Gordon J. Harris, PhD

While the techniques associated with 3D imaging provide an answer to the challenges of proliferating slices in tomographic imaging, the economics and reimbursement of 3D imaging require careful attention.

The purpose of this article is to explore the growing need for 3D imaging services, as well as the financial considerations and challenges to any institution that would like to develop this capability. Specifically presented are the advantages of having a 3D imaging service readily available, the costs of supporting the appropriate staff and equipment, and issues with proper coding of claims and obtaining reimbursement from insurance. A significant problem with Medicare payment for computed tomographic angiography (CTA) is described, as are the steps that Massachusetts General Hospital (MGH) has taken with the American College of Radiology (ACR) to convince the Centers for Medicare and Medicaid Services (CMS) to address this underpayment.

Three-dimensional medical imaging has revolutionized both radiological diagnosis and surgical planning. Reconstructions in three dimensions provide lifelike views that can quickly summarize the relationship among anatomic structures to facilitate accurate diagnosis or to identify the optimal treatment plan before the patient is in the operating room (OR). The benefits include decreased exploratory time in the OR, less damage to healthy tissues, and a lower risk of complications for the patient, all of which contribute to reduced surgical morbidity.  Increased diagnostic sensitivity across all specialties and the likelihood of shorter OR time per procedure contribute to reduced costs. Thus, the quality of care is greatly enhanced, often with a decrease in cost compared to alternative procedures such as catheter angiography (see Figure 1).

Pamela Kassing, RCC

Multi-detector CT (MDCT) opened new possibilities and demand for 3D imaging.1,2 Rapid acquisition using thin slices and improved rendering algorithms facilitates exquisite 3D reformats. The pace of progress is being hastened with the rapid developments in MDCT technology with availability of 4-, 8-, and 16-slice scanners and, more recently, volume CT.3,4 From a practical standpoint, developments in MDCT scanning allow multiple thin slices to be acquired with increased z-axis coverage, resulting in an extraordinary increase in the quantity of acquired data. Review of large numbers of images, often hundreds of slices per study, poses significant constraint on radiologists’ efficiency and may be simply impractical for most radiology departments. Having 3D capability available for diagnosis and surgical planning allows the radiologists and referring physicians to first get a summary view of the entire anatomy in a few concise anatomically clear 3D views, and then refer back to the original 2D data for comparison and confirmation. Thus, 3D reconstruction is more frequently becoming a valuable technique to summarize in a concise and clear way the overwhelming number of slices produced by modern scanners.

Techniques and Applications

3D imaging includes a variety of tools and techniques. CT and MR collect a series of 2D slices through the region of the body under study. These 2D slices can be “stacked” and combined into a single 3D volume of data. However, this requires relatively thin slice acquisition. The thinner the slices and closer to isotropic voxel dimensions, the better the resolution of the reformatted images. For example, most 3D protocols involve 0.5 to 3 mm-thick slice data, while thicker slices, such as 5 mm, produce blurred reconstructions out of plane with a stepping artifact.

Patricia Linton, NP

Once the data is imported into a 3D imaging workstation, the volume of data can be resliced in different planes, or visualized through various volumetric viewing techniques. For example, maximum intensity projection (MIP) or minimum intensity projection (MinIP) displays the maximum or minimum value through the volume along each ray drawn from a particular viewpoint. These techniques are good for instances where the interpreting physician does not want the image blocked at the surface, for example in visualizing vascular stenosis or craniosynostoses. Volume and surface rendering provides an anatomical model that allows selected tissues to be either displayed or stripped away to reveal the structures of interest. Surface rendering and volume rendering are useful in viewing bony anatomy, as well as visualizing vessel surfaces for planning treatment approaches for aneurysms. The specific type of image reconstruction depends on the anatomy, the purpose (diagnosis or surgical planning), and the customer needs (radiologist or surgeon). In general, diagnostic radiologists often prefer MIP views rather than volume- or surface-rendered views because all the data is visible, while the surface- and volume-rendered views make some objects transparent. Surgeons, however, often prefer the rendered views because they display a more anatomically realistic view for surgical planning purposes. In addition to these volumetric visualization techniques, a 3D volumetric dataset can be reformatted in a variety of ways. The simplest type of reconstruction is multi-planar reconstruction (MPR), in which the three primary orthogonal views are produced. If thickness is added to these views, this becomes multi-planar volume reconstruction (MPVR). The reconstructed planes can also be created in any oblique plane, or produced in a curved reformat to create a reconstructed view that follows a curved structure such as a blood vessel, the mandible or maxilla, or the alveolar nerve canal.

