Musculoskeletal: Spiral CT of the Musculoskeletal System: Principles, Techniques, and Clinical Applications
Elliot K. Fishman, M.D.
Despite the widespread belief that Magnetic Resonance Imaging (MRI) is the modality of choice for imaging the musculoskeletal system, computed tomography (CT) remains extremely effective for a wide range of clinical applications. Spiral CT provides definite advantages over standard dynamic CT in nearly all of these applications and in clinical practice is almost universally the technique of choice (1-4). The specific advantages of spiral CT enable us to define its role in musculoskeletal imaging. Spiral CT is particularly well-suited in the following clinical applications.
(1) Examinations where even minimal intrascan or interscan motion may compromise the study. Typical problem areas have been the cervical spine, shoulder, sternum, and wrist. Lack of interscan motion is especially important in any study of these regions when multiplanar or three-dimensional imaging (3D) is planned.
(2) Smaller anatomic regions of interest like the foot or wrist benefit from a spiral acquisition of a volume dataset combining narrow collimation (2-3 mm) and a pitch of 1 with small reconstruction increments (1-2 mm) .
(3) CT studies for suspected infection in muscle or soft tissue, or a suspected soft tissue or muscle mass require iodinated intravenous contrast to optimize detection of the presence and extent of disease (4-6). Spiral CT allows data to be acquired during the phase of maximum contrast enhancement thereby optimizing lesion detection. The enhancement patterns may also prove useful in the differential diagnosis of a lesion. Optimization of vascular enhancement is also valuable for definition of vascular anatomy. Three dimensional images using maximum intensity projection (MIPS) or volume rendering techniques require optimal contrast administration (7-10) if vascular maps are to be generated . Delayed images may be helpful in select cases.
Examination Technique and Scanning Protocols
Specific spiral CT scanning techniques will depend on the clinical problem to be evaluated. Scanning parameters that must be selected include slice thickness or collimation, table speed and interscan spacing. These decisions typically are based on the distance of the area to be scanned and whether multiplanar reconstruction and/or 3D imaging is to be obtained following the routine CT scan. Use of intravenous contrast material will be variable but a good rule of thumb is that if soft tissue or muscle is to be evaluated then contrast is always helpful. This is true whether one is looking at tumor or inflammatory disease.
Several specific scanning protocols are commonly used and can be modified to meet the specific scan parameters available on any commercially available spiral scanner. The parameters listed here should therefore only be used as a guide and may need to be modified. Please consult your CT scanner’s manual or contact your applications specialist for specific recommendations.
If the study is primarily to evaluate isolated musculoskeletal trauma, for example, following a motor vehicle injury to the bony pelvis or shoulder, we would typically not use any intravenous contrast material. Depending on the area to be scanned, the spiral length will vary from 32 to 40 seconds and will generally use a pitch of 1 to 1.5 . Depending on the scanner used, a pitch up to 2 will also be satisfactory in most cases although in our experience will result in images that are a bit fuzzy. Slice thickness is usually 3mm although 5 mm collimation can be used to cover longer distances. Our reconstruction interval would typically be 2 or 3 mm with 3mm collimation and 5-8 mm with 5mm collimation . The kVp is 120 with 280 mAs on a Siemens Somatom-Plus-4 scanner. If the area of interest is a smaller anatomic area, such as the sternoclavicular joint or possibly the carpal bones and then spiral CT can be done with thinner sections (1-2mm) and a pitch of 1-1.6. Data reconstruction will be in 1mm increments. In reconstructing the data in skeletal trauma cases we typically use the high or ultra high reconstruction mode. This tends to make the images sharper, which is of critical importance when generating images to detect subtle fractures. In terms of the quality of the 3D reconstruction, we have generally found little problem with using a high or ultra high filter. Occasionally, the ultra high filter introduces too much noise and the 3D images may be suboptimal. If 3D rendering is required following reconstruction algorithm, a second set of reconstructions can be done with a standard algorithm. Depending on the scanner, this may or may not require saving the raw data from the spiral CT acquisition.
