Research case study

Explainable Pancreatic CT Segmentation

Current research into clinically meaningful 3D segmentation using deep learning and Vision Transformers.

Problem

Pancreatic anatomy is difficult to delineate reliably in volumetric CT, while clinical adoption requires interpretable evidence rather than prediction alone.

Approach

Data: TODO — publishable dataset details
Model: deep learning and ViT-based 3D segmentation approaches
XAI method: TODO — method under development
Validation: model development and evaluation; metrics pending publishable results

Results

TODO: Dice score TODO: clinical validation TODO: model comparison

Figures

Axial medical scan representing pancreatic CT analysis and preprocessing — Pancreatic CT preprocessing view for the current research workflow. demo

Three-dimensional medical segmentation rendering illustrating volumetric model output — Three-dimensional segmentation rendering illustrating volumetric model output. demo

Vision Transformer architecture diagram for image classification and medical segmentation — Vision Transformer architecture overview for image-classification experiments. demo