Patricia Levy-Zuckerman

Currently, the 3D Imaging Service at Massachusetts General Hospital processes an average of more than 80 examinations per day, which represents approximately 5% of the hospital’s CT scan volume, plus about 20% of the MR and ultrasound volume. The 3D Imaging Service processes examinations for vascular, orthopedic, chest, breast MR, GI/GU, emergency, and pediatric examinations, however, about half of the 3D examinations are neuroradiology studies of the brain, head, and neck. Many neuroradiology examinations involve studies of the carotid and/or neurovasculature with CT or MR angiography (CTA, MRA). In addition, a variety of neurology, neurosurgery, and oral-maxillofacial examinations are reconstructed with 3D imaging, including brain surface renderings for neurosurgical planning or stroke and aneurysm imaging; brain tumor volumetric measurements; as well as imaging of facial malformations, craniosynostoses, complex facial fractures/trauma, and surgical planning for dental implants, extractions, and nerve damage evaluations. The 3D Imaging Service also processes all the vascular studies of the body, including scans of the aorta, renal, hepatic, and/or pancreatic vasculature and the runoff CTA and MRA examinations. High-quality, consistent, reliable 3D reconstructions are essential to diagnosis and treatment planning of vascular imaging. For example, aortic aneurysm treatment,5 liver resection,6 and living liver7,8 and renal donor9 planning are greatly aided by presurgical evaluation and planning on computer rather than exploration in the OR. Three-dimensional imaging is also effective in assessing pancreatic cancer and associated vascular involvement,10 and for virtual endoscopy, such as virtual colonography for assessing colon polyps.11,12

Establishing a 3D Service

Staffing. There are two major hurdles to developing a consistent, reliable, and cost-effective 3D Imaging Service at a facility: 1) determining who will perform the 3D processing (ie, appropriate staffing), and 2) capturing sufficient revenue to cover the costs of the service (ie, reimbursement). One consideration in determining the cost-effectiveness of the service would be the cost of not having a dedicated 3D Imaging Service. For example, if a radiologist were to spend a few hours per day performing 3D processing, he would be diverted from reading additional clinical cases, which would have generated added revenue, and the facility might need to hire additional radiologists or moonlighters to read examinations. Likewise, if CT or MR technicians were to perform 3D imaging between scans, it would slow down the scanner throughput, also decreasing revenue. A primary objective in establishing a 3D Imaging Service should be to allow the technologists, radiologists to focus on their primary revenue-generation modes for the practice and hospital, namely, scanning the patient and acquiring the images and interpreting the examinations, respectively. Meanwhile, a dedicated staff for 3D Imaging Services can provide an efficient and consistently processed set of 3D views.

Figure 1. CTA compares favorably twith the more expensive and invasive angiogram procedure (CTAs on left). A stenosis is clearly shown in the top of images with bold modalities. The bottom pair shows the added detail visible on CTA. Image courtesyof Gordon J. Harris,PhD,Massachusetts General Hospital, Boston.

Many complex 3D examinations, CTA for example, require 45 to 60 minutes or more to process all of the associated views, while other less complicated 3D examinations may take only 15 to 30 minutes each to process. If the clinical volume at a facility is sufficient, it would be cost-effective to have one or more dedicated 3D technologists process these examinations (or alternatively, consider outsourcing the 3D image processing to a centralized 3D service). A dedicated 3D technologist should be hired as an experienced, trained, and certified CT or MR technologist who can work closely with the radiologists and referring physicians to create 3D protocols that generate the most useful views for each type of examination. The dedicated 3D technologist is akin to a computer graphics medical artist who can take the cross-sectional scans and craft them into a beautiful set of concise, anatomically detailed 3D views. To accomplish this, the 3D technologist must understand the cross-sectional anatomy and pathology in depth, as well as scan artifacts, and must learn to operate complex computer software. This is a very specialized set of skills and experience, and thus the training period for a dedicated 3D technologist ranges from 3 to 6 months.

