Abdominal Imaging Division

For decades after the discovery of the x-ray by Wilhelm Roentgen in 1895, imaging of the abdomen was primarily obtained using plain film radiography, with just the contrast of the body’s natural contents to produce an image. Gradually, however, the natural contrast of the body was supplemented by exogenous introduction of contrast agents that altered the otherwise expected attenuation of the radiographic beam. Barium-based agents were swallowed, or administered per rectum, to visualize the gastrointestinal tract, including esophagrams, upper GI series, and barium enemas. Specifically designed iodinated agents were concentrated in the bile to visualize the gallbladder; these could be administered orally (for oral cholecystography) or intravenously (for intravenous cholangiography, allowing visualization of the main biliary ducts in addition to the gallbladder).

Iodinated contrast agents were injected intravenously for intravenous urography, in which they were subsequently concentrated and excreted by the kidneys to view the urinary tract. These agents could be injected directly into the lumens of organs, such as the urinary bladder, to visualize these organs. Iodinated agents could also be injected directly into arteries and veins for arteriography and venography to directly visualize these vessels. Arteriography also enhanced the visualization of the organs supplied by the injected artery, and tumors within organs such as the liver could be identified, although these examinations produced two-dimensional images of three-dimensional objects and were not without some mild to moderate risk of complication.

For decades, these techniques composed radiological imaging of the abdomen. In the 1970s and beyond, medical imaging was revolutionized by the invention of devices that could image in three-dimensions: ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI). Although true three-dimensional images can be produced by these devices, the workhouse images are “slices” through the body, as if the patient were sliced like a loaf of bread. The radiologist then creates a mental image of the three-dimensional body by “stacking” the slices in his or her mind. Most importantly, a lesion can be located precisely in space by its appearance at a specific location within a slice of known position. The history of radiology in the latter quarter of the twentieth century and beyond became the story of cross-sectional imaging techniques.

John R. Thornbury took charge of urologic radiology following the departure of Anthony F. Lalli in 1970. Lalli had had great influence on the technique of intravenous urography based on his principles of the “tailored urogram,” later elucidated in a book of the same name. Thornbury would continue leadership in uroradiology and collaborate with Dennis Fryback on early work on the use of decision analysis in imaging and patient management.

In the early 1970s, ultrasonic high-frequency sound waves became widely used in medicine for imaging. Early ultrasound imaging was crude, but the technology improved dramatically, albeit slowly. For abdominal imaging, the ultrasonic transducer, which emitted the high-frequency sound waves and received the echoes back that created the image, would be placed on the skin of the patient; the distance from the skin to the organs of interest, as well as intervening bone and bowel gas that obscured the ultrasonic beam, limited the results. Nevertheless, in part due to the physics of the ultrasonic beam and in part because the technique imaged “slices” through the patient, ultrasonography was able to make diagnoses not previously possible. In 1972, the Radiology Department was granted jurisdiction over sonography, and Terry M. Silver took on the role of division director. Silver advanced the technique of neonatal cranial ultrasonography to study the infant brain and pursued intraoperative sonography; with the latter technique, radiologists were able to image organs when the abdomen was surgically opened and the organs were directly exposed to the ultrasound transducer without intervening tissues.

The advent of computed tomography in the 1970s revolutionized imaging in medicine. Relying on the well understood physics of x-rays supplemented with new computer techniques, CT scanners produced cross-sectional images through the human body. In addition to the improved spatial localization of lesions, CT scanners were able to resolve smaller differences in radiographic attenuation than had been possible with classic film radiography. For example, CT could separate the white and gray matter of the brain, or the liver from the kidney, based on only very small differences in radiographic attenuation, thus showing features never seen before without cutting open the patient for direct inspection. Initially, CT scanners could only image the human head, with the brain as the area of interest, and an EMI head unit was installed at University Hospital in 1975. By 1976, however, faster scanners and improved computing technology allowed imaging of the chest and abdomen.

The first whole body CT scanner in Ann Arbor, an EMI whole body scanner, was not installed in University Hospital, but instead at the nearby Veterans Administration Hospital in 1977, with which the University Medical Center had faculty staffing and teaching agreements. Patients who required body CT scans would be transferred by ambulance from University Hospital to the VA to be imaged under an agreement between the two hospitals. Competition for the available scan time was tight. Often a U-M radiologist, Lawrence Kuhns, would interpret the body CT studies off hours. In 1980, an early General Electric whole body CT scanner was installed at the University Hospital; with this, it was no longer necessary to send UMHS patients to the VA for CT scans.

Adams appointed Gary M. Glazer, a brilliant radiologist and outstanding investigator and mentor, as head of the Body CT Division in 1981. Glazer oversaw rapid expansion of CT with more scanners and more clinical applications as the technology improved and scans became faster and gained higher resolution. (Initially, CT capacity was limited mostly by the institution’s hesitancy to invest in such expensive equipment and was also limited by building constraints in the 1920s-era hospital.) Glazer recruited outstanding fellows, including Isaac R. Francis, who was a fellow from July 1981 to June 1982, and would go on to become a long-time faculty member and one of the most prolific authors of scientific articles in the Department. Francis became the Department’s associate chair for research in 2000.

L. Paul Sonda, a U-M-trained urologist who would become a U-M radiology resident and Radiology faculty member, became closely involved with the Abdomen Division early in his urology career. In the early 1980s, Sonda created the state’s first program in percutaneous renal stone extraction using fluoroscopic guidance in what now would be called minimally invasive surgery. Sonda and abdominal radiologist Marco Amendola had traveled to the University of Minnesota to learn the technique. Later, Sonda partnered with abdominal radiologist Terry Brady and, after Brady’s departure, Edward Woolsey. The procedure was begun by the radiologist accessing the stone-containing kidney in a fluoroscopic room in radiology and dilating a tract large enough to admit an endoscope. The procedure would then be turned over to Sonda, who would use the endoscope to either extract the stone whole (if small enough) or break the stone into smaller pieces that could be extracted or passed. This replaced open renal surgery in an operating room, and in addition to quicker patient recovery and avoidance of general anesthesia (almost all percutaneous cases were performed under conscious sedation), the procedure being done in a radiology room offloaded some cases from the operating rooms at a time when operating room time was limited. Eventually a general fluoroscopic room was replaced with a special procedure room to facilitate the cases. The development of extracorporeal shockwave lithotripsy that began in the late 1980s reduced the need for the percutaneous procedure, but it continues to be performed for patients in whom external shockwave lithotripsy is not appropriate. In the late 1990s, the radiology portion of the procedure was turned over to the Vascular/Interventional Division.

Jonathan M. Rubin joined the Department in 1984 and became co-director of the ultrasound division in 1987. Rubin began expanding research into ultrasound imaging, and would eventually become a National Institutes of Health (NIH)-funded principal investigator and receive the Society of Radiologists in Ultrasound Lawrence Mack Lifetime Achievement Award.

After Martel was appointed chair in October 1982, he continued the efforts begun by Adams to establish subspecialty divisions according to organ or body system and modality. Certain imaging modalities, such as CT and ultrasound, were considered sufficiently new or complex to be regarded as subspecialties in their own right. The general theme of abdomen imaging was spread across the then-divisions of Gastrointestinal Imaging (primarily plain radiography and barium studies), Genitourinary Imaging (plain radiography, intravenous urography, fluoroscopic studies of the genitourinary tract, and interventional GU procedures), Ultrasonography, and Body CT; the latter two divisions also imaged the chest and extremities. Although initially well accepted, this later led to conflict. As the Department expanded, newer recruits who had trained in all modalities did not want to be limited to just one modality; crossing over among multiple Divisions was accepted for some faculty and resisted for others.

Martel combined the Divisions of Gastrointestinal and Genitourinary Imaging with the division of Body CT to form a newly created Abdomen Division in July 1987, appointing James H. Ellis as the first Abdomen Division director. Ellis had been recruited in July 1984 by Martel to be the chief of the Radiology Service at the Ann Arbor VA Medical Center and was proficient in all the studies and procedures of the new division, including the percutaneous nephrostomies and stone extractions that Ellis inherited from Woolsey on the latter’s departure. With the creation of the new Abdomen Division, Glazer became the first division director for the newly designated MR Imaging Division.

In the late 1980s, Joel F. Platt, then a young faculty member, led a team of Rubin, Ellis, and Michael A. DiPietro (of the Pediatric Radiology Section) that began investigating new uses of the ultrasound renal resistive index, a measure of the resistance within the kidney to arterial blood inflow. Although the technique had been in use to evaluate transplant kidneys, this group pioneered the application of renal resistive index within native kidneys to distinguish whether kidneys with dilated collecting systems had physiologically significant obstruction to urine outflow or were dilated for other reasons. Over the years, additional applications of the renal resistive index were investigated, including whether variations in the index might distinguish various medical renal diseases without collecting system dilatation.

