Back in the 1950s, there was a rage for 3-D movies. Everyone wore paper-framed glasses with one red and one green lens and watched monsters and other Hollywood creatures appear to leap from the screen.
Once the novelty wore off, the rage for 3-D movies faded quickly. Today, there is another drive into the third dimension that will last long after it becomes commonplace. The 3-D data sets that can be captured with contemporary imaging equipment permit examination of an anatomic site from any perspective, even after the patient has left the radiology department. And it is not the diagnostic radiologists alone who want 3-D: surgeons and therapeutic radiologists increasingly demand the knowledge such images can provide. Indeed, so valuable is 3-D imaging that Elliott K. Fishman, MD, professor of radiology and oncology at Johns Hopkins Hospital in Baltimore, reports that his department does as many as 70 such procedures every week. This article looks at only a few of the many current uses of 3-D imaging.
An Early Use Continues to Grow
Depictions of the larger blood vessels were among the first applications of 3-D imaging. Renal donor selection and treatment planning for aortic aneurysm repair were among the first indications.
“To do CT angiography well, you need 3-D imaging,” Fishman stresses. “The advent of multidetector array CT and the data sets it provides have driven CT angiography. Because of the detail that is captured, we can see smaller vessels than were identifiable before. We evaluate potential renal donors and people scheduled for liver transplants. We study aortic dissections and aneurysms and identify intestinal ischemia. In fact, we are doing a tremendous number of studies for which people used to do contrast angiography.”
Another replacement for contrast angiography is MRI, and again, 3-D imaging improves the quality of the study. A meta-analysis last year by Patricia J. Nelemans, MD, PhD, and her colleagues at the University of Maastricht in The Netherlands, along with the Cochrane Methods Group on Systematic Review of Screening and Diagnostic Tests, demonstrated that 3-D gadolinium-enhanced MR angiography is superior to present 2-D time-of-flight methods in evaluating peripheral arterial disease.1 Use of various reconstruction methods and examination of the source images improves interpretation accuracy still further.
The greatest challenge to vascular imaging is the coronary arteries, with their constant motion. Francesco Sardanelli, MD, and his colleagues at the University of Genoa examined the hearts of 42 patients by navigator-echo MRI, acquiring two or three transverse slabs with 15% overlap, and looked for coronary stenoses of 50% or more. All of the left main arteries and proximal arterial segments could be seen, although only two thirds of the distal segments were visible. The sensitivity and specificity of the study for proximal stenoses were 90%. Those investigators were less satisfied with the accuracy of MRA in the distal segments of vessels.2
The vasculature of the brain is another subject of 3-D imaging, both for diagnosis and for treatment and follow-up. This subject will be discussed in a future article.
Applications in Urology
Urologists were among the pioneers of minimally invasive surgery, and today, few procedures are done by traditional open methods. As a corollary, urologists have become dependent on imaging. Three-dimensional imaging has proved valuable in defining the site and size of prostate or kidney cancer.
 Figure 1. 3D CT angiogram in the anterioposterior projection demonstrating the arteries of the upper abdomen. Image courtesy of Rendon C. Nelson, MD, department of radiology, Duke University, Durham, NC. |
With serum assays for prostate specific antigen and digital rectal examination, most prostate cancers today are small when they are discovered. Most patients are asymptomatic and are interested in minimally invasive treatment, such as brachytherapy or conformal radiation, rather than the traditional radical surgery. When the size of a cancer makes radical surgery necessary, there is pressure to preserve the neurovascular bundles responsible for erections. The urologist or radiation therapist therefore needs to know precisely how the tumor is distributed in the gland. For larger tumors, the question is whether the tumor has penetrated the capsule, which makes it incurable by surgery. In the past, multiple biopsies generally were used to answer these questions, a method that has a significant error rate.