Although a detailed analysis of the principles and techniques of a 3D imaging is beyond the scope of this chapter, several important concepts should be reinforced. In musculoskeletal 3D imaging, the principle techniques used are shaded surface rendering and volume rendering (11-16). Although both techniques have their advocates, numerous articles have stressed the advantages of volume rendering. Kuszyk et al. (17) evaluated both techniques by comparing 3D images generated with both rendering techniques. He concluded that "surface renderings show gross 3D relationships most effectively, but suffer from more stair step artifacts and fail to effectively display lesions hidden behind overlying bone or located beneath the bone cortex. Volume rendering algorithms effectively show subcortical lesions, minimally displaced fractures, and hidden areas of interest with few artifacts. Volume algorithm show 3D relationships with varying degrees of success depending on the degree of surface shading and opacity. While surface rendering creates me three-dimensionally realistic imaging of the bone surface, it may be of limited clinical utility due to numerous artifacts and the inability to show subcortical pathology. Volume rendering is a flexible 3-D technique that effectively displays a variety of skeletal pathology with few artifacts." Current systems now allow volume rendering to be done in 'real-time' with no pre-editing of the dataset required. This is our technique of choice for all applications.
Another common clinical problem is the case where the primary area of interest is soft tissue or muscle. In these cases we are usually trying to rule out the presence of infection or a mass or to define its extent. Once again the table speed and slice thickness will vary based on the anatomic area that needs to be covered. As stated previously in cases where muscle or soft tissue need to be evaluated, contrast material is mandatory to optimize the differential enhancement between normal and abnormal tissue. We usually use a delay before scanning of approximately 50 seconds from initiation of contrast injection. If images of the lower extremity are to be obtained, a delay of up to 70 seconds may be warranted. The key to timing the acquisition of CT data after contrast administration is an understanding of iodinated contrast and its distribution in muscle. Regardless of the pathologic process, in most cases normal muscle will enhance more than either inflammatory disease or neoplastic disease. Two exceptions might be desmoid tumors and occasional metastatic tumors to muscle which are typically hyperdense or hypervascular (5,18) .
For the majority of muscle or soft tissue evaluations, a typical spiral CT study would be 24 to 40 seconds with a pitch of 1-2. Slice thickness would be 5 or 8mm
(usually 5 mm) depending on the length of the area to be scanned, and the table speed will correspond from 5 or 16mm/sec. The reconstruction interval will vary but typically range from 5 to 8mm. The kVp is 120 with 280 mAs. When soft tissues or muscles are evaluated, the standard reconstruction algorithm (or soft tissue algorithm) is used. Using a high or ultra high algorithm tends to create too much noise and detracts from the quality of the CT image. In select cases, images may need to be reconstructed using both the standard and high resolution/ultra high resolution algorithm.
Occasionally, we have found that delayed scans following the spiral CT may be helpful in better understanding a pathologic process involving muscle. Tumors such as desmoid tumors will show persistent intense delayed enhancement. Also on occasion, abscesses will become more obvious on delayed studies done 10 to 15 minutes after contrast injection. In these cases, a dense enhancing rim may be seen. As with most CT studies the examination should be tailored to the problem at hand.
Clinical Applications
The number of potential applications of spiral CT in musculoskeletal imaging cannot be covered in detail in but one chapter. Therefore, some representative applications are presented which suggest the unique capabilities of this technique.
I. Sternum and Sternoclavicular Joint
Evaluation of the sternum and sternoclavicular (S/C) joint is most often requested to rule out fracture or fracture/dislocation, define the extent of a fracture or dislocation, evaluate suspected inflammation or abscess and occasionally to clarify indeterminant or confusing findings on plain film or bone scan. The sternoclavicular joint is especially difficult to evaluate with plain radiographs and several articles have stressed the value of CT. The use of transaxial CT, supplemented by multiplanar imaging in select obliquities is especially valuable when looking at the sternoclavicular joint (5, 19-21).
Sternal or sternoclavicular injury is most often a direct result of closed chest trauma following a motor vehicle accident . Although often an isolated injury, involvement of the shoulder or ribs is not uncommon. Injuries associated with S/C joint dislocation, especially posterior dislocation, include injury to the aorta and great vessels. In these cases, intravenous contrast should always be used to exclude vascular injury. Spiral CT is invaluable in these cases by providing both an excellent vascular study as well as an excellent study of the skeletal structures involved. In these cases the protocol used is the classic CT Angiography protocol with 3 mm collimation, table speed of 5-6 mm/sec, and reconstruction of data at 1-2 mm intervals. Between 120 and 150 ml of Omnipaque-350 is injected at 3 ml/sec with a 25 second delay from the start of injection and the acquisition of data. 3D rendering is done in all of these cases.
