Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2025
- Goals
- Record
- 92ca89f8-5772-469d-8a55-3f5e5f64f835
- Published
- 2026-10-30
Universitas Dian Nuswantoro (UDINUS), through the participation of Dewi, Ika Novita; Salam, Abu; and Utomo, Danang Wahyu, conducted research on liver and tumor segmentation from medical imaging. The activity focused on a proposed U-Net architecture enhanced with image colorization techniques for precise liver tumor segmentation in clinical CT images. Accurate segmentation is described as important for effective diagnosis and appropriate treatment planning, especially in cases of liver cancer.
The proposed image colorization-based U-Net integrated grayscale-based and RGB-based architectures. The architectures were tested on the LiTS dataset and real clinical data. The evaluation measured effectiveness in liver and tumor segmentation across different imaging conditions. The grayscale-based U-Net achieved a DICE coefficient of 99.95% for liver segmentation and 90.44% for tumor segmentation. The model also achieved an RMSE of 0.019, a PSNR of 82.14, and a pixel accuracy of 0.316. These reported results reflect the model's performance in segmentation, reconstruction, and image quality measures.
The model was additionally tested on clinical data, where it effectively segmented liver and tumor regions across diverse imaging conditions. The research addressed low contrast and variability in tumor characteristics through grayscale-based colorization techniques intended to enhance feature representation. The findings demonstrate the potential of the proposed approach to enhance liver and tumor localization and provide a framework for clinical applications. This UDINUS activity presents health-related research using medical imaging and a colorization-based U-Net approach.