Peritumoral-Aware MIXture learning (PAMIX)

🔥🔥🔥

Updated on April 30 , 2025

✨Paper

This repository provides the official implementation of Multi-Level Peritumoral-Aware Mixture Learning for Breast Tumor Diagnosis

✨Dataset

Organ	Images	Categories	Links
BUS-BRA	1875	Malignant & Benign	BUS-BRA
BUSI	647 (exclude 133 normal)	Malignant & Benign	BUSI
QAMEBI	232	Malignant & Benign	QAMEBI

✨Model & Weights

Models	Weights for BUS-BRA	Weights for BUSI	Weights for QAMEBI
PAMIX	Weights	Weights	Weights
Uni-PAMIX	Weights	Weights	Weights

✨Installation & Preliminary

1. Clone the repository.

   git clone https://github.com/Git-HB-CHEN/PAMIX-Family.git
   cd PAMIX-Family
   

2. Create a virtual environment and activate the environment.

   conda create -n PAMIX-Family python=3.8
   conda activate PAMIX-Family
   

3. Install Pytorch == 1.13.0+cu117, and torchvision==0.14.0+cu117. (You can follow the instructions here)

4. Install other dependencies.

   pip install -r requirements.txt
   

✨Inference using PAMIX & Uni-PAMIX

1. Download the Weights of the PAMIX-Family

2. Place your images in the examples folder

3. Infer your images with the PAMIX

   python infer_PAMIX.py

4. Infer your images with the Uni-PAMIX

   python infer_Uni-PAMIX.py
   

The relevant codes and weights are under preparation and will be made available soon.