HookNet

This is my personal implementation(with Torch) of HookNet. I also summarize HookNet paper in my blog.

HookNet Structure

We propose HookNet, a semantic segmentation model for histopathology whole-slide images, which combines context and details via multiple branches of encoder-decoder convolutional neural networks. Concentric patches at multiple resolutions with different fields of view are used to feed different branches of HookNet, and intermediate representations are combined via a hooking mechanism. - Mart van Rijthoven. (2020)

Environment

pip install -r requirements.txt

Above, I install python 3.6 with CUDA 11.4

Description

Repository Structure

model/hooknet.py: main HookNet [input_image_shape :: (284x284x3) x 2, input_mask_shape :: (70x70) x 1]
model/hooknet_se_resnext101_32x4d.py: main HookNet_SE-ResNeXt101_32x4d model script [input_image_shape :: (512x512x3) x 2, input_mask_shape :: (512x512) x 2]
datagen.py: the data dataloader and augmentation script
functional.py: naming a weight of model and converting outputs to images script
train.py: main training script

Training

Data Preparation

Hooknet
    ├ slide_num_1
    |       ├ input_x100
    |       ├ input_x50
    |       ├ input_x25
    |       ├ input_x12
    |       └ input_y1
    .
    .
    .
    └ slide_num_n
            ├ input_x100
            ├ input_x50
            ├ input_x25
            ├ input_x12
            └ input_y1

if you want to train with hooknet(normal) or quad_scale_hooknet... (This model trains with valid padding.)

input_x1: mpp=1 image patches(284x284x3) directory
input_x2: mpp=2 image patches(284x284x3) directory
input_x4: mpp=4 image patches(284x284x3) directory
input_x8: mpp=8 image patches(284x284x3) directory
input_y1: mpp=1 mask patches(70x70) directory

if you want to train with hooknet(se_resnext101_32x4d)... (This model trains with same padding.)

input_x1: mpp=1 image patches(512x512x3) directory
input_x2: mpp=2 image patches(512x512x3) directory
input_x4: mpp=4 image patches(512x512x3) directory
input_x8: mpp=8 image patches(512x512x3) directory
input_y1: mpp=1 mask patches(512x512) directory

You can get this data structure by using util_multi.py

Train Example

python train.py

You can adjust hyper parameters in config.py

gotjd709 / hooknet

readme