Many hospitals, including MGH, have established a 3D Imaging Service that has become an integral part of the clinical workflow for radiologists and referring clinicians. The 3D Imaging Service at MGH has been in operation since February 1999, with a staff that has grown to six full-time 3D technologists and two image-analysis specialists, a billing coordinator/administrator, a technical director, a systems administrator, and a director. Cases are pulled into the 3D laboratory from the picture archiving and communications system (PACS), and processed on one of several FDA-approved medical image-processing workstations. The 3D views are then saved as a new series of the study on PACS for review, as well as on the Web image server for review by referring physicians on their PCs. Upon completion and after radiologist’s interpretation, the 3D reconstruction procedures are finalized in the clinical information system, associated with the primary examination, and a bill is generated to obtain payment for the service when appropriate. The 3D Imaging Service at MGH is supported entirely by revenue from these services.

Billing and Compliance . The second hurdle in justifying a 3D Imaging Service is to determine how to capture revenue to which the facility is entitled in order to support the 3D imaging staff and equipment cost and to demonstrate cost-effectiveness. Carrier reimbursement from Medicare and other health insurers is the main source of revenue to support 3D imaging, but it is critical that services be billed to Medicare and other payors only when it is appropriate to do so. In all circumstances, the 3D reconstructions must be performed on software that has the required FDA clearances when billing Medicare.

At MGH, Medicare coverage guidelines are used as the benchmark for all carriers. Because payor mix varies widely among hospitals, it is important to confirm the similarity between commercial/private carriers in a specific geographical region with Medicare. This discussion will focus on Medicare reimbursement as the revenue framework for a 3D Imaging Service.

The overarching criterion for Medicare reimbursement is a careful consideration of the medical necessity of the service and procedural code definitions.13 For example, insurers continue to scrutinize the medical necessity of claims for CTA or MRA (rather than another type of imaging). The medical necessity of CTA or MRA can be supported by the need to image the vasculature. This necessity should be reflected in the radiology report by documenting the patient’s history and/or indications for the examination as specified by the referral or patient’s medical record. Appropriate codes must be used to reflect the services provided. With the exception of CTA and MRA, 3D imaging is a separately reportable service with CT, MRI, or ultrasound procedural codes only if both the initial examination and the 3D imaging are medically necessary. CTA and MRA codes, which include the 3D imaging as an integral part of the service, must be used to describe CTA and MRA services. However, if a request were made for previously acquired CT, MR, or ultrasound images to be reconstructed into 3D images for further diagnostic evaluation, then CPT® 76375 would be coded appropriately as an add-on for this service.

The medical necessity of 3D imaging must be supported in one of three ways: 1) a direct order from the requesting physician; 2) positive findings during an ordered examination; or 3) carefully monitored radiology departmental policies.

Figure 2. Displayed are several views of 3D post-processed CTA head and neck images showing (left to right) endoluminal view of carotid bifurcation, volume rendered view of carotid, and curved reformat of carotid from aortic arch to cerebral arteries. The 3D Imaging Service at Massachusetts General Hospital typically creates 30 to 40 standard views, a process that takes 45 minutes to an hour (particularly time-consuming are the curved reformats showing the extended vessels). Prior to 2001, the time and equipment necessary to produce these images was underwritten in part by the reconstruction add-on code. The payment rate for a new code subsequently created for CTA and MRA, however, is inadequate, because Medicare bases the rate on hospital charges, many of which are not charging more for CTA than CT. Image courtesy of Gordon J. Harris, PhD, Massachusetts General Hospital, Boston.

A direct order from a referring physician for a specific scan with 3D imaging (ie, a surgeon requesting a CT or MR examination with 3D reconstruction for surgical planning purposes) supports the medical necessity of 3D imaging. In addition to the requesting physician’s order, medical necessity can be supported by the radiologist’s documentation of the patient history or indications for the examination.

