Emory University (Molony & Samady)
Deep learning platform for fully automatic segmentation and phenotyping of coronary intravascular ultrasound (IVUS) pullbacks, packaged with a desktop GUI and CLI. A convolutional encoder-decoder network delineates the internal (lumen) and external elastic lamina borders on each cross-sectional IVUS frame; downstream rule-based analysis derives lumen area, plaque area, plaque burden, automatically flags lesions with plaque burden exceeding 40%, and reports minimum lumen area and maximum plaque burden along the pullback. Also supports end-diastolic gating and manual contour editing. Trained on 305 clinical IVUS pullbacks (270 train / 35 validation) from Philips and Boston Scientific catheters at Emory University; downstream evaluations have applied DeepIVUS to tasks such as automated detection of stent underexpansion.
Architecture
CNN (2D)
CNN-based encoder-decoder segmentation network delineating internal (lumen) and external elastic lamina borders per IVUS frame, followed by rule-based lesion/plaque-burden analysis
Framework
TensorFlow
Added to catalog
2026-08-12
Apache 2.0
License for model weights only. Associated code may be licensed seperately, check code source for specific terms.
Emory University IVUS Pullback Cohort (DeepIVUS)
305 clinical intravascular ultrasound pullbacks acquired with Philips Eagle Eye Platinum (20 MHz), Philips Revolution (45 MHz), or Boston Scientific OptiCross (40 MHz) catheters; split into 270 training and 35 validation pullbacks with expert-drawn internal/external elastic lamina contours.
Per-frame segmentation of the internal (lumen) and external elastic lamina borders on coronary IVUS images, used to derive lumen area, plaque area, plaque burden, and automated lesion (>=40% plaque burden) flagging