YOLOv9 for Fracture Detection

YOLOv9 for Fracture Detection in Pediatric Wrist Trauma X-ray Images

Comparison

Performance

Model	Test Size	Param.	FLOPs	F1 Score	AP₅₀^val	AP_50-95^val	Speed
YOLOv8	640	43.61M	164.9G	0.59	62.44%	40.32%	3.6ms
YOLOv8+SA	640	43.64M	165.4G	0.62	63.99%	41.49%	3.9ms
YOLOv8+ECA	640	43.64M	165.5G	0.61	62.64%	40.21%	3.6ms
YOLOv8+GAM	640	49.29M	183.5G	0.60	63.32%	40.74%	8.7ms
YOLOv8+ResGAM	640	49.29M	183.5G	0.62	63.97%	41.18%	9.4ms
YOLOv8+ResCBAM	640	53.87M	196.2G	0.62	62.95%	40.10%	4.1ms
YOLOv9-C	640	51.02M	239.0G	0.64	65.31%	42.66%	5.2ms
YOLOv9-E	640	69.42M	244.9G	0.64	65.46%	43.32%	6.4ms

Citation

If you find our paper useful in your research, please consider citing:

@article{chien2024yolov9,
  title={YOLOv9 for fracture detection in pediatric wrist trauma X-ray images},
  author={Chien, Chun-Tse and Ju, Rui-Yang and Chou, Kuang-Yi and Chiang, Jen-Shiun},
  journal={Electronics Letters},
  volume={60},
  number={11},
  pages={e13248},
  year={2024},
  publisher={Wiley Online Library}
}

Requirements

Linux (Ubuntu)
Python = 3.9
Pytorch = 1.13.1
NVIDIA GPU + CUDA CuDNN

Environment

  pip install -r requirements.txt

Overall Flowchart

Dataset Split

GRAZPEDWRI-DX Dataset (Download Link)
Download dataset and put images and annotatation into ./GRAZPEDWRI-DX_dataset/data/images, ./GRAZPEDWRI-DX_dataset/data/labels.
```
  python split.py
```
The dataset is divided into training, validation, and testing set (70-20-10 %) according to the key patient_id stored in dataset.csv.

The script then will move the files into the relative folder as it is represented here below.

 GRAZPEDWRI-DX_dataset
    └── data   
         ├── images
         │    ├── train
         │    │    ├── train_img1.png
         │    │    └── ...
         │    ├── valid
         │    │    ├── valid_img1.png
         │    │    └── ...
         │    └── test
         │         ├── test_img1.png
         │         └── ...
         └── labels
              ├── train
              │    ├── train_annotation1.txt
              │    └── ...
              ├── valid
              │    ├── valid_annotation1.txt
              │    └── ...
              └── test
                   ├── test_annotation1.txt
                   └── ...

The script will create 3 files: train_data.csv, valid_data.csv, and test_data.csv with the same structure of dataset.csv.

Data Augmentation

Data augmentation of the training set using the addWeighted function doubles the size of the training set.

  python imgaug.py --input_img /path/to/input/train/ --output_img /path/to/output/train/ --input_label /path/to/input/labels/ --output_label /path/to/output/labels/

For example:

  python imgaug.py --input_img ./GRAZPEDWRI-DX/data/images/train/ --output_img ./GRAZPEDWRI-DX/data/images/train_aug/ --input_label ./GRAZPEDWRI-DX/data/labels/train/ --output_label ./GRAZPEDWRI-DX/data/labels/train_aug/

The path of the processed file is shown below:

 GRAZPEDWRI-DX_dataset
    └── data   
         ├── images
         │    ├── train
         │    │    ├── train_img1.png
         │    │    └── ...
         │    ├── train_aug
         │    │    ├── train_aug_img1.png
         │    │    └── ...
         │    ├── valid
         │    │    ├── valid_img1.png
         │    │    └── ...
         │    └── test
         │         ├── test_img1.png
         │         └── ...
         └── labels
              ├── train
              │    ├── train_annotation1.txt
              │    └── ...
              ├── train_aug
              │    ├── train_aug_annotation1.txt
              │    └── ...
              ├── valid
              │    ├── valid_annotation1.txt
              │    └── ...
              └── test
                   ├── test_annotation1.txt
                   └── ...

Weights

If you plan to use pretrained models to train, you need put them into ./weights/.

You can get the YOLOv9 pretained models on the MS COCO 2017 Dataset through YOLOv9 official GitHub.
Use gdown to download the trained model from our GitHub:

  gdown https://github.com/RuiyangJu/YOLOv9-Fracture-Detection/releases/download/Trained/weights.zip

Train & Validate

Before training the model, make sure the path to the data in the ./data/meta.yaml file is correct.

meta.yaml

  # patch: /path/to/GRAZPEDWRI-DX/data
  path: 'E:/GRAZPEDWRI-DX/data'
  train: 'images/train_aug'
  val: 'images/valid'
  test: 'images/test'

Arguments

Key	Value	Description
workers	8	number of worker threads for data loading (per RANK if DDP)
device	None	device to run on, i.e. device=0,1,2,3 or device=cpu
model	None	path to model file, i.e. yolov8n.pt, yolov8n.yaml
batch	16	number of images per batch (-1 for AutoBatch)
data	None	path to data file, i.e. coco128.yaml
img	640	size of input images as integer, i.e. 640, 1024
cfg	yolo.yaml	path to model.yaml, i.e. yolov9-c.yaml
weights	None	initial weights path
name	exp	save to project/name
hyp	data/hyps/hyp.scratch-high.yaml	hyperparameters path
epochs	100	number of epochs to train for

Example

  python train_dual.py --workers 8 --device 0 --batch 16 --data data/meta.yaml --img 640 --cfg models/detect/yolov9-c.yaml --weights weights/yolov9-c.pt --name yolov9-c --hyp hyp.scratch-high.yaml --min-items 0 --epochs 100 --close-mosaic 15

Related Works

Expand