If upon the reading of an examination, the radiologist notes positive findings and deems additional views are necessary to address findings that impact the patient’s well-being, both the positive findings and the reason for the additional view(s) must appear in the radiology report to support the medical necessity of the 3D imaging.14

The third way that 3D cases may be performed and billed to Medicare is if the department and/or referring physicians determine that 3D reconstructions should be performed as part of the standard of care for a particular type of examination. This involves a carefully monitored, examination-specific, interdepartmental process. The radiologist writes a protocol policy for an examination to be associated with 3D imaging that presents medical necessity, reasonableness, and standard of care as supported in the medical literature. The policy is then reviewed for medical necessity, coding accuracy, and billing compliance by the chief of radiology, billing and coding offices (professional and technical), compliance offices (professional and intermediary), and radiology administration. When the policy is approved, a notice is sent to all referring physicians to notify them of the medical necessity and rationale for performing 3D in association with a particular examination. The referring physicians are also informed that a 3D charge will be billed to the beneficiary in addition to the charge for the primary requested examination. These policies are restricted to specific diagnostic situations in which omitting 3D imaging would lower diagnostic confidence below practice standards and compromise patient care. The notice also needs to provide the requesting physician with directions on who to contact with questions or concerns regarding the medical necessity of the 3D examination. Once the new policy notice has been communicated to requesting physicians, the policy is made available upon request in the department of radiology and placed in the radiology policies and protocol manual.

To monitor ongoing billing compliance, this type of policy is subject to rigorous interdepartmental review on a yearly basis. Clearly, the role of billing compliance has moved beyond the traditional retrospective analyses to a more integrated role with the practice, a partnership in which fiscal well-being is fully understood to be hand in hand with adherence to government regulations.

Economics and Reimbursement

For 3D reconstruction (other than CTA and MRA), the image postprocessing is billed as an additional code to the primary CT, MRI, or ultrasound examination under CPT code 76375, “coronal, sagittal, multiplanar, oblique, 3-dimensional and/or holographic reconstruction of computerized tomography, magnetic resonance imaging, or other tomographic modality.” This CPT code was originally created for 2D reformatting using “Coronal, Sagittal, Multiplanar and/or Oblique Reconstruction” of CT scans, generally performed on the scanner console at the time of the scan. Then, in 1988, 3D was added to the code description, and in 1998, the description was expanded to include other tomographic modalities such as magnetic resonance imaging and 3D ultrasound.15

Insurance reimbursement is an important factor in developing a self-supporting clinical 3D imaging service. For Medicare outpatients, depending on whether the technical costs (staff, equipment, etc) are billed for by the radiology practice or through the hospital, the technical component of the service would be billed either under the Medicare Hospital Outpatient Prospective Payment System (HOPPS) ambulatory payment classification (APC) system for hospital billing, or under the Medicare Physician Fee Schedule (MPFS) under the Resource Based Relative Value Scale (RBRVS) system for nonhospital billing. For cases when the add-on code can be billed separately, the technical component under HOPPS is approximately $100, and for freestanding clinics billing under the MPFS, it is about $160. The professional component is minimal (about $9). The Medicare payment for 3D images created as part of a CTA or MRA, which do not allow separate billing and payment for the add-on CPT code, has been the source of financial concern. In the next section, the shortcomings of Medicare payment for CTA and MRA are discussed, as well as steps that have been taken to address the problem.

