A 3D Slicer extension for estimating airway stenosis index from bronchoscopic images using cGAN-based depth estimation.
This project provides tools to quantify airway stenosis by comparing depth maps generated from expiration and inspiration bronchoscopic images. A pre-trained CycleGAN generator estimates depth from grayscale bronchoscopic frames, and a configurable threshold produces binary airway segmentations used to compute the Stenosis Index (SI).
v1.0 — Original Module
The initial implementation. Supports loading bronchoscopic image files, ground truth SI calculation from CT segmentation, and estimated SI from depth maps with manual marker placement.
v2.0 — Rewritten Module
A rewrite with live bronchoscope support. Features real-time depth estimation from a video feed via OpenIGTLink, snapshot capture, interactive threshold sliders, and an auto-updating stenosis index display.
This project is licensed under the BSD 3-Clause License.