TY - JOUR T1 - Technologist Based Implementation of Total Metabolic Tumor Volume into Clinical Practice JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 176 LP - 176 VL - 62 IS - supplement 1 AU - Jaiden Sullivan AU - Claire Olson AU - Matthew Thorpe AU - Geoffrey Johnson AU - Manoj Jain AU - Jason Young Y1 - 2021/05/01 UR - http://jnm.snmjournals.org/content/62/supplement_1/176.abstract N2 - 176Introduction: Many methods are used to measure metabolic tumor volume (MTV) via positron emission tomography (PET) (1). These methods may help predict patient outcome and treatment response, especially with diffuse large B-cell lymphoma (DLBCL) (2-5). However, there are challenges with implementing MTV into standard clinical practice given various methods and additional time required to segment tumors. We aim to explore integrating total MTV into the clinical setting using a semi-automated workflow. Methods: An exploratory project design was implemented. Two Nuclear Medicine Technologist Students (NMTS) received a 30-minute training session on distinguishing physiologic from pathologic lymphomatous FDG uptake from a board-certified Nuclear Radiologist. The NMTS had 8-12 hours experience observing the technologist-side of a clinical PET practice (~85 exams per day) yet no other PET education or image interpretation experience. In tandem, the NMTS then independently evaluated 20 FDG PET/CT exams on 10 patients with DLBCL: 10 before (Pre) and 10 after (Post) systemic therapy. The images were reviewed by the NMTS and a Radiologist separately using a “LesionID'' workflow (MIM Inc. Cleveland OH, USA) which automatically segments PET lesions based on an absolute SUVmax threshold. The threshold was set using the SUVmax within a 3 cm spherical region of normal liver. The workflow then segmented everything with SUVmax greater than the threshold. The NMTS and Radiologist separately evaluated each segmentation to distinguish lymphoma from erroneous segmentation. The erroneous segments were deleted (after noting anatomic location) and the total MTV was calculated. If the NMTS were unsure about a lesion being erroneous, it remained to be included in the MTV calculation. Throughout this process, observations around the pearls and pitfalls of technologist-based MTV calculation were recorded. Results: The mean Radiologist-derived MTV values were 446.0 mL (± 555.6) and 38.5 mL (± 77.6) for Pre and Post exams respectively. The mean MTV of NMTS were 414.8 mL (± 597.6) and 27.7 mL (± 57.3) for Pre and Post exams respectively. The mean Radiologist MTV values were 7.5% and 28.0% higher than the NMTS values for the Pre and Post exams respectively. There were 2/10 patients with critical missed segments by the NMTS; a mediastinal mass that was perceived as physiologic heart and a scalp lesion mistaken for physiologic brain. The mean Radiologist number of segments removed was 20.6 (range 8-28) and 18.3 (range 9-41) for the Pre and Post therapy exams respectively. The mean NMTS number of segments removed was 19.2 (range 7-33) and 17.8 (range 7-49) for the Pre and Post therapy exams respectively. Common erroneous sites of segmentation included salivary glands, liver, bowel, kidney, ureter, urinary bladder, muscle, and brain. We uncovered the need for (and helped develop) a two-part MTV calculation workflow: 1) Technologists set a threshold, remove obvious erroneous segments and flag uncertain segments. 2) Radiologist reviews and finalizes the segmentation for MTV calculation. Once final, the Technologist could then review the final segmentation for continuous feedback and quality improvement. Conclusions: Although more data is needed using a larger set of exams, NMTS with little experience may calculate total MTV with relatively favorable accuracy. We suggest Nuclear Medicine Technologists may play a role in calculating total MTV, especially in patients with lymphoma. Capturing revenue from post-processing of images by a Technologist would help with total MTV clinical implementation. However, to the best of our knowledge, there are no Nuclear Medicine service codes for image post-processing in the United States of America (6). Nuclear Medicine service codes for image post-processing are needed to help advance patient care, not only for total MTV calculation but for Nuclear Medicine therapy dosimetry and to help build ground-truth data for artificial intelligence. ER -