Medicare challenges: CTA and MRA

Prior to 2001, appropriate billing for CTA included two separate CPT codes: CT for the specific anatomy involved, plus the 3D reconstruction add-on code, CPT 76375.16 Thus, 3D reconstructions for CTA, as well as for MRA, could be appropriately billed separately, as with other reconstructions, as an add-on to the primary examination. However, effective January 2001, the American Medical Association’s CPT editorial panel created new codes specifically for CTA indicating that 3D reconstructions must be performed or these examinations are not considered CTA.17 The code descriptor for CTA includes the phrase “without contrast material(s), followed by contrast material(s) and further sections, including image post-processing.”18 Thus, the add-on CPT code for 3D reconstruction may not be billed with CTA or MRA as of 2001.17,18

Unfortunately, when CMS established payment rates for the new CTA codes, it did not recognize the additional facility costs of performing CTA as compared with ordinary CT (ie, the additional 3D imaging reconstruction in a CTA). Therefore, CMS assigned CTA the same payment rate as the lesser CT service under both the HOPPS and MPFS payment systems. For example, for hospital-based services billed under HOPPS, CMS originally placed CTA in the same APC grouping as CT studies, meaning the same reimbursement was set for both procedures. Effectively, this eliminated any additional reimbursement for 3D postprocessing; of course, the facility had to bear the costs of the additional equipment and staff time to create the images that distinguished CTA from CT. The reasons for this oversight appear to be confusion about the actual differences in the services and intricacies of how Medicare rates are set under the HOPPS and MPFS systems.

This change in coding for CTA and MRA has had a big impact on the cost-justification for 3D imaging services. CTA and MRA account for half of all 3D examinations processed at MGH. Prior to the coding change in 2001, the added $100 to $160 reimbursement for 3D reconstructions under CPT 76375 supported the costs of 3D workstations, technologist time, and extra images for the 3D imaging service, but since 2001, the add-on 3D code can no longer be billed separately for these examinations. The 3D Imaging Service at MGH negotiated with the hospital finance department in 2001 to develop a reimbursement-sharing model that allows for the 3D Imaging Service to obtain a portion of the total hospital payments for the CTA and MRA services, despite the fact that the hospital gets no additional reimbursement compared with CT to cover the 3D component. To accomplish this, a per-examination fee was agreed upon for 3D processing on the CTA and MRA examinations similar to the regionally adjusted reimbursement for CPT 76375 (approximately $115), and thus the hospital agreed to help cover the costs of providing this service.

During the past 2 years, the 3D Imaging Service at MGH, together with the American College of Radiology, has worked to successfully petition CMS to allow for a differential payment for CTA by removing CTA from the APC group with CT (CTA was moved from APC 0333 to a new APC 0662 in 2003). This new APC does allow for a differential payment amount between the services, but does not guarantee what that difference will be. To set a rate, CMS looks at hospital charges, adjusted by the hospital-specific cost-to-charge ratio, to determine an “imputed cost.” It is these costs that are the basis for each APC’s payment rate.

Unfortunately, CMS uses Medicare claims data to determine the payment rate for the two separate APC groups, and these data, based on 2001 and 2002 Medicare claims, were not a valid reflection of differential hospital resource use. An analysis of the 2001 and 2002 claims data found that less than half of all hospitals that submitted CTA claims charged more for CTA than CTclearly indicating that claims were not reflective of the additional resources needed. Based on this finding, comments were submitted to CMS explaining that, since a CTA involves all the costs of CT plus the additional 3D postprocessing, CTA should always be a higher cost procedure relative to CT, and thus, the claims data were not possibly accurate if CTA charges were equal to or less than CT charges at most hospitals. It was also argued that the claims were misguided, based on documented erroneous and conflicting coding advice in 2001 from the AMA on whether the 3D charge (CPT code 76375) could be billed with CTA.19 However, CMS denied the request that they adjust the reimbursement rate and discard the clearly flawed claims data. Thus, despite the separation of CT and CTA into two APC groups to allow for differential payment, the net differential was only $5 in 2003, and only $24 in the 2004 proposed rule (see Figure 2,). A comment was submitted on this issue for the 2004 proposed rule as well.

Considering that Medicare rates are based on hospital charges, adjusted by a cost-to-charge ratio, it is imperative that hospitals consider the charges for the new CTA codes carefully. Advocates for CTA urge hospital radiology administrators to consult with their chargemaster managers and revise the CTA charges at their hospital so that CTA charges reflect the CT plus the 3D reconstruction costs. If hospitals that are charging erroneously (charging less for CTA than CT) will make this adjustment to reflect the added costs involved with CTA versus CT, then the claims data will be more valid, and, over the next few years, the differential for CTA reimbursements will rise accordingly. Otherwise, it will be more challenging for new 3D laboratories to justify their added costs of staff and equipment to the hospital, and for hospitals to justify the lost revenue of shifting from more invasive procedures such as catheter angiography to less expensive and less invasive 3D imaging procedures such as CTA and MRA, which translates into better patient care.

