Official website for "Video Polyp Segmentation: A Deep Learning Perspective (MIR 2022)"

TransRUPNet for Improved Out-of-Distribution Generalization in Polyp Segmentation (IEEE EMBC)

CT Liver Segmentation Via PVT-based Encoding and Refined Decoding (2024 IEEE ISBI)

Official implementation of TransNetR: Transformer-based Residual Network for Polyp Segmentation with Multi-Center Out-of-Distribution Testing (MIDL 2022)

Polyp segmentation tool utilizing U-Net for accurate medical image analysis, designed to enhance early detection and diagnosis of colorectal cancer. Features a user-friendly Streamlit web app for easy image processing and analysis, leveraging the Kvasir-SEG dataset for improved healthcare outcomes.

[WACV 2024] An implementation of MEGANet for polyp segmentation with multi-scale edge-guided attention

This research will show an innovative method useful in the segmentation of polyps during the screening phases of colonoscopies. To do this we have adopted a new approach which consists in merging the hybrid semantic network (HSNet) architecture model with the Reagion-wise(RW) as a loss function for the backpropagation process.

Polyp recognition and segmentation for colonoscopy images using UNet++ model.

S2ME: Spatial-Spectral Mutual Teaching and Ensemble Learning for Scribble-supervised Polyp Segmentation (MICCAI 2023)

Using DUCK-Net for polyp image segmentation. ( Nature Scientific Reports 2023 )

PyTorch implementation of ResUNet++ for Medical Image segmentation

PyTorch implementation of DoubleUNet for medical image segmentation

Official implementation of DoubleU-Net: A Deep Convolutional Neural Network for Medical Image Segmentation (pytorch implementation)

A multi-centre polyp detection and segmentation dataset for generalisability assessment https://www.nature.com/articles/s41597-023-01981-y

Official implementation of NanoNet: Real-time medical Image segmentation architecture (IEEE CBMS)

Kvasir-SEG: A Segmented Polyp Dataset

Epistemic uncertainty, sometimes referred to as model uncertainty, describes what the model does not know because training data was not appropriate. Modelling epistemic uncertainty is crucial to prevent ill advised discussion making due to over confident models.

Polyp-SAM++ is the first text-guided polyp-segmentation method using segment anything model (SAM).

PraNet: Parallel Reverse Attention Network for Polyp Segmentation, MICCAI 2020 (Oral). Code using Jittor Framework is available.

This projects uses video feeds from endoscopic procedures to identify polyps in the gastrointestinal tract and draw masks around them to aid doctors in identifying precursors of colorectal cancer.