Comparison of a SegNet-based Algorithm Quantitatively Estimating Epifascial Fibrosis in Three-dimensional Computed Tomography Images to the Clinical Lymphedema Grading Method

In lymphedema, proinflammatory cytokine-mediated progressive cascades always occur, leading to macroscopic fibrosis. However, no methods are practically available for measuring lymphedema-induced fibrosis before its deterioration. Technically, CT can visualize fibrosis in superficial and deep locations. For standardized measurement, verification of deep learning (DL)-based recognition was performed. A cross-sectional, observational cohort trial was conducted at a teaching university hospital. The protocol of this study was approved by the University Hospital Institutional Review Board and was registered at the Protocol Registration and Results System (PRS), www. clini caltr ials. gov (NCT04811677: https:// clini caltr ials. gov/ ct2/ show/ NCT04 811677? term= NCT04 81167 7& draw= 2& rank=1). All methods were performed in accordance with the relevant guidelines and regulations. The trial conformed to the tenets of the Declaration of Helsinki. Patients were included if they were clinically diagnosed with unilateral limb lymphedema and had undergone BEI analysis and CT scanning. The subjects provided written informed consent for publication of the case details. Data were collected as close to the CT scanning date as possible. Patients who were diagnosed with deep vein thrombosis, bilateral limb involvement, vascular disease, or local infection were excluded.

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 fibrosis) was trained (65%), validated (15%), and tested (20%). Then, 4 indices were formulated and compared with the standardized circumference difference ratio (SCDR) and bioelectrical impedance (BEI) results. In total, 2138 CT images of 27 chronic unilateral lymphedema patients were analyzed.