In 1989, Melvyn T. Korobkin, an outstanding academic radiologist in body CT who had gone into private practice, was recruited back into academia by Martel and given a co-director role in the Abdomen Division. He became the sole director in July 1992 when Ellis was appointed clinical director for the Department by Dunnick, then the incoming chair.

When Gary Glazer left the Department to become chair of the Department of Radiology at Stanford University in 1989, it left a large hole in the Department’s MR imaging leadership. The Department also had lost its most powerful advocate for abdominal MR imaging. The MR division continued as a quasi-independent division under the leadership of neuroradiologists, given that the majority of applications of MR were in neuroimaging. Additionally, Michigan’s certificate-of-need laws restricted the number of MR units that could be installed, leading to a chronic shortage of patient slots on the MR scanners and long wait times. Preference was often given to patients with neurologic problems because the advantages of MR over CT were clearer for neuroimaging, and some neuroimaging could only be done with MR. Later, applications in the abdomen for MR would become more widespread in radiology.

Rubin and coworkers published a landmark article in 1994 (Rubin JM, Bude RO, Carson PL, Bree RL, Adler RS. “Power Doppler US: A potentially useful alternative to mean frequency-based color Doppler US.” Radiology 1994; 190:853-856) that demonstrated the use of Power Doppler in abdominal imaging. Color Doppler ultrasound was a widely used ultrasound technique to assess vascular flow, both to visualize vessels directly and to demonstrate blood flow within organ tissue. Power Doppler had been used by cardiologists to evaluate vessel stenoses. The Michigan radiologists applied the technique in the abdomen and showed advantages over standard color Doppler, including extended dynamic range that improved visualization of blood flow in low-flow states. Within a few years, power Doppler became standard on ultrasound units produced for the radiology market.

By the early 1990s, chest CT studies were absorbed into the Chest Division and musculoskeletal CT studies were absorbed by the Musculoskeletal Division. In the late 1990s, the interventional genitourinary procedures were transferred to the vascular/interventional division, which put Michigan more in line with other academic radiology departments. The Abdomen and Ultrasound Divisions continued to perform procedures guided by CT and ultrasound, respectively (primarily tissue biopsies, fluid aspirations, and abscess drainages).

In 1992, Dunnick recruited Richard H. Cohan to join the Abdomen Division. Cohan was interested in the use, effects, and complications of contrast agents in imaging, and began a long collaboration with Ellis that produced many publications on the topic. Cohan and Ellis advanced the training of residents and fellows in contrast issues and the management of contrast reactions, and started an annual course in 1995 that employed simulation of contrast reactions as part of the instruction. The course evolved over many years, with Matthew S. Davenport and Jonathan R. Dillman (both former residents who became faculty) joining the training team beginning in 2010. The simulations were taken to a new level in 2012 when the course took advantage of the Health System’s Simulation Center to use high-fidelity patient manikins as practice “patients” experiencing simulated contrast reactions for residents to treat.

Martin R. Prince was recruited from Harvard and Massachusetts General Hospital in 1993 to bring MR angiography to Michigan. Prince’s collaboration with Department MR physicist Thomas Chenevert and Thomas K. Foo, a research scientist at General Electric, led to the development of a bolus triggering gadolinium-contrast-enhanced MRA technique which GE commercialized as “MR Smartprep.” Together with Michigan radiology resident Jeffrey H. Maki, he described the artifacts on 3D gadolinium-contrast-enhanced MRA that were related to contrast concentration variations. Working with Michigan MRI fellow James F. M. Meaney, Prince developed the concept of bolus chase MR Angiography which became widely used for performing peripheral MR Angiography.

By 1995, Korobkin had begun to investigate whether CT scans could provide true tissue diagnosis in one specific situation. The adrenal glands were common sites of metastasis from cancer, but they were also a common site of benign adrenal adenomas. The treatment of many cancers was quite different depending on whether the cancer had metastasized or not, but many cancer patients had an adrenal adenoma posing as a metastasis. Both conditions presented as a nodule in an adrenal gland, and earlier attempts to differentiate them by their morphologic appearance at CT scanning had not produced sufficiently accurate separation. Thus differentiating a benign adrenal adenoma from an adrenal metastasis typically required image-guided biopsy that was difficult and presented some risk to the patient, or long term follow-up that might delay appropriate treatment. Korobkin expanded on and refined earlier work showing that CT scanning might be able to demonstrate that some of these nodules represented a benign adrenal lesion native to the adrenal gland based not on lesion morphology but on lesion radiographic density, potentially obviating the need for biopsy. Korobkin knew that most of the benign adrenal adenomas had cells containing intracellular lipid (“lipid-rich adenomas”), and lipid was lower in density on CT scans than cancer cells that did not contain lipid. But did the adenoma cells contain enough lipid to lower the density sufficiently to make the diagnosis with reasonable certainty? Korobkin et al. compared the CT scans of patients with adrenal adenomas and patients with nonadenomas (primarily metastases), and was able to show that threshold values for CT attenuation could be identified that were sufficiently specific for adrenal adenomas that the diagnosis could be made and biopsy foregone based on noncontrast CT scanning alone (Korobkin M, Brodeur FJ, Yutzy GG, Francis IR, Quint LE, Dunnick NR, Kazerooni EA. “Differentiation of adrenal adenomas from nonadenomas using CT attenuation values.” American Journal of Roentgenology 1996; 166:531-536). The determination of intracellular lipid in these adrenal masses could also be made using specific MR techniques (Korobkin M, Giordano TJ, Brodeur FJ, Francis IR, Siegelman ES, Quint LE, Dunnick NR, Heiken JP, Wang HH. “Adrenal adenomas: relationship between histologic lipid and CT and MR findings.” Radiology. 1996 Sep;200(3):743-747). Although this research enabled lipid-rich adenomas to be diagnosed noninvasively, lipid-poor adenomas still could not be differentiated from metastases. Later, Korobkin and co-workers were able to show that many of the lipid-poor adenomas could be distinguished from metastases by using a different approach. Observing that intravenously administered iodinated contrast medium washes out of adrenal masses at different rates over time, Korobkin and his colleagues developed a quantitative method that accurately and consistently distinguished adenomas from metastases, even those lipid-poor adenomas that did not meet the noncontrast-CT criteria for being an adenoma (Korobkin M, Brodeur FJ, Francis IR, Quint LE, Dunnick NR, Londy F. “CT time-attenuation washout curves of adrenal adenomas and nonadenomas.” American Journal of Roentgenology 1998; 170:747-752). Together, these noncontrast and contrast-enhanced CT techniques have become the standard of care for evaluating adrenal nodules.

As cross-sectional imaging became more and more important in clinical medicine, older traditional abdominal imaging techniques became less favored. Fortunately, the Department was able to maintain expertise in traditional genitourinary imaging by the long tenure of a large number of faculty with this interest: Cohan, Dunnick, Ellis, and Platt. Newer faculty such as Davenport became expert as well. In gastrointestinal imaging, the Department recruited graduating resident L. Paul Sonda in 1997 to take a leadership role; the Department funded a six-month fellowship at outside institutions (Hospital of the University of Pennsylvania, Beaumont Hospital) to sharpen his skills in traditional gastrointestinal imaging. Sonda had been a urologist at the University of Michigan since 1977; he began his residency in radiology at Michigan in July 1993.

Like Korobkin, Robert L. Bree was a radiologist who returned from private practice to an academic career at Michigan and thrived. An expert in ultrasound imaging and imaging utilization, Bree also was a talented leader and administrator. He recognized that image-guided biopsies, fluid aspirations, and abscess drainages were split between the ultrasound and CT sections of the Abdominal Imaging division, leading to fragmented service and difficulty with scheduling (sometimes clinical services would be bounced back and forth between ultrasound and CT as each side thought the other was better suited to perform the procedure). Furthermore, the performance of these interventional procedures of uncertain length interrupted the efficiency of the interpretation of regular diagnostic imaging studies. Bree not only saw that a specialized image-guided cross-sectional procedure service would improve patient care, but he also recognized that, over time, caseload would increase as obtaining procedures became easier and new applications could be introduced. He also shifted many of the procedures from CT to ultrasound, freeing up slots in the Michigan certificate-of-need-limited CT scanner resources. In yet another insight, Bree appreciated that mid-level providers could deliver service continuity and could perform procedures under guidance that would increase case numbers beyond what could be provided by radiology faculty alone. This led to the recruitment of the Department’s first physician assistant (PA), Ellen Higgins, MS, PA-C in 1998. Once the contributions of PAs to patient care, service efficiency, and quality improvement were recognized, the number of PAs in the Department expanded rapidly, and by 2014 there were 11 in the Department.