In a study presented at the Radiological Society of North America meeting in 1997 and published in 1999, J?rgen Scheidler, MD, and his colleagues at the University of California-San Francisco examined 53 men who had biopsy-proved prostate cancer with a combination of endorectal MRI and 3D MR proton spectroscopy (MRS).3 Each sextant of the prostate image was evaluated separately for the presence of cancer. On MRS, a ratio of choline + creatine to citrate that was 2 SD or more above the population norm was considered possible evidence of cancer, and a ratio of 3 SD or more above the norm was taken as proof of cancer. When judged by the histopathologic findings, the combination of MRI and MRS had a specificity for cancer as great as 91%. When either MRI or MRS indicated cancer, the sensitivity was as high as 95%.
 Figure 2. 3D CT angiogram in the left posterior obligue projection demonstrating a single right renal artery. Image courtesy of Rendon C. Nelson, MD, department of radiology, Duke University, Durham, NC. |
These radiologists also explored the ability of this combination of imaging methods to detect extracapsular extension.4 The? addition of MRS data significantly improved the ability of the less experienced radiologist to identify extracapsular spread; the accuracy of the more experienced reader also was increased. Significantly, use of MRS as well as MRI reduced interobserver variability in the assessment of extracapsular spread.
The greater sensitivity of current cross-sectional imaging modalities is not an unmixed blessing. It is not unusual today to discover an asymptomatic renal mass in a patient undergoing imaging for some non-urologic indication. If this mass proves to be cancerous, radical nephrectomy has been the traditional treatment, but more urologists now believe that for smaller masses, partial nephrectomy, perhaps by laparoscopy, is sufficient if the surgical margins are tumor free. In planning such surgery, contrast-enhanced spiral CT is invaluable for portraying the true extent of the tumor, its blood supply, and its relations to the collecting system. One team of urologists and radiologists used volume-rendering software to create videotapes of 3-D renal images from triphasic spiral CT scanning data. The tapes were reviewed periodically in the operating room to assist in tumor resection and reconstruction of the renal remnant.5
Cancer Treatment
Another cancer for which less radical therapy is becoming popular as a result of earlier diagnosis is breast cancer. Again, success depends on accurate definition of the extent of the cancer. Preoperative mammography is not sufficiently sensitive: in some series, as many as half of women who underwent lumpectomy have been found to have positive margins and require additional surgery. G. Amano, MD, and his colleagues in the Department of Surgical Oncology at Tohoku University School of Medicine, Sendai, Japan, correlated the findings from 3-D MRI with the histopathological findings in 58 women with breast cancer.6 They identified three patterns of tumor growth: localized, segmentally extended, and irregularly extended. Among the 30 localized cancers, 10 had sent out thin fingers of ductal carcinoma in situ or invasive lobular carcinoma that could not be seen by MRI. For the other two patterns, however, 3-D MRI did well in identifying peripheral invasive cancer and was more accurate than mammography and ultrasonography, either alone or in combination. “MRI may provide additional information concerning carcinoma extension prior to surgery,” those authors concluded, making selection of the appropriate procedure easier.
Squamous-cell carcinoma or other cancer of the head and neck can be well defined by 3-D spiral CT. Radiologists at the University of Chicago compared four methods of reconstruction of contrast-enhanced spiral CT data as a means of staging cancers of the head and neck.7 Collimations of 3 and 5 mm were used with reconstruction at no more than 3-mm intervals. Images were examined in sagittal, coronal, and oblique views. Various reconstruction algorithms proved useful for different purposes. For example, maximum intensity projections were valuable for assessing the patency, compression, occlusion, or displacement of blood vessels, whereas a combination of shaded surface display and multiplanar reconstruction provided excellent depictions of the lymph nodes and the relations of the tumor to surrounding tissues. Three-dimensional reconstructions were valuable for localizing tumors in the neck and for surgical or radiation therapy planning.
