BRNet

code for "Automated assessment of BI-RADS categories for ultrasound images using multi-scale neural networks with an order-constrained loss function"

Requirements

• PyTorch 1.0+
• Python 3.x

Running

train:

python train.py --config cfgs/brnet.yaml --net inception_v3_brnet --model mixup_model --gpu 0

eval:

python eval.py --config cfgs/brnet.yaml --net inception_v3_brnet --model mixup_model --gpu 0 --save_name InceptionV3-BRNet_trail_1 --ckpt_path path-to-checkpoint.pth

Data

The data folder contains a portion of our dataset.