Prior to Bree’s recruitment to become chair of the Department of Radiology at the University of Missouri (Columbia), he had acquired the equipment and had introduced radiofrequency ablation (RFA) to the cross-sectional interventional service. This was a minimally invasive technique that used percutaneous probes, placed under imaging guidance, to heat and destroy unwanted tissue (primarily cancers). With Bree’s departure, the Department recruited Hanh V. Nghiem to continue and advance tumor treatment via image-guided percutaneous techniques. These cases rapidly expanded in number and complexity, and additional faculty were added. Major applications were the biopsy and ablation of renal tumors, precluding the need for resection, and ablation of liver cancers as either primary therapy or as a method to keep patients alive and well enough for liver transplantation.

In 1999, the Department recruited Hero K. Hussain from the Mallinckrodt Institute of Radiology at Washington University in St. Louis to take charge of abdominal MR imaging. Although abdominal MR imaging would remain administratively a part of the abdomen division, Hussain rapidly became the “go-to” person in abdominal MR, setting protocols and directing workflow. She took on the responsibilities of a division director without the title. Hussain quickly focused abdomen MR on liver imaging, a field where MR was clearly superior to CT for lesion characterization, although other applications, including MR urography and later MR enterography, were not ignored. Hussain worked with clinical and transplant liver specialists to develop imaging criteria for the non-invasive diagnosis of hepatocellular carcinoma, a primary liver malignancy. Many patients needing liver transplantation have cirrhosis and are at risk for hepatocellular carcinoma, and many of these patients have liver lesions that enhance on CT during the arterial phase of contrast administration, but not all of these lesions are cancers.

With the appointment of Platt as director of the Abdomen Division in July 2002, the Ultrasound Division was folded into the Abdomen Division.

In 2005, Hussain published a seminal paper describing highly specific MR imaging features for the diagnosis of hepatocellular carcinoma in the cirrhotic liver (Marrero JA, Hussain HK, Nghiem HV, Umar R, Fontana RJ, Lok AS. “Improving the prediction of hepatocellular carcinoma in cirrhotic patients with an arterially-enhancing liver mass.” Liver Transpl 2005; 11:281-289). These criteria were subsequently adopted by all major American and European professional societies dealing with liver disease and the Organ Procurement and Transplantation Network, thus eliminating the need for invasive biopsy confirmation of hepatic malignancy status in most cases.

Under the divisional leadership of Platt and with the contributions of various individuals, new techniques were brought to the patients of the University of Michigan Health System. With the onset of helical and then multi-slice CT, scans could be obtained faster, enabling better visualization of the intravenously administered contrast material while it was still briefly in the arterial system. For example, CT was able to improve the assessment of patients who were being evaluated as living donors of a kidney to another patient with kidney failure. Previously, in order to decide if these donors were candidates for donation and which kidney to take if they could be a donor, potential donors needed a catheter angiogram and an intravenous urogram to assess both the vessels supplying the native kidneys and the integrity and anatomy of the renal parenchyma and urinary collecting systems. Using CT, the evaluation could be performed in one examination and require only a single intravenous injection of iodinated contrast material (Platt JF, Ellis JH, Korobkin M, Reige K. “Helical CT evaluation of potential kidney donors: findings in 154 subjects.” American Journal of Roentgenology. 1997 Nov;169(5):1325-1330). CT also replaced the intravenous urogram in the evaluation of patients with urinary tract stone disease, and later, with the development of CT urography, in part advanced and refined at Michigan by Cohan and Elaine M. Caoili and others (Caoili EM, Cohan RH, Korobkin M, Platt JF, Francis IR, Faerber GJ, Montie JE, Ellis JH. “Urinary tract abnormalities: initial experience with multi-detector row CT urography.” Radiology. 2002 Feb;222(2):353-360, Caoili EM, Inampudi P, Cohan RH, Ellis JH. “Optimization of multi-detector row CT urography: effect of compression, saline administration, and prolongation of acquisition delay.” Radiology. 2005; 235(1):116-123), essentially completely replaced the intravenous urogram in the evaluation of the urinary tract for all diseases where imaging played a role. Still later, MR urography was added to the imaging armamentarium.

The small bowel was difficult to evaluate, as it was beyond the reach of the gastroenterologist’s endoscope, and the barium- and radiography-based small bowel follow-through primarily imaged only the mucosal surface of the small bowel, and not the small bowel wall where much pathology was located. CT and MR enterography techniques were able to image both the lumen and the wall of the small bowel, adding greatly to the understanding of small bowel diseases such as Crohn disease, with young faculty members Mahmoud M. Al-Hawary and Ravi K. Kaza taking leadership roles.

Abdominal radiologist Ruth C. Carlos initiated comparative effectiveness research and technology assessment in diagnostic testing, as well as research into health care reform and imaging utilization. In 2010, she was named assistant chair for clinical research in the Department. She became a member of the Board of Review for the General Electric-Association of University Radiologists Radiology Research Academic Fellowships, a renowned national training program for health services research in radiology, and in 2012, she became the chair of the Board of Review.

New techniques and applications in abdominal imaging continue to be developed and refined in the Department of Radiology at U-M, and will likely play an increasing role in abdominal imaging in the future. Vascular imaging continues to expand, and pre-operative planning for tissue flaps is rapidly converting from catheter angiography to noninvasive CT angiography under the leadership of Peter S. Liu. New CT reconstruction techniques that are able to reduce radiation dose to the patient are under evaluation by faculty such as Kaza and Katherine E. Maturen. Dual-energy CT may provide additional ways to improve diagnostic quality and/or reduce radiation dose under the direction of Kaza and Platt. A rejuvenation of MR imaging of the prostate gland began under Davenport’s guidance. MR staging of rectal cancer is being championed by Al-Hawary.

In 2012, Davenport, a former chief resident, was recruited back to the Department after fellowship training from Duke University. Davenport had great interest in the use of radiographic contrast media and published many articles on this topic. Davenport was the lead author on several defining articles from the Michigan contrast interest group that applied a new statistical technique, propensity score matching, to the long-standing question of the renal toxicity of iodinated contrast agents (the agents given intravenously for CT scans, and other uses). Propensity score matching allowed a retrospective analysis to more closely emulate a randomized controlled trial by reducing the selection bias that plagued other retrospective reviews of renal deterioration in patients who did and did not receive intravenous iodinated contrast agents during their illnesses. One of the publications (Davenport MS, Khalatbari S, Cohan RH, Dillman JR, Myles JD, Ellis JH. “Contrast material-induced nephrotoxicity and intravenous low-osmolality iodinated contrast material: risk stratification by using estimated glomerular filtration rate.” Radiology 2013; 268:719-728) was co-awarded the 2013 Alexander R. Margulis Award for Scientific Excellence as the best original scientific article of the year published in Radiology, the premier radiological journal.

Basic Radiological Sciences Division

Following the 1981 recruitment of Paul L. Carson to lead the Basic Radiological Sciences division, the division has seen substantial growth and increase in prominence. Beginning with expertise in ultrasound physics and imaging, the division would subsequently expand to bioengineering, biochemistry, medicinal chemistry, biostatistics, microwave and optical imaging, magnetic resonance, and magnetic resonance spectroscopy.

In ultrasound, Brian Fowlkes, Carson and colleagues including Oliver Kripfgans and Mario Fabiilli used ultrasound to create microbubbles in blood vessels and the urinary bladder to enhance sonographic diagnosis and therapy. Later, by using ultrasound to vaporize injected perfluorocarbon microdroplets and then agitate the resulting microbubbles in the location of specific cells, their group targeted these cells for imaging and for the delivery of drugs, hormones, or acoustic energy density if administered to the entire body (Kripfgans OD, Fowlkes JB, Miller DL, Eldevik OP, Carson PL. “Acoustic droplet vaporization for therapeutic and diagnostic applications.” Ultrasound Med. Biol. 2000; 26:1177-1189). This has helped them deliver cancer therapies to specific cells and injectable dyes to study drug release, effects, and retention, among other applications.

The ultrasound group has also made important progress toward the use of 3D automated ultrasound in conjunction with breast tomosynthesis to improve breast cancer detection and assessment (Padilla F, Roubidoux M, Paramagul C, Sinha S, Goodsitt M, Le Carpentier G, Chan H-P, Hadjiiski L, Fowlkes B, Joe A, Klein AL, Nees A, Noroozian M, Patterson SK, Pinsky RW, Hooi FM, Carson PL. “Breast mass characterization using 3D automated ultrasound as an adjunct to digital breast tomosynthesis: A pilot study.” J Ult Med 2013; 32:93-104). Fowlkes, working with Charles Cain, professor and founding chair of the Department of Biomedical Engineering, in collaboration with biomedical engineering and urology colleagues, have developed a noninvasive surgical technique called “histotripsy” (Parsons JE, Cain CA, Fowlkes JB. “Spatial variability in acoustic backscatter as an indicator of tissue homogenate production in pulsed cavitational ultrasound therapy.” IEEE Trans. Ultrason. Ferroelectr. Freq. Control 2007; 54:576-590). The procedure uses highly focused ultrasound waves as an “acoustic scalpel,” dissolving lesions without breaking the patient’s skin or damaging surrounding tissue. Douglas Miller published seminal studies in the bioeffects of ultrasound, detailing the conditions under which the most vulnerable tissues might be damaged, particularly with ultrasound contrast agents (Williams AR, Wiggins RC, Wharram BL, Goyal M, Dou C, Johnson KJ, Miller DL. “Nephron injury induced by diagnostic ultrasound imaging at high mechanical index with gas body contrast agent.” Ultrasound Med. Biol. 2007; 33:1336-1344).

