Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images

Manual delineation of liver segments on computed tomography (CT) images for primary/secondary liver cancer (LC) patients is time-intensive and prone to inter/intra-observer variability. Therefore, we developed a deep-learning-based model to auto-contour liver segments and spleen on contrast-enhanced CT (CECT) images. We trained two models using 3d patch-based attention U-Net (MpaU-Net) and 3d full resolution of nnU-Net (MnnU-Net) to determine the best architecture (BA) . BA was used with vessels ( M-Vess) and spleen (Mseg+spleen) to assess the impact on segment contouring. Models were trained, validated, and tested on 160 (C-RTTrain), 40 (C-RTVal), 33 (C-LS), 25 (C-CH) and 20 (C-PVE) CECT of LC patients. MnnU-Net outperformed MpaU-Net across all segments with median differences in Dice similarity coefficients (DSC) ranging 0.03-0.05 (p < 0.05). Mseg+spleen , and MnnU-Net were not statistically different (p > 0.05), however, both were slightly better than M-Vess by DSC up to 0.02. The final model, Mseg+spleen, showed a mean DSC of 0.89, 0.82, 0.88, 0.87, 0.96, and 0.95 for segments 1, 2, 3, 4, 5-8, and spleen, respectively on entire test sets. Qualitatively, more than 85% of cases showed a Likert score >= 3 on test sets. Our final model provides clinically acceptable contours of liver segments and spleen which are usable in treatment planning.