PyTorch implementation of MICCAI 2018 paper "Liver Lesion Detection from Weakly-labeled Multi-phase CT Volumes with a Grouped Single Shot MultiBox Detector"

Last update: Oct 12, 2022

Overview

Grouped SSD (GSSD) for liver lesion detection from multi-phase CT

Note: the MICCAI 2018 paper only covers the multi-phase lesion detection part of this project. The liver segmentation with U-net is currently WIP.

Pixel-wise liver segmentation & multi-phase lesion detection from CT image in PyTorch.

This repo uses in-house industrial CT dataset, so the code for data loading is little nasty & unstructured, and I cannot release the dataset.

But structures of the liver segmentation data are nearly identical to 3Dircadb (http://www.ircad.fr/research/3dircadb/). And the lesion detection dataset looks like the figure 1 from our MICCAI 2018 paper.

train_liverseg_unet.py: U-net training for liver segmentation. Assumes 512x512 slices of dicom CT image and their corresponding binary segmentation masks.

ssd_liverdet/train_lesion_multiphase.py: GSSD training for liver lesion detection. Assumes 4-phase 512x512 slices of dicom CT image and their corresponding bounding box annotations.

The original SSD implementation is hard-forked from here.

PyTorch implementation of MICCAI 2018 paper "Liver Lesion Detection from Weakly-labeled Multi-phase CT Volumes with a Grouped Single Shot MultiBox Detector"

Related tags

Overview

Grouped SSD (GSSD) for liver lesion detection from multi-phase CT

Owner

Sang-gil Lee

Build a medical knowledge graph based on Unified Language Medical System (UMLS)

Link prediction using Multiple Order Local Information (MOLI)

Image morphing without reference points by applying warp maps and optimizing over them.

CondenseNet: Light weighted CNN for mobile devices

Prior-Guided Multi-View 3D Head Reconstruction

A deep learning framework for historical document image analysis

TEA: A Sequential Recommendation Framework via Temporally Evolving Aggregations

A Real-Time-Strategy game for Deep Learning research

PyTorch implementation of Algorithm 1 of "On the Anatomy of MCMC-Based Maximum Likelihood Learning of Energy-Based Models"

NumQMBasic - A mini-course offered to Undergrad physics students

Integrated Semantic and Phonetic Post-correction for Chinese Speech Recognition

Designing a Minimal Retrieve-and-Read System for Open-Domain Question Answering (NAACL 2021)

Code for DisCo: Remedy Self-supervised Learning on Lightweight Models with Distilled Contrastive Learning

Frequency Spectrum Augmentation Consistency for Domain Adaptive Object Detection

Code for "NeRS: Neural Reflectance Surfaces for Sparse-View 3D Reconstruction in the Wild," in NeurIPS 2021

Point detection through multi-instance deep heatmap regression for sutures in endoscopy

OMLT: Optimization and Machine Learning Toolkit

Catch-all collection of generative art made using processing

Fully Adaptive Bayesian Algorithm for Data Analysis (FABADA) is a new approach of noise reduction methods. In this repository is shown the package developed for this new method based on \citepaper.

Shared Attention for Multi-label Zero-shot Learning