With the recruitment of Xueding Wang, a new group in optical imaging was created in close association with ultrasound. His work is on the cutting edge of photoacoustic imaging, which processes ultrasound waves generated when laser pulses are applied to biological samples, expanding the optically absorbing tissues. He was the first to direct the technique to imaging of inflammatory arthritis (Chamberland DL, Wang X, Roessler BJ. “Photoacoustic tomography of carrageenan-induced arthritis in a rat model.” J. Biomed. Optics 2008; 13:011005) and has attached gold nanoparticles to antirheumatic drugs to visualize their uptake by inflamed joint tissues. Because photoacoustics is capable of quantifying physiological biomarkers of cancer, like oxygenation and pH levels, Wang and research investigator Guan Xu are using photoacoustics to study prostate cancer. Wang, Carson and Zhixing Xie built a breast 3D photoacoustic imaging system with the deepest reported breast imaging of 5 cm.

Heang-Ping Chan joined the division in 1989, rapidly establishing her leading work in computer-assisted diagnosis of breast cancer; she would later become the Paul L. Carson Collegiate Professor of Radiology. Recently she and her colleagues have developed what is probably the most advanced CAD system for detection of microcalcifications in digital breast tomosynthesis (Sahiner B, Chan H-P, Hadjiiski L, Helvie M, Wei J, Zhou C, Lu Y. Computer-aided detection of clustered microcalcifications in digital breast tomosynthesis: a 3D approach. Med. Phys. 2012; 39:28-39). Her lab not only develops computerized methods of detection but uses machine learning methods to build decision support systems for risk prediction, diagnosis, and treatment response monitoring. In recent years they have been applying these methods to other diseases and organ systems besides the breast.

A promising CT hepatocyte-specific contrast agent was developed and advanced to a commercial small animal imaging agent by Jamie Weichert. Meyer’s laboratory developed the first practical method for warped, 3D registration and fusion of multimodality data sets (CT, MRI, PET, US) to improve diagnosis and treatment in many conditions (Meyer CR, Boes JL, Kim B, Bland PH, Zasadny KR, Kison PV, Koral K, Frey KA, Wahl RL. “Demonstration of accuracy and clinical versatility of mutual information for automatic multimodality image fusion using affine and thin-plate spline warped geometric deformations.” Medical Image Analysis 1996/7; 1:195-206).

Brian D. Ross, the Roger A. Berg Radiology Research Professor, and Thomas L. Chenevert, the BRS Collegiate Professor of Radiology, collaborated on research using MR imaging to track the diffusion of water through tumor cells as an indicator of chemotherapy response (Chenevert TL, McKeever PE, Ross BD. “Monitoring early response of experimental brain tumors to therapy using diffusion magnetic resonance imaging.” Clin Cancer Res. 1997; 3:1457-1466). Because cancer cells slow the diffusion of water, an increase in diffusion correlates to a positive response to therapy. P31 spectroscopy was developed to evaluate muscle physiology, and magnetization transfer imaging was used by Scott Swanson to clarify factors that determine MR contrast (Henkelman RM, Huang X, Xiang QS, Stanisz GJ, Swanson SD, Bronskill MJ. “Quantitative interpretation of magnetization transfer.” Magnetic Resonance in Medicine 1993; 29:759-766). Leading edge hyperpolarized xenon MRI research was performed by Swanson in cooperation with the Department of Physics.

Ross developed the molecular imaging center with colleague Alnawaz Rehemtulla, professor of radiation oncology. The large funding for this center, Ross’s program project grant, and regional small animal imaging facility complemented by the consistent ultrasound, CAD, image processing, and nuclear medicine grant support, has placed U-M Radiology and Radiation Oncology in the top tier in National Institutes of Health radiologic research support since 2002. Gary Luker and Kathy Luker, Department faculty and members of the Molecular Imaging Center, specialize in using bioluminescence imaging and intravital fluorescence microscopy (Luker KE and Luker GD. “Applications of bioluminescence imaging to antiviral research and therapy: Multiple luciferase enzymes and quantitation.” Antiviral Res. 2008; 78:179-187) to analyze the signaling pathways regulating tumor growth, metastasis, and drug resistance in cell and animal models of breast cancer. Their work has allowed the visualization of ligand-receptor binding and the activity of proteasomes (cellular organelles responsible for degrading proteins), both of which are targets for chemotherapy. The group made significant progress in analyzing the mechanisms of cancer progression and treatment.

Breast Imaging Division

In the 1970s, mammography was performed with either film screen techniques or xerography. Although each technique had advantages and advocates, the lower radiation dose enabled film screen systems to become the dominant technique. Magnification views proved useful in the detection and characterization of calcifications. A hooked needle was developed that could be inserted into a suspicious non-palpable lesion and left in place to direct the surgeon to the removal of the cancer.

With improvements in mammography technique, clinical trials were performed that demonstrated the value of mammography as a screening procedure. The Health Insurance Plan of Greater New York project was the first appropriately powered study and demonstrated a significant decrease in mortality among screened women 40 years of age or older. This study and a subsequent Breast Cancer Detection Demonstration Project led the American College of Radiology to develop guidelines for screening mammography in 1976.

U-M was an original participating site of the BCDDP in the 1970s which showed the feasibility of widespread breast cancer screening by mammography. Results of the BCDDP and additional randomized controlled trials from Europe led to national recommendations for routine screening mammography beginning in the mid-1980s.

The original breast imaging facility was located on West Washington Street in Ann Arbor, not at the Hospital, which created some logistical issues. Women needing wire-localizations of non-palpable breast masses traveled from West Washington Street where the localization was performed to the hospital on the other side of Ann Arbor for their surgery. By the mid-1980s, mammographic equipment was located at the main hospital.

Mammographic studies were interpreted by radiologists with additional training in breast imaging. Clinical mammography followed techniques used and developed in the BCDDP and all women initially had physical examination by nurses followed by mammography. Thermography, an initial component of the BCDDP, was discontinued. Test interpretation usually occurred while the patient waited, and additional studies were completed as needed. In the 1990s, screening mammography was offered; the patients had their mammograms performed but did not see nurses or physicians at that appointment, and the interpretations of the images were performed “off line” in batches; the screening methodology increased efficiency and allowed more patients to undergo mammography at lesser cost and resource utilization. In 1998, 35 percent of the mammograms were screening studies. The number of screening examinations relative to diagnostic studies continued to rise, and in 2003, more screening studies were done than diagnostic mammograms.

Breast radiologists not only identify suspicious areas in the breast on mammography, but also characterize the mammographic lesion as well as palpable lesions through additional studies such as ultrasound, fine needle aspiration, needle localization, and core biopsies. In the 1980s, U-M was one of the first sites in the U.S. to offer image-guided percutaneous biopsy based upon techniques brought to U-M by Ingvar Andersson of Sweden. This allowed a less invasive (than open surgery) diagnosis of image-detected lesions. These procedures began in earnest in the late 1980s and by the year 2000, over 2000 procedures were performed annually by breast radiology faculty.

In addition to clinical research, the division has focused upon computer-aided detection and diagnosis (CAD) of breast imaging findings under the leadership of Heang-Ping Chan, advanced ultrasound techniques under the leadership of Paul Carson, and breast MR under the leadership of Tom Chenevert. Many of the clinical research advances and research techniques developed are now in routine clinical use. Faculty helped develop the BIRADS lexicon and establish the scientific basis for its use, now the standard throughout the U.S. and much of the world. Image-guided needle biopsy has largely replaced surgical biopsy. Every mammogram and MR is now “double read” using computer vision technology in addition to the radiologist. Breast ultrasound has evolved from a rarely used tool, to an indispensable clinical tool for screening and diagnosis. Magnetic resonance imaging is highly sensitive in identifying breast lesions, including malignancies. Thus, MR has proven to be useful in screening, in the further evaluation of some breast abnormalities, and for directing breast biopsies. Non-contrast breast MR was shown to be insufficient for clinical use at U-M in the 1980s. Research studies of contrast-enhanced breast MR in the early 1990s at U-M paved the way for clinical breast MR exams to begin in 2005. MR has become a commonly used modality for specific high risk screening indications.