Lung and Colon Cancer Screening
Three-dimensional imaging is finding two more roles in cancer diagnosis and treatment. One is histopathologic analysis. There is a growing interest in screening high-risk persons for lung cancer using low-dose CT. Many nodules are found, most of which are not malignant. Among the methods being used to identify those that are is 3-D image extraction with isotropic resampling. This method can measure the volume of 3- to 4-mm artificial lesions with an accuracy of 3%. By measuring the change in the volume of a lesion at intervals, it is possible to calculate the doubling time and thereby distinguish a cancer from a benign mass without a biopsy.8
Perhaps the most exciting application of 3-D imaging techniques is for the creation of virtual images of the interior of the body; eg, the gastrointestinal tract. The hope is to eliminate many endoscopic examinations.
The bulk of the recent work on virtual endoscopy has been done in the colon with the hope of detecting clinically significant polyps before they become invasive cancers. In a trial supported by a grant from the National Cancer Institute’s Prostate-Lung-Colon-Ovary Screening Trial, Elizabeth G. McFarland, MD, of Washington University in St Louis and her colleagues at Yale and the University of Vermont demonstrated that 3-D reconstructions of data acquired by spiral CT increased the accuracy of polyp identification.9 For polyps or cancers >10 mm, the sensitivity and positive predictive value were as high as 100%. Use of 3-D imaging increased interobserver agreement.
Unfortunately, virtual colonoscopy is not yet ready for routine use. The best results with the fewest ventilatory artifacts appear to be obtained with multidetector array scanners,10 putting the examination out of the reach of many centers at present. Also, the study requires a full colon preparation, insufflation of the colon, and examinations in both the prone and supine positions, making it unappealing to patients. Moreover, the radiologist usually must construct transverse, multiplanar reformation, and three-dimensional endoluminal images. Finally, all published trials have been in patients with known polyps or cancers; screening studies have apparently not been undertaken.
Nevertheless, C. Daniel Johnson, MD, of the Mayo Clinic and Abraham H. Dachman of the University of Chicago suggest that CT colonography has “enormous potential for colorectal screening,” and they suspect that the study will be “the next colon screening examination.”11 A research team from the Warren Grant Magnuson Clinical Center at the US National Institutes of Health and the Department of Radiology at the Mayo Clinic is working on a shape-based algorithm that will identify polyps automatically, which would be a boon in screening examinations.12
Another method in development is intraluminal ultrasonography with miniaturized probes. M. H?nerbein, MD, and associates at the Robert R?ssie Hospital of Humboldt University in Berlin obtained what they described as life-like views of cancers of the esophagus and colorectum. The images improved the assessment of local invasion and the relations of the tumor to adjacent structures.13
Monitoring of Cancer Therapy
In an effort to reduce toxicity and costs, some medical oncologists are using volumetric imaging to determine whether chemotherapy is effective. For example, K.H. Shin, MD, and colleagues in the Department of Orthopaedic Surgery at Yonsei University College of Medicine, Seoul, Korea, used 3-D MRI to calculate changes in the volume of osteosarcomas in 41 patients receiving neoadjuvant chemotherapy.14 There was a significant correlation between changes in volume and histopathologic findings in the surgical specimen. After three cycles of chemotherapy, the 3-D study could identify a good response with a sensitivity of 85%, a specificity of 76%, and a positive predictive value of 88%.
The bronchial tree is another site where virtual endoscopy is being used. “Insurance companies do not specifically pay for virtual bronchoscopy, so it is difficult to know how much of it is being done for other than research purposes,” reports Atul Mehta, MD, head of bronchology, Cleveland Clinic. He expects it to become common, especially as the resolution improves.
“Virtual bronchoscopy will improve the yield of present methods such as flexible bronchoscopy. It also will eliminate much of the bronchoscopy that we do now on patients with nonspecific symptoms such as cough and for follow-up of patients who have had endobronchial treatment such as brachytherapy or laser photoresection for lung cancer. Once you have your CT scanner, the technology is very easy.”
Mehta is enthusiastic about another application of virtual endoscopy that does not require high resolution.
