Most residency programs for radiation oncology physicists do not reflect the heightened importance of medical imaging

With the widespread application of highly conformal radiotherapy techniques, imaging has taken on increased importance both in planning and delivery. Radiation oncology physicists increasingly have to use imaging systems without necessarily having in-depth knowledge as to how these systems work. This is due to the educational programs they attended, which allowed students who planned to specialize in radiation oncology, to graduate without the same level of detailed knowledge of imaging required of their counterparts who wanted to specialize in imaging. It might be hoped that current radiation oncology physics residency programs have remedied this situation by including sufficient education in all the imaging modalities used in therapy, but it has been suggested that this is not the case for most programs. This is the topic debated in this month's Point/Counterpoint. Medical imaging plays an increasingly important role in radiation oncology.1 Magnetic resonance imaging (MRI) and positron-emission tomography (PET), together with computed tomography (CT), have become a part of the standard imaging tools used for radiation therapy treatment planning and tumor response monitoring. Intensity modulated radiation therapy has significantly improved dose conformity and has led to more stringent requirements for immobilization and localization in radiation treatment delivery. Image-guided radiation therapy (IGRT), which enables in-room target localization, has been developed to meet these challenges. Imaging technologies available for IGRT include electronic portal imaging, ultrasound-based techniques, kilovoltage (kV) x-ray imaging, integrated CT/linear accelerator systems, tomotherapy with megavoltage (MV) CT, and cone beam CT (CBCT) using kV and MV x rays. Clinical radiation oncology physicists ought to have in-depth knowledge of these imaging technologies. Radiation oncology physics residency programs are the formal training programs for future clinical physicists. Recognizing the need for structured clinical training for physicists wishing to practice professional medical physics, the AAPM published a report on essentials and guidelines for hospital-based residency training programs in 1990,2 with an updated version in 1996 (AAPM Report No. 90).3 The CAMPEP (Ref. 4) has accredited medical physics residency programs based on AAPM Report No. 90 (Chapter 3). Section 3.4.4, Training Content, of the report does not explicitly include imaging modalities that are important to radiation oncology except radiographic/fluoroscopic and CT images for simulation, although Sec. 3.5.4.A3 on IGRT does briefly mention MRI, PET, ultrasound, and image registration and fusion in addition to CT. Unless individual residents have had previous training in imaging physics, the current curriculum of radiation oncology physics residency programs based on the AAPM Report No. 90 does not provide the in-depth training in imaging physics which is critical to IGRT. For example, radiation oncology physics residents do not have sufficient training to understand factors affecting image quality for CBCT, commonly used for in-room IGRT, and MRI, widely used for defining target volume and soft tissues. One could argue that radiation oncology physicists do not need to have in-depth knowledge of imaging physics: When needed, they could seek help from their colleagues who specialize in imaging physics. But radiation oncology physicists trained in all aspects of the radiation oncology workflow have a better appreciation of the effects of image quality on the planning and delivery of radiation dose distributions. They are responsible for the problems that can prevent them from planning, and safely and accurately delivering doses. Radiation oncology physicists with a better understanding of imaging physics can make better clinical judgments. Therefore, residency programs for radiation oncology physicists should provide an in-depth training in medical imaging in this era of IGRT. It is true that during the past decade, it has become increasingly common to use multiple imaging modalities to define and localize the treatment volume. All treatment planning systems are able to use multiple imaging modalities such as CT, PET, MRI, or U.S. for external beam and brachytherapy planning. Also, all treatment delivery systems can be purchased with some form of built-in imaging modality such as cone beam CT (kV or MV) to localize the treatment volume. But currently, most clinics use just CT imaging for external beam treatments or U.S. for prostate brachytherapy. Direct use of other imaging modalities in treatment planning systems is still quite limited. The topic of how much a radiation oncology physicist should know about imaging will be debated for years. The AAPM Task Group Report on academic programs for graduate degrees in medical physics recommends that “To some degree, image science is required knowledge for any medical physicist, but details of magnetic resonance image science are more pertinent to the specialist.”5 The same can be said for radiation oncology physics residents. End users of imaging modalities in treatment planning do not need to know the imaging system in detail. What they need to know is how to administer a comprehensive quality assurance (QA) program for different imaging modalities such as those presented in AAPM Task Group Reports.6,7 Graduate and residency programs for radiation oncology physicists do not have to cover in detail how each modality functions, but should instead teach how to perform QA for these imaging modalities. Ultimately, it is up to the ABR to tackle this issue rather than individual programs. In the Part I: General section of the Examination Study Guide for Radiologic Physics,8 two imaging modalities, nuclear magnetic resonance and ultrasound, are already included. Moreover, for the Oral Examination4 in Radiologic Physics, out of the five subject categories included, one is “image acquisition, processing and display.” So the requirement for a practical understanding of medical imaging for radiation oncology physicists is already in place. The case for increasing the imaging component of radiation oncology physics residency programs is premature at best. Dr. Das acknowledges the importance of medical imaging in the era of IGRT but suggests that radiation oncology physics residents only need to learn how to establish and manage a comprehensive QA program for the imaging devices used in radiation oncology, and do not need to have in-depth knowledge about how each imaging modality functions. I beg to differ. First, simply administering a comprehensive QA program for imaging devices used in the clinic is not sufficient. As asserted in my opening statement, radiation oncology physicists should know more about medical imaging than currently required in residency programs because only they have an appreciation of the effects of image quality on the ability to plan and deliver radiation treatments appropriately. Second, even if just for establishing an effective QA program, radiation oncology physicists should know more about how each imaging modality functions. Without sufficient knowledge, they will not be able to establish an effective and efficient QA program for the imaging devices used. I agree with Dr. Das that the curriculum for an individual program should not be drastically changed until the AAPM, CAMPEP, and the ABR have established new guidelines. It is also true that there is a medical imaging component for radiation oncology physicists in the current ABR examinations. But I would argue that in this era of IGRT, with the heightened importance of medical imaging, the medical imaging component of the exam is insufficient in terms of depth of knowledge required and, therefore, current training of radiation oncology physics residents in imaging is inadequate. Dr. Zhu states that if radiation oncology physicists have a better understanding of imaging physics, they can make better clinical judgments. I could not disagree more. Clinical judgments should be made by radiation oncologists and not by radiation oncology physicists. Since radiation oncologists are using these imaging modalities to derive the planning target volume and localizing it for daily treatment, they should consult their imaging counterparts, the radiologists, in making these decisions. A physicist should provide them with information on uncertainties associated with these modalities. Physicists should also be extensively involved in QA for these imaging devices, thereby instilling confidence in the definition and localization of targets and normal tissues. Dr. Zhu's claim that residency programs should provide an in-depth training in medical imaging for residents in radiation oncology physics is not practical. Other than a handful of large academic institutions, most radiation oncology physics departments/sections do not have experts in imaging. In my institution (University of Wisconsin), there are three courses on imaging (CT, MRI, and ultrasound) offered each year. Each of these courses is three credit hours. Each credit hour corresponds to one hour of class per week for a semester. Graduate students interested in specializing in imaging physics are required to take these courses for better understanding and in-depth knowledge. So a total of nine credit hours will be required to give a radiation oncology resident an in-depth training on these imaging devices. Enlarging the curriculum of a radiation oncology physics training program to imaging physics at this level will be burdensome for the training institutions. I still believe that implementation of a comprehensive QA program on the imaging modalities used in a radiation oncology clinic might be all that we need. Not only will it give some insight to the residents on the imaging modalities that are being used in the clinic, but it will also teach them the importance and benefits of such a QA program.