The Division of Breast Radiology had been directed by Barbara Threatt, who left the University to enter private practice in 1984. Visiting Professor Andersson led the division on an interim basis and trained his successor, Dorit D. Adler. Adler had completed her residency training at Michigan in 1984, and did fellowship training in mammography and MR in the 1984-1985 academic year. In July 1986, she was named director of breast imaging. Mark Helvie was appointed division director in 1992 when Adler reduced her appointment to part-time; Adler continued as associate division director until her retirement in 2009.

The value of screening mammography continues to be debated, especially with regard to the benefit of annual mammography in women less than 50 years of age. Since the preponderance of scientific studies demonstrates benefit, screening mammography continues to grow and is now offered at U-M imaging facilities in Briarwood, Brighton, Canton, East Ann Arbor and Livonia, in addition to the Cancer Center. In a project led by Marilyn Roubidoux, Michigan breast faculty also provide mammographic screening for Native Americans; mammograms are obtained in the northern Great Plains using a van equipped with mammographic equipment, and the images are sent to U-M for interpretation.

Cardiothoracic Imaging Division

Entering the 1980s, the Chest Radiology Division was a small operation concerned only with interpretation of chest radiographs. When computed tomography (CT) was introduced and developed in the 1980s, a Body CT Division was formed to interpret scans of the neck, chest, abdomen, pelvis, and bones, including pediatric body CT studies. However, a major focus of body CT research in the 1980s concerned chest CT, particularly staging lung cancer. Under Gary Glazer’s leadership, many important articles about the technique and accuracy of chest CT in lung cancer staging came from U-M.

For most of the 1980s, the Chest Division consisted of two or three faculty who did mostly (but not exclusively) chest work. Although originally hired at Michigan in 1982 as a member of the Body CT Division after fellowship training in body CT and ultrasound, Barry Gross became Chest Division director in 1983. Later in the 1980s, Gross helped to start a process of sub-specialization within CT to allow David Spizarny, fellowship-trained as a chest radiologist, to read chest CT without being a member of the Body CT division. By the early 1990s, chest CT separated from body CT. Initially, scans of the chest, abdomen, and pelvis were not split apart for reading; instead, if the disease process originated in the chest (for example, lung cancer or esophageal cancer) the study was read in chest CT; if disease was primarily abdominal (for example, colon cancer) the study was read in abdomen CT. After a number of years, it was decided to split almost all combined chest/abdomen/pelvis CT scans, with chest reading all chest CTs and abdomen reading all abdomen and pelvis CTs. As of 2014, the exceptions are aortic CTs (read by chest if ordered by cardiac surgery, cardiology, or prompted by chest symptoms, and read by abdomen if ordered by vascular surgery, general surgery, or prompted by abdominal symptoms) and lung and esophageal cancer cases (because they need to be prepared for presentation at the Multidisciplinary Thoracic Oncology Conference). Today, there are more CT patient encounters in cardiothoracic imaging than in any other division in the Department.

Philip Cascade became the division director in 1991. He brought a significant background in imaging of coronary arteries, even though this was almost exclusively the province of cardiologists throughout the United States (including at U-M). Cascade’s extensive knowledge of the subject and his outstanding interpersonal skills were very helpful over the years, as radiologists assumed a significant role in cardiovascular evaluation because of advances in CT technology. He was also a major driver in the creation of the American College of Radiology Appropriateness Criteria, leading the ACR Task Force that developed all 10 panels. Cascade developed a method of assessing the competence of radiologists by analyzing missed diagnoses. This method was used extensively in the Chest Division before dissemination throughout the U-M Department of Radiology. Ultimately, this competency assessment formed the basis for the American College of Radiology Radpeer™ program, used nationally to evaluate the performance of radiologists.

While members of the Chest Radiology division had sporadically trained fellows in chest radiology since the mid-1980s, the fellowship program began to be filled regularly under Cascade’s leadership, and it became a major source of faculty recruitment, particularly in the first decade of the 21st century. Cascade brought in and mentored his successor as division director, Ella Kazerooni, and handled the potentially difficult situation of having his predecessor, Barry Gross, return to the division (after a stint as chair of Henry Ford Hospital Radiology) with aplomb. He turned the reins over to Kazerooni in 2000, and remained a key member of the division until his retirement at the end of 2013.

Kazerooni oversaw a period of unprecedented growth in the division beginning in 2000. This decade saw major expansion in chest imaging, incorporating magnetic resonance imaging (MRI), cardiac CT, and coronary artery CT. Given these changes, the division was renamed Cardiothoracic Radiology. The new technology completely changed the workup of many clinical problems. Pulmonary emboli are now almost exclusively diagnosed by CT, with nuclear medicine and catheter angiography no longer playing major roles. CT scanning of the aorta with detailed aortic measurements is a major supporting component of the University’s large cardiology and cardiovascular surgery program, including trans-catheter aortic valve replacement. Coronary CT angiography plays a growing role in the evaluation of emergent chest pain, and cardiac MRI has grown to support many specialty programs in cardiology, such as the hypertrophic cardiomyopathy program and the adult congenital heart disease program. The Cardiothoracic Division also provides major support for expanded pediatric cardiac MRI services to accommodate increased collaboration with pediatric cardiology and cardiac surgery.

The growth of the division significantly relied on the hiring of fellows who had trained in our own fellowship program. By 2014 there were 14 full-time cardiothoracic radiologists covering University Hospital and the Ann Arbor VA, and 2 others splitting time with other divisions (one doing some pediatric radiology and one doing some body CT). At one time, all cardiothoracic radiologists did the same assignments, consisting of chest CT, biopsies, and plain chest radiographs. By 2014, a further degree of sub-specialization had been introduced, with the establishment of assignments covering cardiovascular magnetic resonance imaging, led by Gisela Mueller and Prachi Agarwal, cardiovascular CT, led by Smita Patel, and thoracic interventional procedures. Sub-specialization of thoracic interventional procedures under the direction of Baskaran Sundaram allowed the application of advanced interventional techniques such as radiofrequency ablation procedures. With growth in cardiac imaging, cardiology colleagues have been incorporated into our imaging program, beginning with Adam Dorfman in pediatric cardiac MRI.

Kazerooni did an outstanding job of mentoring junior faculty, helping multiple former fellows advance to senior academic ranks. Seven faculty members received an Association of University Radiologists General Electric Radiology Research Academic Fellowship (GERRAF) Award or a Radiological Society of North America or American Roentgen Ray Society Scholar Award as a result. Kazerooni encouraged junior faculty members to enhance their knowledge of research basics and statistics, with many division members obtaining master’s level training from U-M’s School of Public Health to enhance the quality (and quantity) of academic output. She fostered close collaboration of our physician faculty with several groups in the basic sciences division of the Department, which has led to numerous National Institutes of Health-funded grants with radiology principal investigators. This includes working with Heang-Ping Chan’s group on computer assisted diagnosis for the detection and characterization of lung nodules, pulmonary emboli, and coronary artery disease, Charles Meyer’s program in image analysis, and Brian Ross’s lab in the area of advanced chronic obstructive pulmonary disease imaging using the parametric response mapping technique. Kazerooni was also a key player in major research initiatives involving pulmonary medicine, thoracic surgery, and pathology on the subjects of interstitial and obstructive lung diseases. This program has been recognized as an area of tremendous expertise at U-M. She brought participation in major cardiothoracic imaging trials to the division, including Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) II and III trials evaluating CT and MRI for pulmonary embolus detection, the National Emphysema Treatment Trial for lung volume reduction surgery, and the National Lung Screening Trial for early lung cancer detection. The latter is the largest clinical trial conducted by the U-M Radiology Department to date, enrolling over 850 subjects. In addition to her work as the division director, since 2008 she has served as the Department of Radiology’s associate chair for clinical services.

Musculoskeletal Imaging Division

Musculoskeletal imaging has seen significant growth and improvement since the 1980s, largely due to technological advances. Similarly, the number of imaging studies and fellowship-trained musculoskeletal faculty have also dramatically increased. In 1999, Curtis Hayes was recruited as director of the Musculoskeletal Division, and in 2003 Jon Jacobson became division director. From 1999 to 2015, the number of musculoskeletal faculty increased from 3 to 11. In addition to musculoskeletal radiology faculty on site at University Hospital and the outpatient Taubman Center, by 2015 musculoskeletal radiologists were embedded in the MedSport Orthopaedic Clinic at Domino’s Farms and at the newly established Northville Health Center.

While radiography has been and is still considered an essential imaging method in evaluation of the skeletal system, it is MR imaging and ultrasound that has allowed detailed assessment of the soft tissues as well as other areas not adequately assessed with radiography. In addition, advances in CT as well as digital image transfer and reporting have had a significant impact on patient care.