“Virtual bronchoscopy simulators produce images that look exactly like the endobronchial tree. These simulators are the wave of the future for physician training, because you would prefer not to have them work first on patients. Instead, like airline pilots, they will acquire their skills on a simulator before doing it in the real world.”
This list of applications is only a sampling of what can be done with 3-D imaging. For example, real-time 3-D imaging of the fetus is being used for evaluation of anomalies (one can imagine a demand by parents for these images!). In the chest, 3-D ultrasound imaging reduces the cardiac artifacts and enables radiologists to quantify the extent of emphysema. Also, echocardiography is used routinely at some medical centers. Many other techniques will be developed to exploit the resolution and convenience of 3-D images. No special glasses are needed.?
References
1. Nelemans PJ, Leiner T, Vet HCW de, Engelshoven JMA van. Peripheral arterial disease: meta-analysis of the diagnostic performance of MR angiography. Radiology. 2000;217:105-114.
2. Sardanelli F, Molinari G, Zandrino F, Balbi M. Three-dimensional navigator-echo MR coronary angiography in detecting stenoses of the major epicardial vessels, with conventional coronary angiography as the standard of reference. Radiology. 2000;214:808-814.
3. Scheidler J, Hricak H, Vigneron DB, et al. Prostate cancer: localization with three-dimensional proton MR spectroscopic imaging-clinicopathologic study. Radiology. 1999;213:473-480.
4. Yu KK, Scheidler J, Hricak H, et al. Prostate cancer: prediction of extracapsular extension with endorectal MR imaging and three-dimensional proton MR spectroscopic imaging. Radiology. 1999;213:481-488.
5. Coll DM, Uzzo RG, Herts BR, Davros WJ, Wirth L, Novick AC. 3-Dimensional volume rendered computerized tomography for preoperative evaluation and intraoperative treatment of patients undergoing nephron sparing surgery. J Urol. 1999;161:1097-1102.
6. Amano G, Ohuchi N, Ishibashi T, Ishida T, Amari M, Satomi S. Correlation of three-dimensional magnetic resonance imaging with precise histopathological map concerning carcinoma extension in the breast. Breast Cancer Res Treat. 2000;60:43-45.
7. Franca C, Levin-Plotnik D, Sehgal V, Chen GT, Ramsey RG. Use of three-dimensional spiral computed tomography imaging for staging and surgical planning of head and neck cancer. J Digit Imaging. 2000;13(suppl 1):24-32.
8. Yankelevitz DF, Reeves AP, Kostis WJ, Zhao B, Henschke CI. Small pulmonary nodules: volumetrically determined growth rates based on CT evaluation. Radiology 2000;271:251-256.
9. McFarland EG, Brink JA, Pilgram TK, et al. Spiral CT colography: reader agreement and diagnostic performance with two- and three-dimensional image-display techniques. Radiology. 2001;218:375-383.
10. Hara AK, Johnson CD, MacCarty RL, Welch TJ, McCollough CH, Harmsen WS. CT colography: single- versus multi-detector row imaging. Radiology. 2001;219:461-465.
11. Johnson CD, Dachman AH. CT colography: the next colon screening examination? Radiology. 2001;216:331-341.
12. Summers RM, Johnson CD, Pusanik LM, Malley JD, Youssef AM, Reed JE. Automated polyp detection at CT colography: feasibility assessment in a human population. Radiology. 2001;219:51-59.
13. H?nerbein M, Ghadimi BM, Gretschel S, Schlag PM. Three-dimensional endoluminal ultrasound: a new method for the evaluation of gastrointestinal tumors. Abdom Imaging. 1999;24:445-448.
14. Shin KH, Moon SH, Suh JS, Yang WI. Tumor volume change as a predictor of chemotherapeutic response in osteosarcoma. Clin Orthop. 2000;376:200-208.
Judith Gunn Bronson, MS, is a contributing writer for Decisions in Axis Imaging News.
