Image Segmentation Model

Installation

git clone https://github.com/quangdungluong/semantic-segmentation-model.git
cd semantic-segmentation-model
pip install -r requirements.txt

Flask App demo

mkdir brain_mri/model_ckpt
# download the weight then move to brain_mri/model_ckpt directory
python app.py

• Weight: Google Drive

Usage

from model import *
from trainer import *
from metrics import *
import torch

model = AttU_Net(img_ch=3, output_ch=1)
optimizer = torch.optim.Adam(model.parameters(), lr=0.0001)
scheduler = torch.optim.lr_scheduler.ReduceLROnPlateau(optimizer, 'max', patience=3, factor=0.5)
measures = {'dice_coef':dice_coef_metric,
           'iou':iou,
           'precision':precision,
           'recall':recall,
           'fscore':fscore}

train_log, val_log = train_model("R2UNet", model, dataloader, bce_dice_loss, optimizer, scheduler, measures, num_epochs)

Dataset 1

ISIC 2018 Challenge Task 1: Lesion Boundary Segmentation

The dataset (training and validation) was split into three subsets, training set, validation set, and test set (included original validation set), which the proportion is 80%, 10% and 10% of the whole dataset, respectively. The entire dataset contains 2694 images where 2075 images were used for training, 259 for validation and 360 for testing models.

Result

Method	ResNeXt50-UNet	Attention UNet
Dice coefficient	0.901	0.850
IoU	0.822	0.743
Precision	0.894	0.865
Recall	0.913	0.841
F1-Score	0.901	0.850

In the combined image:

• Purple: True Positive (model prediction matches lesion area marked by human)
• Blue: False Negative (model missed part of lesion in the area)
• Green: False Positive (model incorrectly predicted lesion in an area where there is none)

Dataset 2

Brain MRI Segmentation LGG Segmentation Dataset

• This dataset contains brain MR images together with manual FLAIR abnormality segmentation masks.
• The images were obtained from The Cancer Imaging Archive (TCIA).
• They correspond to 110 patients included in The Cancer Genome Atlas (TCGA) lower-grade glioma collection with at least fluid-attenuated inversion recovery (FLAIR) sequence and genomic cluster data available.

Result on test set

Method	UNet	ResNeXt50-UNet	Attention UNet
Dice coefficient	0.848	0.819	0.877
IoU	0.742	0.737	0.788
Precision	0.852	0.801	0.872
Recall	0.850	0.816	0.892
F1-Score	0.847	0.793	0.877

The project is still in progress...