The first clinical MR imaging machine was installed at U-M in 1983. At the time, this state-of-the-art equipment had a field strength of 0.35 Tesla. Since then, the strength of magnets commonly used in clinical practice has increased to 3 Tesla, which has markedly improved the quality of MR imaging. MR imaging of the joints, such as the knee and shoulder, became routine, followed by intra-articular administration of gadolinium-based contrast material to improve the evaluation of the joints for internal derangement; clinicians came to rely on the imaging information in their treatment decisions. By 2014, the number of MR imaging machines had increased, with 10 located throughout the University of Michigan Healthcare System. Two of these MR imaging machines are designed with a “true open” configuration, which is helpful to patients with claustrophobia. An additional dedicated 3 Tesla research MR imaging machine provides access to patients enrolled in research studies.

Improvements in technology have also directly affected ultrasound and its use in evaluation of the musculoskeletal system. While early machines were limited by the lower frequency transducers, current clinical ultrasound machines have transducers with frequencies greater than 15 MHz. This has enabled ultrasound to provide a detailed assessment of superficial structures, such as tendons and ligaments, with resolutions higher than MR imaging. Another advantage of ultrasound is the ability to dynamically assess anatomic structures, for example during extremity movement or muscle contraction. Musculoskeletal ultrasound has also grown as ultrasound units have been placed in or adjacent to clinics to provide quick and efficient patient access. In 1999, musculoskeletal ultrasound was performed one half day per week. By 2015, there were five dedicated musculoskeletal ultrasound rooms with full schedules each day. A strong clinical, educational, and research focus on musculoskeletal ultrasound resulted in U-M being recognized as a national leader in musculoskeletal ultrasound education and research. In addition to diagnostic ultrasound, there has been growth with ultrasound-guided procedures, such as joint injections, as needle placement is most accurate using ultrasound guidance. Research in ultrasound at Michigan has also had a dramatic impact on clinical imaging. For example, Jonathan Rubin and his basic science team developed power Doppler imaging in the 1990s, and this technique became part of every clinically-available ultrasound unit. Former Division Director Ronald Adler, former faculty member Marnix Van Holsbeeck and Rubin were early leaders in ultrasound applications in musculoskeletal imaging. Rubin’s research team continued to make significant contributions to ultrasound research.

Computed tomography is commonly used in the musculoskeletal system to evaluate for fracture and other osseous abnormalities. Current CT scanners have a multichannel design, which significantly reduces scanning times. For example, it is now possible to scan an entire joint in seconds rather than minutes, and the thinner imaging slices have enabled images to be reformatted in any imaging plane with high resolution. Newer technology has also allowed significant improvements in reducing artifact, which is important when imaging in the presence of metal, such as a hip replacement. New technology has also allowed imaging to be completed with a goal of reducing radiation to a minimum, which is important to younger patients, especially children. CT is commonly used to guide percutaneous biopsy of soft tissue and bone tumors to characterize tumors prior to treatment.

Lastly, technological advances in digital transfer of images and reporting have improved patient care. Images acquired at any site in the medical system can now be viewed almost instantaneously at any computer or imaging workstation, by both the radiologist and clinician. Speech recognition dictation systems allow the radiologist to quickly and efficiently produce a finalized report that immediately is uploaded into the patient medical record. This combination of digital image transfer and speech recognition dictation has reduced the time from imaging to reporting to a matter of minutes. Such “on line” reporting by subspecialized radiologists can have a dramatic impact on patient care.

Neuroradiology Division

Fred Jenner Hodges, chair of Radiology at Michigan from 1931-1965, recognized the need for subspecialization in neuroradiology. So he arranged for Trygve O. Gabrielsen, who did his residency in radiology at the U-M from 1957-59 and joined the faculty in 1962, to obtain additional training in neuroradiology with Torgny Greitz at the Karolinska Institute in Sweden in 1965. Walter M. Whitehouse, who succeeded Hodges as chair of Radiology from 1965-1979, even more strongly supported subspecialization in neuroradiology and made it possible for Gabrielsen to undergo more highly specialized neuroangiography training with Per Amundsen in Oslo in 1968. This allowed Gabrielsen to modernize cerebral angiography at U-M using femoral arterial access techniques. The result was a switch from surgical direct puncture and injection of the carotid artery to cerebral angiography performed by catheterization of the femoral artery and advancement of a catheter through the arterial system to the point of contrast media injection. Michigan was one of the first institutions in the country to adopt this technique, which is now the standard access technique for angiography. Because of this, neuroradiology was given control of all invasive diagnostic neuroradiologic procedures, and a neuroradiology fellowship program was established at U-M. Gabrielsen spent 24 years as the director of the Division of Neuroradiology. Under Gabrielsen, the first U-M CT scanner, an EMI head imaging unit, was installed in 1975.

Despite disagreements with other departments regarding control of MRI, Radiology retained jurisdiction for magnetic resonance imaging. Under the guidance of William Martel, chair of Radiology from 1982-1992, the first clinical MRI facility in Michigan was established in 1983 using a 0.35 Tesla Diasonics superconducting magnet. This Diasonics unit was one of the earliest superconducting systems installed in a clinical facility in the U.S. As Martel’s successor as chair, Reed Dunnick markedly expanded MRI services, including installation of MRI units in outlying clinics.

In 1988, James A. Brunberg was recruited to become director of the Division of Neuroradiology. He left in 1998 to become chair of the Radiology Department at the University of California at Davis, whereupon leadership of the Neuroradiology Division was provided by Stephen Gebarski and then Suresh Mukherji. Gebarski and Mukherji worked to improve patient care by expanding neuroradiology services, acquire and optimally utilize the most modern imaging devices, and maintain ACGME certification of the Neuroradiology Fellowship. They continued the Gabrielsen tradition of innovative high quality neuroradiological clinical care, teaching, and research. In 2012, Mukherji became chair of radiology at Michigan State University. Ashok Srinivasan became division director of neuroradiology, building on the Gabrielsen legacy.

Nuclear Medicine Division

In the years immediately following World War II, intense interest pervaded both government and academia on how to develop peaceful uses of atomic energy. In 1946, publications described how radioiodine treatments, first with 130-I and later 131-I, had eliminated hyperthyroidism and reduced thyroid cancer in patients. The Oak Ridge National Laboratory in Tennessee offered both education in the employment of clinically useful radiation and therapeutic quantities of 131-I to personnel from approved hospitals. The concept was new, and its implementation introduced yet to be fully evaluated hazards: the radiation emanating from patients and their excreta would expose families and the public to this invisible energy. The Atomic Energy Commission, later the Nuclear Regulatory Commission, was charged with the regulation and oversight of diagnosis and treatment of patients with radioactivity.

Cyrus Sturgis, the chair of the Department of Internal Medicine, had treated hyperthyroid patients with thiouracil medications; he recognized the need for a better alternative to surgical thyroidectomy. William H. Beierwaltes, then an assistant professor in Internal Medicine who had also used thiouracil therapeutically, agreed to travel to Oak Ridge and subsequently accept the burden and the opportunities surrounding these radioactive compounds. In 1952, the Radioisotope Service was established in the Department of Internal Medicine with Beierwaltes as director. One of the first books on what was becoming a subspecialty, Clinical Use of Radioisotopes, was co-authored by Beierwaltes; a colleague, Phillip Johnson; and the physicist on the service, Arthur Solari.

The diagnosis and treatment of thyroid diseases with 131-I remained prominent components of the Service. But the horizon was rapidly widening. Unsealed radioisotopes in multiple pharmacologic preparations were expanding to permeate diagnosis and treatment of a myriad of human disorders, and by 1959, a number of young physicians at Michigan and across the nation sought education in this subspecialty. In accommodation, the service became the Nuclear Medicine Division with Beierwaltes, then a full professor, as chief.

Under Whitehouse and Martel, nuclear medicine remained within the Department of Internal Medicine. Substantial advancements were made by faculty such as James H. Thrall, who developed radionuclide ventriculography and diuresis renography; Thrall later became chair of radiology at Henry Ford Hospital and subsequently at Massachusetts General Hospital. John W. Keyes, Jr. fabricated the first single photon emission computed tomography (SPECT) unit in Michigan.

David E. Kuhl, who had an illustrious and internationally renowned career in Nuclear Medicine, was recruited to replace the retiring Beierwaltes in 1986. In the same year the Division of Nuclear Medicine, along with all components of the U-M Health Center, moved to the B-1 floor of the new hospital building adjacent to the Department of Radiology. The exception was Pediatric Nuclear Medicine which had been situated on the 4th floor of the Mott Children’s Hospital and eventually moved to the third floor adjacent to Pediatric Radiology. Patient care during the movements was maintained at the highest levels by the assiduous direction of Barry Shulkin.