In the patient with S/C joint injury, we routinely do multiplanar and 3D imaging. The coronal views and oblique coronal images are most helpful in defining any involvement of the sternum including displacement due to fracture. A Z-axis 3D study is optimal for looking at the orientation of S/C joint dislocations and in helping to determine the mechanism of injury. These reconstructions can be done with either a surface rendering or volumetric technique. We have found that these 3D images are most useful after associated bony structures have been edited.
Because chest trauma is often complex, it is important to carefully evaluate the entire shoulder joint complex and not overlook fractures in other structures including the scapula (22) . Scapular injuries are often overlooked in over 40 percent of cases on plain radiographs. CT excels at detecting the full extent of injuries and is especially valuable at evaluating pathology involving the scapula.
Infection of the sternoclavicular joint is more common in the patient with a history of either drug abuse, steroid use or prior surgery to the head and neck region. It is especially common in the HIV positive or AIDS patient although we have recently seen several cases in patients without any known risk factors. Spiral CT excels at defining both the soft tissue and muscle component of disease as well as the presence of associated of bony involvement . Extension into adjacent muscle (pectoralis major, pectoralis minor or sternocleidomastoid muscle) and retrosternal soft tissues is not uncommon. In rare cases, untreated mediastinitis may, in fact, develop . Spiral CT can be used to both plan therapeutic intervention and monitor response to therapy whether it be surgical and/or medical. In cases of suspected infection contrast enhancement is particularly useful. We recently reviewed a series of seven patients with infection of the sternoclavicular joint and found spiral CT to be invaluable at arriving at a correct diagnosis and in guiding management (15). When evaluating the sternoclavicular joint we have found the use of a high spatial filter to be valuable in accentuating subtle details including early bone erosions or periosteal reaction .
Spiral CT is also of value for the detection of bone fragments or foreign matter in the shoulder joint. The combination of narrow slice collimation with closely spaced sections (1-2 mm) and targeted images is invaluable for the detection of small intra-articular fragments. Multiplanar reformations or 3D reconstructions can help localize the lesion in three planes and/or perspectives. Similarly, complex humeral injuries can be mapped with spiral CT scanning.
Spiral CT scanning with narrow collimation is also valuable in the evaluation of the acromion either pre- or post-operatively . This may prove useful in patients with shoulder instability or in those with an impingement syndrome. When evaluating the shoulder with spiral CT, a protocol might be 3mm slice thickness, 3mm/sec table speed with reconstructions at 1-2mm intervals.
III. Skeletal Trauma
Possibly the most obvious ideal application for spiral CT is in the evaluation of musculoskeletal trauma. Since closely spaced scans can be obtained in a short scanning cycle, there is likely to be less chance for intrascan or interscan motion. Once a volumetric data set is generated the images can be used for multiplanar and three- dimensional reconstruction. The value of rapid acquisition is particularly apparent in trauma patients when the trauma involves areas where patients may have difficulty remaining still such as the shoulder, sterno-clavicular joint, elbow, or wrist. Because arbitrary interscan spacing can be chosen following data acquisition, it is possible to obtain very closely spaced scans as needed for image post-processing. Prior work has shown that the quality of multiplanar reconstruction and three-dimensional images from helical or spiral CT is equal to that of conventional CT (24). In an article by Ney (24) to test whether spiral CT scans provided as images equal to or superior to those of standard dynamic CT, two objects were used to study the effects of spiral CT: an angled cylindrical bone phantom and a human cadaver femur specimen with a simulated 1mm fracture. Both objects were scanned in a water bath, and a series of spiral and standard CT scans were obtained with various parameters. Volumetric rendering was then done with the resultant datasets to create three-dimensional images. Three radiologists reviewed the images to rate fidelity, accuracy, and diagnostic usefulness. The results showed that for similar parameters (slice thickness, interscan spacing) spiral and dynamic CT data resulted in images similar in quality. However, since spiral CT is approximately five times faster than dynamic CT it is possible to use thinner collimation and obtain more sectional data with spiral CT.
