Deep learning-based quantitative estimation of lymphedema-induced fibrosis using three-dimensional computed tomography images

Abstract

In lymphedema, proinfammatory cytokine-mediated progressive cascades always occur, leading to macroscopic fbrosis. However, no methods are practically available for measuring lymphedemainduced fbrosis before its deterioration. Technically, CT can visualize fbrosis in superfcial and deep locations. For standardized measurement, verifcation of deep learning (DL)-based recognition was performed. A cross-sectional, observational cohort trial was conducted. After narrowing window width of the absorptive values in CT images, SegNet-based semantic segmentation model of every pixel into 5 classes (air, skin, muscle/water, fat, and fbrosis) was trained (65%), validated (15%), and tested (20%). Then, 4 indices were formulated and compared with the standardized circumference diference ratio (SCDR) and bioelectrical impedance (BEI) results. In total, 2138 CT images of 27 chronic unilateral lymphedema patients were analyzed. Regarding fbrosis segmentation, the mean boundary F1 score and accuracy were 0.868 and 0.776, respectively. Among 19 subindices of the 4 indices, 73.7% were correlated with the BEI (partial correlation coefcient: 0.420–0.875), and 13.2% were correlated with the SCDR (0.406–0.460). The mean subindex of Index 2 (PFibrosis in Affected−PFibrosis in Unaffected/PLimb in Unaffected)presented the highest correlation. DL has potential applications in CT image-based lymphedema-induced fbrosis recognition. The subtraction-type formula might be the most promising estimation method.