Laboratory space was made available in the research buildings to accommodate expanding ideas and scope of the basic science in Nuclear Medicine.

In 2000, under the direction of the Medical School dean, Allen S. Lichter, the Nuclear Medicine Division was incorporated into the Department of Radiology.

Kuhl stepped down as chief in 2002, but continued his research inquiries. Barry L. Shulkin accepted the role of interim chief. Kirk A. Frey, already a prolific member of the Division, was appointed chief in 2004; he subsequently was named the David E. Kuhl Professor of Radiology.

The Radioisotope Service of the Ann Arbor Veterans Administration Hospital began operation late in October 1954 as a small unit on the second floor of the hospital under the supervision of a Radioisotope Committee. Beierwaltes acted as a consultant to the Committee. From 1955 to 1969, Walter DiGiulio was chief of the Service. In 1970 the Radioisotope Service was renamed the VA Nuclear Medicine Service. Rodney Pozderac became chief from 1972 to 1979, and James Thrall served as interim chief from 1979 to July 1980. Then Milton Gross joined the U-M faculty and became chief of the VA Service.

In response to increasing demands for a nuclear medicine service in multiple VA facilities in Michigan and elsewhere, a telenuclear medicine network was created in the late 1980s. The Ann Arbor VA Nuclear Medicine Service became the interpretative hub for affiliate Nuclear Medicine laboratories located in the Saginaw and Battle Creek VA Hospitals. Expansion subsequently included the Toledo, Ohio, VA Outpatient Clinic; the Fort Wayne, Indiana, VA Hospital; and a private cardiology practice in Flint, Michigan.

In 1990, Milton Gross became the national program director of the VA Nuclear Medicine Service. He was joined in 1991 by Lorraine Fig, as the deputy director, National VA Nuclear Medicine Service and Associate Chief, Nuclear Medicine Service.

After a major renovation and expansion of the Ann Arbor VA Hospital, the Nuclear Medicine Service moved to its present seventh-floor location in 1997. In an innovative VA national program, the Ann Arbor VA, in the early 2000s, began to share a PET scanner with the Nuclear Medicine Division at U-M; the association was one of the first where a Department of Veterans Affairs facility shared a PET scanner sited at its academic affiliate. This enhanced capability at the VA was followed in the mid-2000s by additional clinical innovations: a “mobile” PET/CT scanner in 2011, a fixed site PET/CT, and a SPECT/CT scanner in 2012.

For the Positron Emission Tomography (PET) Center, the additions of Richard Ehrenkauffer (Radiochemistry) and Richard Hichwa (Physics) saw the U-M PET Program develop visibly as a world-leading enterprise. The scientific “core” of the UM PET Center was then supercharged by the recruitment of Kuhl, and, he, in turn, brought in Robert Koeppe, Gary Hutchins (Physics) and Michael Kilbourn (Radiochemistry).

Two chemists, Raymond Counsel and Donald Wieland, had already been recruited by Dr. Beierwaltes with specific aims of synthesizing radiopharmaceuticals to concentrate in endocrine organs and tumors.

Single Photon Emission Computed Tomography (SPECT) had been introduced by W. Leslie Rogers and John Keyes. Subsequently, Kuhl directed and gave impetus to the research programs especially emphasizing the new emission tomography. Consequently, this attracted talented medical scientists, physicists, chemists and engineers to become full partners. Molecular imaging with emission tomography (SPECT plus PET) soon became essential components in the care of patients with cancer, as well as cardiac, endocrine and brain disorders.

The PET Program had begun with the application to the NIH for construction and equipping of a new facility. Through the efforts of a distinguished triumvirate of Beierwaltes, Sidney Gilman and Bernard Agranoff, a facility equipped with a cyclotron (30 MeV Cyclotron Corporation CS-30), a radiochemistry laboratory and a state-of-the-art human PET scanner was dedicated in late 1982, and marked the beginning of a long-running program of human PET studies of physiology and disease. The size and complexity of the PET program increased with the arrival of Kuhl to head it. He spearheaded the expansion of the facility’s resources, the number of investigators, and the research grant revenues. Upon the acceptance of PET imaging as a valuable clinical imaging tool, additional PET and then PET/CT scanners were added to activities in the Nuclear Medicine hospital site.

In 2003-2005, a new cyclotron (GE PETtrace) was purchased, expanding the research capabilities, and replacement radiochemistry facilities were constructed. The new facility, complete with cyclotron and multiple hot cells for radiochemistry, became fully equipped and good manufacturing practice (GMP)-compliant, and thus provided the resources for continuing success in radiopharmaceutical development. In turn, deliveries of unique and valuable agents for research and patient care in the Division of Nuclear Medicine were enabled.

Kuhl forecast that, in seeking earlier diagnosis, this research program might be linked to manifold drug development. New pharmacological classifications of patient groups (e.g., spatial patterns of metabolism, neurotransmitters, enzymes, proteinaceous accumulations) might redefine long-established clinical classifications and thus lead to earlier diagnoses and new treatments. Cancer research, hand-in-hand with drug development, gives hope that the targeting of new drugs might be matched to specific features of an individual’s tumor (e.g., proliferation, receptors, angiogenesis, hypoxia) and support personalized treatment that will be more effective. Real benefits to patients will continue to be driven by these kinds of long-term research projects, guided, in part, by lessons learned during this period.

In 2009, Kuhl was awarded the prestigious Japan Prize from the Japan Prize Foundation for the technological integration of medical science and engineering represented by his contributions to tomographic imaging in nuclear medicine.

The development of new brain imaging agents for neurological applications has long been a focus of research in nuclear medicine. Numerous unique radiopharmaceuticals were developed in U-M laboratories and introduced into clinical research studies: of these, four in particular ([123I]IBVM (iodobenzovesamicol), [11C]DTBZ (dihydrotetrabenazine), [11C]PMP (N-methylpiperidinyl propionate), and [18F]FEOBV (fluoroethoxybenzovesamicol)) have been utilized extensively to characterize specific neurochemical changes in studies of neurodegenerative diseases such as Parkinson’s and Alzheimer’s; the preparation and application of these radiopharmaceuticals have been duplicated at institutions world-wide.

A vibrant and innovative radiochemistry group headed by Donald Wieland has been exceptionally productive over decades. Agranoff supported Kirk Frey in his first novel radiotracer project, targeting radiotracer imaging of myelin in the brain. Lipophilic tracers with radio-iodinated iodobenzene, targeting white matter on the basis of its differential lipid content, were developed. However, since MRI provided outstanding white matter images, the enthusiasm for radiotracer-based imaging approaches diminished. Beierwaltes, Agranoff and Gilman succeeded in obtaining an NIH NINDS Program Grant to explore applications of positron tomography to human neurologic disorders.

As of 2014, the UM PET Center researchers are involved with the bench-to-bedside translations of numerous research radiotracers, imaging and quantifying novel CNS targets of interest in neurologic disease. This era evolved, logically, from an emphasis on proof of imaging feasibility to an increasing emphasis on disease-based image results, applicable to better diagnosis and to better understanding of pathophysiology. A major emphasis of the PET Center scientists, led by Kuhl, was in degenerative dementia — Alzheimer’s disease and related disorders. The group targeted improved understanding of cerebral glucose hypometabolism in dementia, and, with leadership from Satoshi Minoshima, discovered diagnostically-useful early disorders in the posterior cingulate/precuneus cortical region in the brains of AD patients. In addition, the group identified the metabolic distinction of dementia with Lewy bodies on the basis of hypometabolism in the occipital cortex. Members also focused on the introductions of new tracers to explicate the cerebrocortical acetylcholine system and the role of acetylcholine esterase (AChE), which were identified as abnormal in pathologic studies of AD patients. New tracers enabled the identification of the presynaptic acetylcholine vesicular transporters ([I-123]iodobenzovesamicol and [F-18]fluoroethylbenzovesamicol), acetylcholinesterase ([C-11]PMP) and muscarinic cholinergic receptors (scopolamine, TRB and NMPB).

The scientific achievements of the Michigan PET program have involved a long list of accomplished individuals, and, in addition to those mentioned above, have been key to the successes of the program, especially Roger Albin, a neurologist, who provided decades of leadership in selection, design and implementation of brain imaging projects.

In the early 1970s, Beierwaltes was aware of futile attempts to radiographically visualize the human adrenal glands. To overcome this diagnostic void, he hired a lipid chemist, Raymond Counsell. At the time, Counsell knew that, when injected intravenously, cholesterol labeled with 14-C concentrated in the cortices of adrenal glands in animals. Based on this information, he synthesized 131-I-19-cholesterol or iodocholesterol, also known as NP-59. Following the infusion of this innovative radiopharmaceutical, the adrenal glands of dogs were scintigraphically portrayed. In 1971, with the aid of a number of collaborators, the imaging of the human adrenal glands with a single photon camera became a reality. Subsequently in 1971, aided by the clinical experience of Jerome Conn, chief of the Endocrinology and Metabolism Division, Beierwaltes depicted a small and most-difficult-to-locate adrenal tumor, an aldosteronoma.