Gordon J. Harris, PhD, is director, 3D Imaging Service, Massachusetts General Hospital, and associate professor of radiology, Harvard Medical School, Boston

Pamela Kassing, RCC, is senior director of Economics and Health Policy, American College of Radiology, Reston, Va

Patricia Linton, NP, is MGH/MGPO compliance analyst and educator, Massachusetts General Hospital, Boston

Patricia Levy-Zuckerman, is principal, Covance Health Economics and Outcomes Services, Inc, Gaithersburg, Md.


  1. Rubin GD. 3-D Imaging with MDCT. Eur J Radiol. 2003;45(suppl 1):S37-41.
  2. Berland L, Smith JK. Multidetector-array CT: once again, technology creates new opportunities [editorial; comment]. Radiology. 1998;209:327-329.
  3. Klingenbeck-Regn K, Schaller S, Flohr T, Ohnesorge B, Kopp AF, Baum U. Subsecond multi-slice computed tomography: basics and applications. Eur J Radiol. 1999;31:110-124
  4. Hu H, He HD, Foley WD, Fox SH. Four multidetector-row helical CT: image quality and volume coverage speed. Radiology. 2000;215:55-62.
  5. Tillich M, Hill BB, Paik DS, et al. Prediction of aortoiliac stent-graft length: comparison of measurement methods. Radiology. 2001;220:475-483.
  6. Wigmore SJ, Redhead DN, Yan XJ, et al. Virtual hepatic resection using three-dimensional reconstruction of helical computed tomography angioportograms. Ann Surg. 2001;233:221-226.
  7. Bogetti JD, Herts BR, Sands MJ, Carroll JF, Vogt DP, Henderson JM. Accuracy and utility of 3-dimensional computed tomography in evaluating donors for adult living related liver transplants. Liver Transpl. 2001;7:687-692
  8. Herfarth C, Lamade W, Fischer L, et al. The effect of virtual reality and training on liver operation planning. Swiss Surg. 2002;8:67-73.
  9. Kawamoto S, Montgomery RA, Lawler LP, Horton KM, Fishman EK. Multidetector CT angiography for preoperative evaluation of living laparoscopic kidney donors. AJR Am J Roentgenol. 2003;180:1633-1688.
  10. Fishman EK, Horton KM. Imaging pancreatic cancer: the role of multidetector CT with three-dimensional CT angiography. Pancreatology. 2001;1:610-24
  11. Rubin GD, Beaulieu CF, Argiro V, et al. Perspective volume rendering of CT and MR images: applications for endoscopic imaging. Radiology. 1996;199:321-330.
  12. Fenlon HM, Nunes DP, Schroy CP, Barish MA, Clarke PD, Ferrucci JT. A comparison of virtual and conventional colonoscopy for the detection of colorectal polyps. N Engl J Med. 1999;341:1496-1503.
  13. Centers for Medicare and Medicaid Services. CMS Program Manuals—Medicare Carrier (Pub. 14), Part III—Claims Process. October 25, 2002. Transmittal 1776:1.
  14. Centers for Medicare and Medicaid Services. CMS Program Manuals—Medicare Carrier (Pub. 14), Part III—Claims Process. 15021 Ordering of Diagnostic Tests. October 25, 2002. Transmittal 1776:1-5.
  15. American College of Radiology. Code 76375 and coding issues. ACR Bulletin. July 1999:6.
  16. American Medical Association. CPT Assistant. April 1996;6(4):11.
  17. American Medical Association. CPT Assistant. July 2001;11(7):3.
  18. American College of Radiology. New and revised CPT codes for 2001: part I. ACR Bulletin. December 2000:4.
  19. American Medical Association. CPT 2001 Coding Symposium—Radiology, Commonly Asked Questions. 2001.