Several other studies have shown similar results. McEnery et al. (25) studied the capabilities of spiral CT vs. conventional CT to represent minimal fracture displacement on multiplanar reconstruction images and found that with correctly selected scan parameters that "spiral CT derived reconstructions demonstrate similar edge profile resolution to reconstructions obtained from conventional CT." Link et al. (26) studied artificial spine fractures with spiral and conventional CT and found that "helical CT requires thinner collimation for fracture detection comparable with that of conventional CT."
Several authors have recently questioned the value of conventional CT vs. spiral CT for musculoskeletal image reconstruction and 3D imaging (27). However, these studies have been in phantom experiments where real life problems such as patient cooperation and potential interscan or intrascan movement as well as time to complete study are not considered. It is our experience that in the clinical environment, spiral CT is the technique of choice in the trauma patient.
Although this chapter cannot review in detail all of the specific applications for CT imaging of trauma, several of the more common applications should be noted. These anatomic zones including the pelvis and acetabulum, the knee including the tibial plateau, the ankle joint the wrist and the spine. Even in the pre-spiral CT era it became clear that transaxial CT supplemented by multiplanar reconstruction (MPR) and 3D imaging could have a major impact on both diagnosis and patient management. Although precise numbers vary by clinical application we found a consistent 20-30 percent rate of altered management by using MPR and 3D imaging vs. transaxial CT alone. These change in management were predominantly two types: tentative surgery scheduled due to a situation worse than anticipated and acute surgery deferred in favor later definitive arthrodesis or arthroplasty, again usually when the images revealed a clinical picture worse than anticipated.
In acetabular and pelvic trauma, spiral CT datasets coupled with a real-time 3D volume rendering program allows visualization of the entire pelvis through any plane or perspective (28). The interactive nature of an imaging display was previously shown to be of value in creating arbitrary 360 degree rotation into any inlet or tangential view desired. Editing of the dataset is especially of value in isolating the fracture and in select cases disarticulating the femur from the acetabulum may be useful. By scanning and creating 3D maps of the entire pelvis we can easily detect any associated sacral or sacroiliac injuries .
Finally in a single examination we can also use intravenous contrast to create vascular maps of the iliac and femoral vessels to rule out any associated vascular injuries . Vascular injuries are critical components of complex pelvic fractures and in the past were studies exclusively with catheter angiography. With spiral CT angiography the need for classic angiographic studies may be replaced in may cases.
CT is especially useful in lower extremity trauma involving either the knee joint or the ankle. In the patient with a tibial plateau fracture, spiral CT with sagittal and coronal reformatting of data is an important study in defining whether or not a patient needs surgical intervention. The use of these displays coupled with 3D images is ideal for defining plateau depression and quantifying it . In cases of proximal tibiofibular dislocation the 3D images are especially valuable. McEnery et al. (29) did an analysis to determine the value of spiral CT for detecting displacement of fractures of the tibial plateau. The authors found that "spiral CT can detect clinically important inferior depressions of tibial plateau fractures." The authors felt that to achieve optimal results a scan protocol of 2mm section collimation, 2mm/sec table speed and image reconstruction at 1mm intervals. Trauma to the distal femur can also benefit from spiral 3D studies in select cases
Severe ankle trauma also illustrates the role of multiplanar and 3D imaging in finalizing assessment and surgical planning (30) . Pilon fractures, with severe impaction and destruction of the articular plafond, may be triaged into those patients needing immediate surgery and those who will be treated later with arthroplasty. Injuries to the talus, calcaneus or tarsal bones are well imaged with spiral CT protocols. Followup of patients whether management be surgery or a more conservative is easily done with spiral CT. MPR and 3D images are often successful even in the face of metal pins and plates .
Spiral CT with direct coronal reconstructions is an excellent approach to the traumatized wrist (31). This technique combines 2mm collimation, 2-3 mm/1sec table speed and reconstruction of data at 1mm intervals. The technique is successful in evaluating occult or complex fractures, and evaluate the postsurgical wrist to determine healing. High resolution algorithms are helpful in these cases as are targetting image reconstructions in order to achieve the best details in the dataset .
