowel
    Eva Mendes Serrao et al.
    Cancer Imaging (2022) 22:27 https://doi.org/10.1186/s40644-022-00463-5 
Stomach

  • “High-resolution CT is the gold standard diagnostic imaging study for staging of gastric malignancies. Cinematic rendering creates a photorealistic evaluation using the standard high-resolution CT volumetric data set. This novel display method ofers unique possibilities for the evaluation of gastric masses. Here we present further observations of the role of cinematic rendering in the evaluation of gastric masses at a large tertiary care center. We ofer three valuable teaching points for the application of the cinematic rendering for gastric masses with several case examples for each teaching point, discuss potential limitations of cinematic rendering, and review future directions for cinematic rendering in this setting.”
    Implementation of cinematic rendering of gastric masses into clinical practice: a pictorial review  
    Claire Brookmeyer · Steven P. Rowe  · Linda C. Chu  · Elliot K. Fishman
    Abdom Radiol (NY) 2022 Jul 11.doi: 10.1007/s00261-022-03604-3.
Trauma

  • “Traumatic arterial injuries of the extremities are a rare but potentially fatal event. Computed tomography (CT) angiography of the extremities has become the technique of choice and can provide rapid accurate detection and characterization of vascular lesions. Vascular injuries can be classified in active hemorrhage, vasospasm, occlusion, post-traumatic arteriovenous fistula, pseudoaneurysm, and patterns of intimal injuries. The learning objectives of this pictorial essay are to review the normal arterial anatomy of the upper and lower limbs, describe the technique of CT angiography in vascular trauma of the extremities, describe and illustrate the CT-angiography findings of traumatic arterial injuries, and know the potential pitfalls when interpreting a CT-angiography of the extremities.”
    Traumatic arterial injuries in upper and lower limbs: what every radiologist should know
    Zhao Hui Chen Zhou et al.
    Emergency Radiology (2022) 29:781–790
Vascular

  • “The true mortality associated with undiagnosed pulmonary embolism is estimated to be less than 5%, but recovery from pulmonary embolism is associated with complications such as bleeding due to anticoagulant treatment, recurrent venous thromboembolism, chronic thromboembolic pulmonary hypertension, and long-term psychological distress. Approximately half the patients who receive a diagnosis of pulmonary embolism have functional and exercise limitations 1 year later (known as post–pulmonary-embolism syndrome), and the health-related quality of life for patients with a history of pulmonary embolism is diminished as compared with that of matched controls. Therefore, the timely diagnosis and expert management of pulmonary embolism are important.”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.

  • Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • • Pulmonary embolism is a common diagnosis and can be associated with recurrent venous thromboembolism, bleeding due to anticoagulant therapy, chronic thromboembolic pulmonary hypertension, and long-term psychological distress.
    • A minority of patients who are evaluated for possible pulmonary embolism benefit from chest imaging (e.g., computed tomography).
    • Initial treatment is guided by classification of the pulmonary embolism as high-risk, intermediate-risk, or low-risk. Most patients have low-risk pulmonary embolism, and their care can be managed at home with a direct oral anticoagulant.
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • • Patients with acute pulmonary embolism should receive anticoagulant therapy for at least 3 months. The decision to continue treatment indefinitely depends on whether the associated reduction in the risk of recurrent venous thromboembolism outweighs the increased risk of bleeding and should take into account patient preferences.
    • Patients should be followed longitudinally after an acute pulmonary embolism to assess for dyspnea or functional limitation, which may indicate the development of post–pulmonary-embolism syndrome or chronic thromboembolic pulmonary hypertension.
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • “Occult cancer is detected in 5.2% of patients within 1 year after a diagnosis of unprovoked pulmonary embolism. An extensive screening strategy may detect more cancers than limited screening, but data are limited as to whether such screening is associated with better patient outcomes. Experts recommend limited cancer screening guided by medical history, physical examination, basic laboratory tests and chest radiographs, and age-specific and sex-specific cancer screening.”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • "Appropriate management of subsegmental pulmonary embolism (a single isolated subsegmental pulmonary embolus or multiple emboli, without the presence of pulmonary embolism in segmental or more proximal pulmonary vessels and without deep-vein thrombosis in the legs) is uncertain. Although some guidelines suggest clinical surveillance instead of anticoagulation in patients with low-risk subsegmental pulmonary embolism, a recent prospective cohort study involving such patients who were treated without anticoagulation therapy showed a higher-than expected  incidence of recurrent venous thromboembolism during 90-day follow-up A randomized, placebo-controlled trial of clinical surveillance as compared with anticoagulation in this patient population is ongoing (ClinicalTrials.gov number, NCT04263038).”
    Pulmonary Embolism
    Susan R. Kahn, Kerstin de Wit
    N Engl J Med 2022;387:45-57.
  • • Serial cardiac troponin (cTn) biomarkers, preferably high-sensitivity cardiac troponin (hs-cTn), are useful for rapid detection and exclusion of myocardial injury (class 1 strength of recommendation; level B-NR quality of evidence [nonrandomized]).
    • Structured risk assessment and evidence-based clinical decision pathways(CDPs) should be used to facilitate disposition and guide diagnostic evaluation (class 1 strength; level B-NR quality).
    • Low-risk patients with acute or stable chest pain may be discharged home without urgent cardiac testing (class 2a strength for acute chest pain, class 1 strength for stable chest pain; level B-R quality [randomized]).
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022
  • • For intermediate-risk patients with acute chest pain and no known coronary artery disease (CAD), coronary computed tomographic angiography (CCTA) is useful for exclusion of atherosclerotic plaque and obstructive CAD (class I strength; level A quality).
    • For intermediate-risk patients with acute chest pain and no known CAD, functional testing (eg, exercise electrocardiography, stress echocardiography, stress positron emission tomography/single-photon emission computed tomography myocardial perfusion imaging, or stress cardiac magnetic resonance) is useful for diagnosis of myocardial ischemia (class I strength; level B-NR quality [nonrandomized]).
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022
  • • For intermediate-risk patients with acute chest pain and no known CAD, functional testing (eg, exercise electrocardiography, stress echocardiography, stress positron emission tomography/single-photon emission computed tomography myocardial perfusion imaging, or stress cardiac magnetic resonance) is useful for diagnosis of myocardial ischemia (class I strength; level B-NR quality[nonrandomized]).
    • Clinically stable patients presenting with chest pain should be included in decision-making. Information about risk of adverse events, radiation exposure, costs, and alternative options should be provided to facilitate the discussion.
    Evaluation and Diagnosis of Chest Pain
    David G. Beiser, Adam S. Cifu, Jonathan Paul
    JAMA Published online July 1, 2022

Privacy Policy

© 1999-2022 Elliot K. Fishman, MD, FACR. All rights reserved.