submission_1127
About
Summary
Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive malignancy with poor prognosis, underscoring the critical need for early and accurate diagnosis. We present a novel two-stage pipeline for PDAC detection from contrast-enhanced computed tomography (CECT) scans, leveraging the nnU-Net framework for segmentation and a highresolution multi-task convolutional neural network (CNN) for joint lesion segmentation and classification. The first stage employs a low-resolution nnU-Net to segment the pancreas region, while the second stage refines segmentation and performs lesion classification.
Mechanism
This algorithm produces a likelihood score for the presence of pancreatic ductal adenocarcinoma (PDAC) in an input venous-phase contrast-enhanced computed tomography scan (CECT), together with a detection map with the lesion localization. we introduce a two-stage pipeline for pancreas region of interest (ROI) localization followed by joint cancer segmentation and classification. The localization is based on a low-resolution nnU-Net model and we further extend it to multitask network by adding a classification head. The algorithm was trained with 2238 images from the PANORAMA public development data set.
For more details: https://github.com/bowang-lab/PANORAMA/blob/main/Panorama_report.pdf
Interfaces
This algorithm implements all of the following input-output combinations:
| Inputs | Outputs | |
|---|---|---|
| 1 |
Validation and Performance
Challenge Performance
| Date | Challenge | Phase | Rank |
|---|---|---|---|
| Jan. 23, 2025 | PANORAMA | Development Phase - Tuning | 6 |
| Feb. 25, 2025 | PANORAMA | Testing Phase | 4 |
Uses and Directions
This algorithm was developed for research purposes only.