Trauma to the spine can be routinely visualized successfully with a combination of transaxial CT, MPR images and 3D studies. Nunez et al. (32) analyzed the standard radiographs and spiral scan in 88 patients and found that in 32 patients (N=50) fractures were either missed or incompletely defined on conventional studies and defined on CT. The authors in fact felt strongly enough to recommend routine screening with spiral CT in polytrauma victims.
Similar success has also been noted in the rest of the spine. In the thoracic and lumbar spine, CT coupled with sagittal reconstruction and 3D images helps define the presence and extent of injury. Specific applications include in addition to fracture detection, detecting subluxations and locked facts, as well as localizing foreign matter like bullets. Sacral fractures, which can be overlooked on plain radiographs are easily detected with spiral CT .
IV. Soft Tissue or Muscle Evaluation
The evaluation of soft tissue or muscle infection and/or tumor is another clinical application for spiral CT. With the use of iodinated contrast material, we can scan through an area of suspected musculoskeletal abnormality during peak levels of contrast enhancement. It has been previously documented that contrast enhancement allows for better detection of intramuscular pathology whether it is inflammatory or neoplastic (33-34). With spiral CT, we can get an excellent idea of the vascularity of the lesion, and its detectability is enhanced by optimizing normal muscle enhancement. The local vascular supply is also optimally visualized which can be valuable in patients with tumors where resection is being considered. Vascular thrombosis, aneurysms or pseudoaneurysms are also easily detected with these techniques . The timing of spiral image acquisition after contrast injection is critical. We have found that a delay of about 50-60 seconds is ideal in most clinical situations. In cases where an AV malformation or another arterial process is suspected then images at 25-30sec are preferred. In cases of vascular processes, surface or volumetric three-dimensional reconstruction or maximum intensity projection images may be helpful.
One patient population where we have found spiral CT of the muscle to be especially helpful is in the AIDS patient. This population appears to have an increased incidence of muscle infections and abscesses although in many cases the presentation is best occult or subtle. Contrast-enhanced spiral CT optimizes detection of lesion during the pre-equilibrium phase, even in patients with poor tissue planes . The extent of involvement is well-demonstrated on multiplanar reconstruction of the spiral CT dataset. Multiplanar imaging is particularly useful in surgical planning especially when the inflammatory process is extensive . In many of these cases, delayed scans may also be helpful in defining the full extent of pathology.
Spiral CT is also useful in distinguishing vascular masses from hematoma, abscess, or tumor. In the post-operative vascular patient, aneurysms or pseudoaneurysms may present as a palpable mass. Although in many cases the diagnosis can be made clinically or with Doppler ultrasound, in other cases spiral CT with 3D rendering is most valuable. As more experience is gained with spiral CT angiography, its role continues to expand in vascular imaging across a wide range of applications.
One interesting finding related to spiral CT is the frequency of detection of musculoskeletal metastasis as incidental findings. We have seen many cases of metastasis to muscle presenting as hypervascular lesions ranging in size from 5 mm to 5cm. In most cases the patient have not been symptomatic due to these lesions. The common tumors where we have found this to occur have been lung cancer, breast cancer, and lymphoma.
V. Skeletal Tumors
Spiral CT is helpful in defining the full extent of primary or metastatic bone tumors. This information can be used for either image analysis or therapy planning whether it is surgical, radiation therapy, or chemotherapy. Multiplanar and/or 3D reconstruction may be of special value in this group of patients. Spiral CT is especially valuable in areas such as the ribs, spine, sternum, and shoulder when other examinations are equivocal for tumor infiltration . The combination of thin section CT (2-4 mm) with narrow interscan spacing (1-3 mm) can be useful for the detection of subtle tumor infiltration.
In bony metastases or primary tumors, spiral CT can help define the presence and extent of disease. Although bone scans are an excellent screening study for metastases, CT is more valuable when symptoms are localized to a specific zone or anatomic region. In these cases, targeted exam are the ideal screening study. Multiplanar and 3D studies are often critical in these cases for defining the true extent of lesions for the referring orthopedic surgeon or oncologist . The recognition that image display is important for the referring physician is often overlooked in a busy radiologic practice.
