Chest: The Advantages of Multidetector CT Scanning with a Focus on CT Angiography

Elliot K. Fishman M.D.
Karen Horton M.D.

Introduction

As we enter the first year of this new millennium we are promised a changing world whether that change be in telecommunications, computing, marketing or medical care. This first year is surely shaping up as an exciting one in the field of CT scanning with the beginning of wider distribution of multidetector/multislice CT scanners (MDCT/MSCT). While we can at times become almost blasé toward system improvements and upgrades (oh boy! another upgrade) the development of MDCT/MSCT is in our opinion nearly equal to the change from dynamic CT to spiral CT, or from classic pitch 1 spiral CT to subsecond variable pitch spiral CT scanning. To summarize our conclusions of MDCT/MSCT we can simply say that every study we could do before we can do better (i.e. search for lung or liver metastases, oncologic tumor staging, CT angiography) and that we now have the capabilities to expand our horizon to studies previously outside the realm of CT scanning (i.e. extended length CT angiography or coronary artery angiography). In this article we will review many of these changes and look at their impact on daily practice focusing on the applications in CT angiography and 3D imaging. Due to the constraints of article size we will not attempt to comment on every single advantage or application but provide an impetus for your further exploration.

Technical Aspects of MDCT/MSCT

Over the years their have been multiple improvements in CT scanning beginning with the earliest scans where single slice acquisition times and well as reconstruction times were measured in minutes to today's scanners where they are measured in fractions of a second. We have always realized several basic concepts in our drive to optimize CT scanning techniques and protocols. This can be summed up as the need for faster scan acquisition (goal is as close to zero as possible), faster data reconstruction (goal is as close to zero as possible) and better image resolution (goal is submillimeter range) . Although specifications and published performance standards are important, the practicing radiologist tends to worry less about chip clock speed and underlying operating system (Unix vs. Windows NT vs. Linux), and more about possible scanning protocol and CT image quality. MDCT/MSCT meets the radiologists needs by combining fast acquisition, fast data processing and high spatial resolution.

The need for faster data acquisition is addressed by faster gantry rotation times (.5-.75 sec) as well as the use of multiple detectors as opposed to a single detector. Depending on the manufacturer (Siemens Medical Systems, General Electric Medical Systems etc.) and model of your MDCT/MSCT scanner as well as the parameters selected the multislice scanner can be up to 8 times faster than a 1 second, single slice spiral scan. Once acquired the data can be processed at rates of around 2 slices per second. The incremental spacing used can range from as little at .5mm to whatever interval is deemed appropriate. Finally scanner spatial resolution is now in the range of 32 line pairs, which is a substantial improvement to earlier scanners. (1-3)

Our MDCT/MSCT ( Siemens Somatom Plus 4 Volume Zoom) uses four detectors at a time and can use 1mm detectors (4x1mm), 2.5 mm detectors (4x2.5) or 5 mm detectors (4x5). Depending on the parameter selection images can be reconstructed at slice thickness ranging from 1 mm to 10 mm. For example with the 2.5mm detectors the slice widths possible are 3-8 mm and 10mm. Selection of the slice collimation can be done after the data is acquired unlike with single detector CT where the slice collimation is chosen prior to data acquisition and is fixed . However the selection of the detector channels used becomes critical as the 1mm detectors (4x1 mm) are needed if slice widths of 1, 1.25, 1.5, 2 mm are to be used. Slice widths of 3-5 mm can also be obtained with 1 mm collimation as well.

Post processing of these large volume datasets is another challenge, which is being met with variable success in the marketplace. While in the recent past system requirements for such processes as 3D imaging dealt with 80-150 slices today's workstations need to be able to routinely process 200-500 scan slices. This is being met by several workstations at the present time including the GE Advantage Workstation and the Siemens 3D Virtuoso workstation. These and other systems are focusing on current needs as well as future directions with real time imaging and volume rendering as a focus. Typically the workstation platforms are from leaders like Silicon Graphics Inc (SGI- Mountain View, Ca.), and Sun Microsystems Inc. (SUN- Mountain View, Ca.) where technology continues to provide ever faster systems with better performance specifications. Other challenges with these large datasets are core problems including image transfer rates on the system and/or hospital network as well as storage requirements both on and off line. These large data sets are also a challenge for radiologists who rely on film interpretation rather than computer reading stations.

Clinical Applications

The growth in CT volumes being experienced today can be attributed to numerous factors but surely would be in great part be due to the tremendous success of CT as witnessed by the hundreds of published articles helping to define its varied clinical applications ranging from applications like suspected appendicitis or renal calculi to staging neoplasms like lung cancer and pancreatic cancer. These applications are all enhanced by the use of MDCT/MSCT. Although specific statistics are hard to quote because of the newness of this technology we can look at some logical reasons for improvement in some of the classic CT applications. In addition MDCT brings CT angiography and 3D imaging to a new plateau and will drive it to become a part of mainstream CT. The use of CT Angiography will be the focus of our review of clinical applications.

Liver Tumors

Whether the clinical problem is to detect the presence of liver metastases in a patient with colon cancer or to evaluate a patient with cirrhosis for possible hepatoma the ability to acquire the information in a shorter time frame allows for truer arterial phase images and well as better timed portal venous images. (3-4) The pure arterial phase images are ideal for detection of small but highly vascular liver tumors. The routine use of narrower collimation and closer interscan spacing also play a role in lesion detection. 3D imaging of the dual phase images is especially valuable in planning hepatic resection but is also useful for creating 3D arterial or venous maps. With MDCT we have found that we routinely can follow vessels and their branching patterns for up to 30% further out. The sharpness and detail of vessels whether viewed with volume rendering technique or maximum intensity projection also increases. The presence of arterial or venous invasion is optimized with MDCT.