Over the following years, 131-I-19-iodochoesterol was found useful in differentiating disorders of the adrenal cortex. Although computed tomography and magnetic resonance imaging have largely replaced NP-59 scintigraphy, at its inception the agent was broadly acknowledged as a large leap in clinical science.

Beierwaltes also brought strategic vision and financial support for investigations of the adrenal medulla and related structures. Pheochromocytomas — neuroendocrine tumors found in the adrenal medulla and in other components of the sympathetic nervous system — became the major focus. These tumors secreted norepinephrine and occasionally epinephrine, hormones that evoked hypertension and devastating symptoms that were not infrequently lethal. Tests to detect the presence of such tumors were becoming more accurate, but their anatomic locations often remained elusive.

Donald Wieland was initially hired to synthesize a radiopharmaceutical that would delineate pheochromocytomas and related tumors. Building on prior research with the anti-hypertensive drug bretylium, Wieland synthesized meta-iodobenzylguanidine, which, when labeled with 131-I, was found to concentrate in the adrenal medullas of dogs. Now called 131-I-MIBG or MIBG, the agent was selectively sequestered by pheochromocytomas and similar neuroendocrine neoplasms which, when located on scintigraphic images, would direct excisions by skilled surgeons. It was a major clinical breakthrough that was initially reported by James Sisson and the investigational group. Subsequently, 131-I-MIBG in large doses reduced the size and hormonal secretions of malignant pheochromocytomas; a number of these studies were directed by Brahm Shapiro.

Another neuroendocrine tumor, neuroblastoma, is a lethal malignancy not uncommonly found in children. 131-I-MIBG was concentrated and retained in neuroblastomas to an extent that the radiation emanating from the decay of 131-I was therapeutic. Gregory Yanik, a pediatric oncologist, has played a prominent role in national and international studies that are designed to define the optimal doses and schedules for treatments with 131-I-MIBG.

Wieland also synthesized 11-C-hydroxyephedrine which, when infused slowly into children by Barry Shulkin, clearly portrayed endocrine tumors by employing PET technology. The tumor images were clarified by additional imaging with 18-F-fluorodeoxyglucose.

Beginning in the 1970s and early 1980s, there was much interest in scintigraphically displaying the heart and its functions. Edward Carr demonstrated that Rubidium-86 accumulated in the myocardium, but it was soon discovered that Thallium-201 yielded improved images. James Thrall collaborated with cardiologist Bertram Pitt to reveal not only the absence of Thallium-201 in areas of myocardial infarctions, but also Technetium 99m-pyrophosphate concentrations in the infarcted areas to portray the ischemic and dying myocardial cells.

131-I-MIBG readily entered the sympathetic neuron terminals in the myocardia of animals, and thereby hearts of patients were scintigraphically displayed. As expected, MIBG was absent in regions of denervation such as infarcts and diabetic neuropathy. For more precise information, the higher fidelity of PET imaging was sought. Wieland synthesized a pharmacologically related agent, hydroxy-ephedrine labeled with 11-C and denoted as HED. Markus Schweiger, a nuclear cardiologist, employed PET imaging of 11-C-HED to demonstrate that, as expected, transplanted human hearts manifested no sympathetic innervation. Then in 1991, he reported that over months, transplanted hearts were becoming re-innervated.

Richard Wahl, with a colleague in Medical Oncology, Mark Kaminski, employed 131-I-Tositumomab (Bexxar), a radio-labeled antibody that selectively accumulated in non-Hodgkin lymphomas to develop a therapeutic protocol. These designed treatments led to remissions of this lethal neoplasm.

Vascular/Interventional Division

Interventional radiology (IR) is a subspecialty of radiology that diagnoses and treats a variety of disorders using percutaneous methods guided by imaging. Image-guided interventional procedures today are safer, faster, minimally invasive and typically more cost-effective than traditional surgery. These interventional radiologic procedures can be grouped into vascular and non-vascular interventions. Included in the category of vascular interventional procedures are angioplasty, catheter-directed thrombolysis and thrombectomy, transcatheter embolotherapy, venous access, endovascular aneurysm repair (EVAR), and arterial, venous and oncologic interventions. Included in the category of non-vascular interventions are percutaneous image-guided biopsy and drainage, tumor ablation, and hepatobiliary, gastrointestinal and genitourinary interventions.

Vascular and interventional radiology began at the Michigan when Melvin Figley, a faculty member from 1950 to 1961, pioneered angiocardiography, assessment of cardiac physiology and the quantitative analysis of cardiac function. He described the angiographic diagnosis of aortic coarctation, ductus arteriosus, constrictive pericarditis, and congenital heart disease. He also introduced translumbar aortography and splenoportography at the University, and his description of the arteries of the abdomen and pelvis in both normal patients and in occlusive disease was widely used. Over the past 50 years, the outstanding faculty members in the Division have continued pioneering various interventional radiologic procedures, thereby paving the way for the practice of world-class modern interventional radiology at the University.

Joseph J. Bookstein succeeded Figley as head of the angiography section. He contributed significantly to advancing angiocardiography, diagnostic angiography and magnification angiography. His method of renal pharmacoangiography allowed hemodynamic assessment of renal artery stenosis. In the 1970s, Helen C. Redman and Stewart R. Reuter led the angiography section at the University-affiliated Wayne County General Hospital where they developed techniques of gastrointestinal angiography, selective arterial embolization for control of gastrointestinal bleeding, pharmacoangiography, angiographic diagnosis of pancreatic carcinoma, and angiographic evaluation of portal hypertension. In 1974, Reuter moved to the University, succeeding Bookstein. During these years, the program became nationally known and the fellowship attracted superior trainees.

In 1976, after Reuter moved to California, Kyung Cho headed the angiography division. He provided mentorship and an inspiring role model as well as outstanding leadership in many areas of medical research. These included development of embolization therapy to treat tumors and bleeding of internal organs, and angiographic treatment of renovascular hypertension caused by segmental renal artery stenosis. His investigation of hepatic microcirculation and the peribiliary vascular plexus helped advance the effective use of diagnostic imaging to understand hepatic circulation. Among all of his accomplishments, however, it is perhaps Cho’s work exploring the use of carbon dioxide (CO2) as a contrast agent for diagnostic imaging and endovascular intervention that may have the widest clinical application. CO2 can provide safe vascular imaging in patients who are allergic to the commonly-used iodinated contrast media used in angiography or who may be susceptible to the potential nephrotoxicity of iodinated contrast media; this has the potential to save many lives and prevent long-term complications. To advance this practice Cho wrote the world’s first textbook on CO2 angiography to demonstrate its safety in comparison with other contrast agents and to educate angiographers about CO2’s advantages, uses, and contraindications.

In the 1980s, the technical ability to record radiographic images digitally had a major impact on catheter-based angiography. Digital subtraction angiography (DSA) became an important imaging tool in the evaluation of vascular diseases. It largely replaced cut film angiography in both diagnostic and therapeutic radiologic vascular procedures. All angiography suites in the Division are equipped for digital subtraction angiography.

The division directors from 1996 to 2013 included Victoria Marx, David Williams, and James Shields. Wael Saad was appointed as Director in 2013. The Division has outstanding faculty members including Kyung Cho, Narasimham Dasika, Joseph Gemmete, Venkat Krishnamurthy, Paula Novelli, Ranjith Vellody, Wael Saad, James Shields, Jonathan Willatt, David Williams, and Zishu Zhang. Cho retired in 2015 after 42 years serving the Department and the patients of the University of Michigan.

Catheter-based angiography remains the most reliable means of visualizing blood vessels. Today, blood vessels in almost any region of the body can be catheterized allowing the use of the circulatory system as a roadway to perform therapies for a variety of diseases. This pioneering quest for minimally invasive image-guided treatment of pathologic conditions represents a core value of interventional radiology. The outstanding interventional radiologists at the University continued to develop new interventional procedures including intravascular ultrasound, endovascular intervention, endovascular repair of aneurysm (EVAR), percutaneous fenestration for the treatment of type 2 aortic dissection with malperfusion syndrome, oncologic intervention, transcatheter embolotherapy, and percutaneous hepatobiliary intervention.

Changes in the health care landscape dictate that in order to remain competitive and thrive as a leading healthcare provider for minimally invasive, image-guided therapy, IR at the University has adapted to a new paradigm that includes both marketing and collaboration. To achieve this model, IR has undertaken a strategy to advance image-guided therapy, improve clinical care management skills, maintain an outpatient clinic, and promote IR services to the patient.