CT is also worthwhile in patients with benign tumors such as suspected osteoid osteoma . The use of narrow collimation (2 mm) and small interscan spacing (1mm) allows for detection of the nidus even in the most difficult of cases. This information can potentially be used for percutaneous CT-guided removal of the nidus or for standard surgical planning (35).
VI. Problem Solving Studies
The initial role of CT scanning in the musculoskeletal system was as a problem solving tool. CT often became the final arbiter in cases when there were conflicting findings between different radiologic studies, clinical presentation, and/or physical examination . Spiral CT helps increase the value of CT in this clinical situation. Spiral CT scanning provides a volume dataset that can allow for multiple sections through an area of suspicious pathology. Spiral CT is excellent in detecting the presence of subtle lesions with even minimal destruction or resorption. In most cases when bone involvement is evaluated, intravenous contrast material is not used. It may be helpful, however, in demonstrating soft tissue extension of tumor and vascular invasion. It is therefore important to understand the clinical questions to be answered prior to performing the spiral CT examination.
Conclusion
Spiral CT combined with multiplanar reconstruction and 3D imaging has replaced conventional tomography in a wide range of applications. The role of spiral CT musculoskeletal imaging undoubtedly will increase with more clinical experience as well as with the advancement of spiral CT technique and the refinement of reconstruction algorithms to improve image quality.
In early publications on spiral CT several specific limitations of musculoskeletal spiral CT scanning were noted. The key limitations included a relatively low mA value on some systems for the basic 24-32 second spiral as well as bone reconstruction algorithms that did not seem to create high resolution images that are as sharp as those generated from non spiral CT data. Additionally, the length of the spiral CT scan (32sec) was often not long enough when a larger area was to be imaged for multiplanar or 3D reconstruction. However, these problems have been solved on most systems and the potential roadblocks to imaging h as been removed.
The current trend by major CT manufacturers toward workstations will expand the routine use of multiplanar and 3D imaging, two applications for which spiral CT is particularly well suited. Other applications, including surgical planning, custom hip design, and joint reconstruction will all benefit from the use of spiral CT and we look forward with cautious optimism to its continued growth . The impact of multidetector CT will be interesting in the next few years.,
References
1. Fishman EK, Wyatt SH, Bluemke DA, Urban BA. Spiral CT of musculoskeletal pathology: preliminary observations. Skeletal Radiol 1993; 22:253-256.
2. Heiken JP, Brink, JA, Vannier, MW. Spiral (Helical) CT. Radiology 1993; 189:647-656.
3. Fishman EK. Spiral CT evaluation of the musculoskeletal system In: Spiral CT: Principles, Techniques and Clinical Applications, Fishman EK, Jeffrey RB, eds. New York: Raven Press Ltd, 1995.
4. Pretorius ES, Fishman EK. Helical (Spiral) CT of the musculoskeletal system. Radiol Clin NA 1995; 33(5):949-978.
5. Tecce PM, Fishman EK. Spiral CT with multiplanar reconstruction in the diagnosis of sternoclavicular osteomyelitis. Skeletal Radiol 1995; 24:275-281.
6. Scott WW Jr, Fishman EK. Soft Tissue Masses (in) Computed Tomography of the Musculoskeletal System (editors) Scott WW Jr, Magid D, Fishman EK. Churchill Livingstone, NY, NY, 1987; 8:1-27.
7. Dillon EH van Leeuwen MS, Fernandez MA, Mali WPTM. Spiral CT angiography. AJR 1993; 160:1273-1278.
8. Costello P, Gaa J. Spiral CT angiography of the abdominal aorta and its branches. Eur Radiol 1993; 3:359-365.
9. Rubin GD, Dake MD, Napel S et al. Spiral CT of renal artery stenosis: comparison of three-dimensional rendering techniques Radiology 1994; 190:181-189.
10. Scott WW Jr, Fishman EK. Spiral CT arthrography of the shoulder. Work in progress .
11. Fishman EK, Magid D, Ney DR, Chaney EL, Pizer SM, Rosenman JG, Levin DN, Vannier MW, Kuhlman JE, Robertson DD. Three-dimensional Imaging. Radiology 1991; 181:321-337.
12. Magid D, Fishman EK, Sponseller PD, Griffin PP. 2D and 3D computed tomography of the pediatric hip. RadioGraphics 1988; 8:901-934.
