NIH Research Festival
FARE Award Winner
Pheochromocytomas and paragangliomas (PPGLs) are respectively intra-adrenal and extra-adrenal neuroendocrine tumors whose pathogenesis and progression are greatly regulated by genetics. Identifying PPGL‚Äôs genetic clusters (SDHx, VHL/EPAS1, kinase signaling, and sporadic) is essential as PPGL's management varies critically on its genotype. Genetic testing for PPGLs is expensive and time-consuming. Contrast-enhanced CT (CE-CT) scans of PPGL patients are acquired at the beginning of patient management for PPGL staging and determining the next therapeutic steps. Given a CE-CT sub-image of the PPGL, this work demonstrates a two-branch vision transformer (PPGL-Transformer) to identify each tumor‚Äôs genetic cluster. The standard of reference for each tumor included two items: its genetic cluster from clinical testing, and its anatomical location. One branch of the PPGL-Transformer identifies PPGL's anatomic location while the other one characterizes PPGL's genetic type. A supervised contrastive learning strategy was used to train the PPGL-Transformer by optimizing contrastive and classification losses for PPGLs‚Äô genetic group and anatomic location. Our method was evaluated on a dataset comprised of 1010 PPGLs extracted from the CE-CT images of 289 patients. PPGL-Transformer achieved an accuracy of 0.63 ¬± 0.08 and balanced accuracy (BA) of 0.63 ¬± 0.06 on five-fold cross-validation and outperformed competing methods by 2-29% on accuracy and 3-18% on BA. The performance for the sporadic cluster was higher on BA (0.68 ¬± 0.13) while the performance for the SDHx cluster was higher on recall (0.83 ¬± 0.06). The proposed method may lead to faster and more widely available PPGLs‚Äô genetic characterization offering PPGLs‚Äô timely personalized management.
Scientific Focus Area: Clinical Research
This page was last updated on Monday, September 25, 2023