Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/ayush1997/Xvision
Chest Xray image analysis using Deep learning !
https://github.com/ayush1997/Xvision
chest-xray-images computer-vision deep-learning medical-imaging python tensorflow transfer-learning vggnet
Last synced: 7 days ago
JSON representation
Chest Xray image analysis using Deep learning !
- Host: GitHub
- URL: https://github.com/ayush1997/Xvision
- Owner: ayush1997
- Created: 2017-01-02T11:56:15.000Z (over 7 years ago)
- Default Branch: master
- Last Pushed: 2019-10-01T09:25:07.000Z (almost 5 years ago)
- Last Synced: 2024-03-07T19:31:45.819Z (4 months ago)
- Topics: chest-xray-images, computer-vision, deep-learning, medical-imaging, python, tensorflow, transfer-learning, vggnet
- Language: Python
- Homepage:
- Size: 7.29 MB
- Stars: 301
- Watchers: 28
- Forks: 129
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Lists
- my-awesome-stars - ayush1997/Xvision - Chest Xray image analysis using Deep learning ! (Python)
- Awesome-Medical-Imaging - X-ray Classification
README
# Xvision
Chest Xray image analysis using **Deep Learning** and exploiting **Deep Transfer Learning** technique for it with Tensorflow.
The **maxpool-5** layer of a pretrained **VGGNet-16(Deep Convolutional Neural Network)** model has been used as the feature extractor here and then further trained on a **2-layer Deep neural network** with **SGD optimizer** and **Batch Normalization** for classification of **Normal vs Nodular** Chest Xray Images.
## Nodular vs Normal Chest Xray
## Some specifications
| Property |Values |
| ------------- | ------------- |
| Pretrained Model | VggNet-16 |
| Optimizer used | stochastic gradient descent(SGD) |
| Learning rate | 0.01|
|Mini Batch Size| 20 |
| Epochs | 20 |
|2 Layers| 512x512 |
|GPU trained on| Nvidia GEFORCE 920M|
## Evaluation
### Confusion Matrix and Training Error Graph
| | **Normal** | **Nodule** |
|------|---------|---------|
| **Precision**| 0.7755102| 0.55555556 |
|**Recall**| 0.76 | 0.57692308 |
**Accuracy** : **69.3333 %**
## DataSet
[openi.nlm.nih.gov](https://openi.nlm.nih.gov/gridquery.php?q=&it=x,xg&sub=x&m=1&n=101) has a large base of Xray,MRI, CT scan images publically available.Specifically Chest Xray Images have been scraped, Normal and Nodule labbeled images are futher extrated for this task.
## How to use ?
The above code can be used for **Deep Transfer Learning** on any Image dataset to train using VggNet as the PreTrained network.
### Steps to follow
1. Download Data- the script download images and saves corresponding disease label in json format.
```python scraper.py ```
2. Follow the ```scraper/process.ipynb``` notebook for Data processing and generate
* Training images folder - All images for training
* Testing images Folder - All images for testing
* Training image labels file - Pickled file with training labels
* Testing image labels file - Pickled file with testing labels
3. Extract features(**CNN Codes**) from the **maxpool:5** layer of PreTrained CovNet(VggNet) and save them beforehand for faster training of Neural network.
```python train.py ```
* Train images codes folder - Path where training images codes will be stored
* Test images codes folder - Path where testing images codes will be stored
4. The extracted features are now used for training our **2-Layer Neural Network** from scratch.The computed models are saved as tensorflow checkpoint after every **Epoch**.
```python train_model.py ```
5. Finally the saved models are used for making predictions.Confusion Matrix is used as the Performance Metrics for this classifcation task.
```python test_model.py ```
## Some Predictions
## References
> 1. [Learning to Read Chest X-Rays: Recurrent Neural Cascade Model for Automated Image Annotation](https://arxiv.org/pdf/1603.08486.pdf)
> 2. [Deep Convolutional Neural Networks for Computer-Aided Detection: CNN Architectures,
Dataset Characteristics and Transfer Learning](https://arxiv.org/pdf/1602.03409.pdf)
## Contribute
If you want to contribute and add new feature feel free to send Pull request [here](https://github.com/ayush1997/Xvision/pulls) :D
To report any bugs or request new features, head over to the [Issues page](https://github.com/ayush1997/Xvision/issues)
## To-do
- [ ] Implement saliency map or use Deconv for better visualizations.