Parallel/Opposed: IMRT QA using treatment log files is superior to conventional measurement‐based method

In the recent years, more and more researchers believe measurement‐based quality assurance (QA) method for intensity‐modulated radiotherapy (IMRT) is insensitive in detecting various types of failures. 1 , 2 , 3 Machine delivery log‐file analysis has been proposed to be a more effective and efficient approach in verifying IMRT delivery accuracy in terms of gantry, collimator, jaws, and MLCs. 1 , 4 , 5 , 6 , 7 , 8 , 9 However, whether log file measurements can replace conventional QA methods remains a major debate in current medical physics society. This is addressed as our first parallel/opposed topic. Nathan Childress is parallel to the argument. Dr. Childress received his Ph.D. in Medical Physics in 2004 from University of Texas‐M.D. Anderson Cancer Center, his M.S. in Nuclear Engineering, and his B.S. in Chemical Engineering from the University of Missouri‐Columbia, both in 2001. His dissertation focused on bulk IMRT QA analysis and film dosimetry. He worked as a clinical physicist for six years at The Methodist Hospital before founding Mobius Medical Systems, LP in 2010. Mobius has designed and developed software packages that perform linear accelerator QA, treatment plan QA, and IMRT QA. He created and maintains www.medphysfiles.com, a site for the free sharing of files related to clinical medical physics. Dr. Childress is a Section Editor for the JACMP and is certified by the American Board of Radiology in Therapeutic Radiological Physics. Quan Chen is opposing the argument. Dr. Chen obtained his Ph.D in medical physics from University of Wisconsin‐Madison, Madison, WI in 2004. He joined TomoTherapy Inc. the same year as a medical physicist, focusing on research and innovations. He joined University of Virginia in 2011 as an assistant professor in the radiation oncology department. His main research interests include tomotherapy, high‐performance computing, optimization and dose calculation, Monte Carlo, and innovative QA methods. He has authored and co‐authored over 50 papers in peer‐reviewed journals and holds seven granted patents. Dr. Nathan Childress (Mobius Medical Systems, LP) The goal of IMRT QA is to ensure each patient receives safe and effective treatment. Linear accelerator log files, when used intelligently, are the optimal means of performing IMRT QA. Not only can log file‐based IMRT QA verify information transfer integrity and delivery performance, it can do so more accurately and efficiently than conventional methods. Conventional IMRT QA methods of measuring the dose distribution in a plastic phantom are laborious, insensitive to some error types, and devoid of specificity. This leads some physicists to view IMRT QA as a tool to detect large machine calibration errors, whereas the true goal of IMRT QA is detecting errors specific to an individual patient's plan. Conventional IMRT QA produces a single, integrated result with limited pathway to identify and remedy error sources. Instead of fixing root causes, physicists end up either aimlessly repeating IMRT QA measurements until they pass, or they spend copious amounts of time investigating individual components. The lack of sophisticated methods may be one reason why approximately 20% of institutions fail RPC heterogeneous phantom irradiations at 7%/4 mm criteria (10) despite having acceptable conventional IMRT QA outcomes. Performing IMRT QA using log files offers several advantages. First, the delivered 3D dose can be calculated from log files in the patient CT and compared directly with the prescribed treatment plan. Not only does this comparison validate information transfer from the planning system, but it also allows for a comprehensive and quantitative assessment of the impact of delivery performance on the three‐dimensional dose in a patient. Second, log files are produced each time the plan is delivered, meaning the method can be utilized during patient deliveries, in addition to a single pretreatment measurement. This extends the current IMRT QA paradigm to include the entire course of delivery. Third, the entire process can be completely automated, yielding a detailed analysis of dose in patient's anatomy available within minutes of delivery. This reduces the workload required for IMRT QA on the medical physics team, and enables them to more efficiently focus their effort on evaluating the QA outcome rather than producing it. Finally, a system that employs an accurate, independent dose calculation algorithm to recalculate the dose in the patient CT using both the planned values and the data contained in the log files allows for a clear separation of errors (treatment plan versus treatment delivery), thus allowing physicists to fix root causes of problems. It is important to note that log file‐based IMRT QA must be augmented with independent commissioning measurements and a robust routine QA program (such as AAPM TG 142) to verify machine calibration. Once external systems are used to validate machine calibrations, the high temporal and spatial resolution of log files can identify patient‐specific errors and their sources almost automatically. Detection of certain errors in IMRT QA may then trigger additional machine‐specific tests. The use of log files for IMRT QA, when intelligently utilized as part of a robust QA program, can give physicists more time and information to analyze the clinical impact of detected errors and effectively mitigate them.