13. Magid D, Fishman EK. Imaging of musculoskeletal trauma in three dimensions. Radiol Clin North Am 1989; 27:945-956.
14. Ney DR, Drebin RA, Fishman EK, Magid D. Volumetric rendering of computed tomographic data: principles and techniques. IEEE Comput Graphic Applic 1990; 10:24-32.
15. Drebin RA Carpenter L, Hanrahan P. Volume rendering. Comput Graphics 1988; 22:65-74.
16. Drebin RA, Magid D, Robertson DD, Fishman EK. Fidelity of three-dimensional CT imaging for detecting fracture gaps. J Comput Assist Tomogr 1989; 13:487-489.
17. Kuszyk BS Heath DG, Bliss DF, Fishman EK. Skeletal 3-D CT: advantages of volume rendering over surface rendering. Skeletal Radiol 1996; 25:207-214.
18. Olson PN Everson LI, Griffiths HJ. Staging of musculoskeletal tumors. Radiol Clin North Am 1994; 32:151-162.
19. Pretorius ES, Scott WW Jr, Fishman EK. Acute trauma to the shoulder: role of spiral computed tomographic imaging. Emergency Radiology 1995; 2(1):13-17.
20. Kuhlman JE, Fishman EK, Ney DR, Magid D. Complex shoulder trauma: Three-dimensional CT imaging. Orthopedics 1988; 11:1561-1563.
21. Kuhlman JE, Fishman EK, Scott WW Jr, Magid D, Siegelman SS. Two-dimensional and three-dimensional evaluation of the painful shoulder. Orthop Rev 1989; 18:1201-1208.
22. Stiles RG, Otte MT. Imaging of the shoulder. Radiology 1993; 188:603-613.
23. Wilson A Totty WG, Murphy WA, et al. Shoulder joint: arthrography CT and long-term follow-up with surgical correlation. Radiology 1989; 173:329-333.
24. Ney DR, Fishman EK, Kawashima A, Robertson Jr DD, Scott WW Jr. Comparison of helical and serial CT with regard to three-dimensional imaging of musculoskeletal anatomy. Radiology 1992; 185(3):865-869.
25. McEnery KE, Wilson AJ, Murphy Jr WA. Comparison of spiral computed tomography versus conventional computed tomography multiplanar reconstructions of a fracture displacement phantom. Invest Radiol 1994; 29(7):665-670.
26. Link TM, Meier N, Rummeny EJ, Garmann S, Steinbeck J, Schroder M, Oelerich M, Haussler MD. Artificial spine fractures: detection with helical and conventional CT. Radiology 1996; 198(2):515-519.
27. Kasales CJ, Mauger DT, Sefczek RJ, et al. Multiplanar image reconstruction and 3D imaging using a musculoskeletal phantom: conventional versus helical CT. J Comput Assist Tomogr 1997; 21(1):162-169.
28. Scott WW Jr., Fishman EK, Magid D. Optimal imaging of acetabular fractures. Radiology 1987; 11:1017-1020.
29. McEnery KW, Wilson AJ, Pilgram TK, Murphy Jr WA, Marushack MM. Fractures of the tibial plateau: value of spiral CT coronal plane reconstructions for detecting displacement in vitro. AJR 1994; 163:1177-1181.
30. Magid D, Michelson JD, Ney DR, Fishman EK. Adult ankle fractures: Comparison of plain films and interactive two- and three-dimensional CT scans. AJR 1990; 154:1017-1023.
31. Kuszyk BS, Fishman EK. Direct coronal CT of the wrist: helical acquisition with simplified patient positioning. AJR 1996; 166:419-420.
32. Nunez Jr DB, Zuluaga A, Fuentes-Bernardo DA, Rivas LA, Becerra JL. Cervical spine trauma: how much more do we learn by routinely using helical CT? RadioGraphics 1996; 16(6):1307-1318.
33. Magid D Fishman EK. Musculoskeletal infections in patients with AIDS: CT findings. AJR 1992; 158:603-607.
34. Ayala A, Murray JA, Erling MA. Osteoid osteoma: intraoperative tetracycline-fluorescence demonstration of the nidus. J Bone Joint Surg (Am) 1988; 68:747-751.