One potential problem with MDCT is that we are increasing the number of small (2-4 mm) hepatic lesions detected. Although the majority of these are benign even in the oncologic patient, their detection can prove to be a clinical nightmare since they are typically too small to biopsy and require simple follow up only.

Pulmonary Embolism

The published results with single detector spiral CT scanning have shown an accuracy rate that has been reported as over

90% for up to 5th order branches of the pulmonary artery. (5) Yet, limitations do exist including the potential problems with long breatholds in these patients needed to acquire a scan with narrow (3 mm or less) collimation. Motion during to insufficient breathold can lead to both false positive and negative studies. MDCT limits these problem by providing both faster acquisition while at the same time providing narrow collimation. A routine pulmonary embolism study from the diaphragm to above the arch can be done with 3 mm collimation in as little as 10-14 seconds depending on the volume that needs to be scanned. An additional advantage of the MDCT is that we easily follow a scan of the chest with delayed scans of the pelvis and thigh in order to exclude these as sites of thromboembolism. This promises to be a single one stop exam that may replace not only Nuclear Medicine Ventilation Perfusion scanning but doppler ultrasound as well for this application. (6) Although 3D imaging with vascular mapping is not a classic part of pulmonary embolism evaluation it may prove helpful in select cases.

Aortic Aneurysm/Dissection

Spiral CT coupled with CT angiography has been shown to be equal to conventional angiography for the detection and mapping of aneurysms and dissection. CT Angiography has proven to be useful for not only diagnosis but is used as a guide for stent selection and placement. (7-9) MDCT allows a more detailed analysis of the aorta by acquiring data in a shorter breathold and with a shorter acquisition time. This limits artifact both related to cardiac pulsations as well as to patient breathold. This is especially important in imaging the ascending aorta and aortic root as well as the smaller branch vessels. The longer spiral acquisition also provides the capability to easily scan the iliac and femoral vessels in a single breathold. The higher spatial resolution of MDCT coupled with decrease/elimination of artifact allows vessel to be seen in better detail, which is important in defining branch vessel stenosis or occlusion.

MDCT is also proving valuable as an imaging tool for non-invasively following patients after aneurysm repair with either graft or endovascular stent. Complications including graft occlusion or leak can be easily recognized. The use of MDCT also allows the use of 20-40% less IV contrast which is important in scanning the older population where renal function may be less than optimal.

Renal Transplant Donors

Although we have been successful in the preoperative imaging of potential renal transplant donors with single detector spiral CT (10-11) the use of MDCT provides more detail than previously possible. We are able to now get true arterial phase images without the problem of significant filling of the venous structures like the renal veins. We can see the branching of the renal arteries more clearly and for a further distance within the kidney. Small accessory renal arteries are easier to detect especially when they are in more atypical locations.. Similarly on the venous phase images we can now routinely define smaller structures like the adrenal vein, which are important especially when harvesting of the transplant kidney is done by laprascopic nephrectomy. The technique for renal transplant donor requires a dual phase spiral CT followed by a delayed topogram to visualize the ureters. Our typical scan delay times are a 25-second delay for the arterial phase and a 55-second delay for the venous phase imaging.

Oncology: Tumor Staging and Therapy Design

The ability to detect smaller tumors in such diverse anatomic areas as the lungs, liver and pancreas is increased with MDCT as long as scan protocols are optimized to take advantage of the unique capabilities of MDCT . In terms of CT angiography, the ability to better define vascular anatomy should improve staging of tumors especially those where vascular involvement is the key to a patient being a surgical or nonsurgical candidate. For example in pancreatic cancer the creation of arterial and venous vascular maps are the final steps in the concept of a single comprehensive imaging study. We can not only use the dual phase images to pick up smaller non-enhancing tumors (adenocarcinoma) or enhancing tumors (insulinomas) but to better determine eventual resectability. Although axial images are usually diagnostic CT angiographic maps with 3D-volume rendering are ideal to better define the presence and extent of vascular involvement. (12-13 ) Variations in vascular anatomy like accessory hepatic arteries are also easier to define preoperatively with MDCT. Similarly vessel stenosis can be accurately defined with MDCT. With this information it is not suprising that in the future conventional angiography should play little if any role in staging pancreatic tumors.

Similarly in renal tumors dual phase MDCT scanning coupled with CT angiography provides a comprehensive study with optimal lesion detection as well as providing the opportunity for presurgical planning. Whether classic nephrectomy or partial nephrectomy is planned CT angiography provides the roadmap for optimizing the chance for a successful outcome. MDCT is especially useful in cases where a partial nephrectomy is planned by providing a well defined map of the tumor and its relationship to the renal vessels and pelvocalyceal system. (14) In cases of partial nephrectomy we routinely obtain the second dataset at a late venous phase closer to equilibrium. With over 40% of renal cancers now being picked up by serendipity it is important to remember than if patient selection is done properly the outcome of classic nephrectomy and partial nephrectomy are similar.

Future Directions

The future of MDCT appears to be unlimited. Applications ranging from evaluation of intestinal ischemia (figure )to coronary artery imaging are directions where MDCT with 3D imaging is going. However carefully constructed clinical trials will be needed to document statistically many of the advances that we described or many of those that will be addressed in the coming months. The ability to use lower radiation dose CT scan protocols as well as the ability to diminish iodinated contrast volumes will need to be addressed. The challenges in data processing and manipulation needed to take advantage of this new technology will also need to be addressed in the short term if MDCT is to meet its true manifest destiny. Despite these challenges what we find especially exciting is the seemingly endless opportunities for MDCT or to quote President Franklin D. Roosevelt " The only limit to our realization of tomorrow will be our doubts of today. " If there are any doubts about the future of CT, MDCT eliminates them.