Skip to content

Latest commit

 

History

History
14 lines (9 loc) · 461 Bytes

File metadata and controls

14 lines (9 loc) · 461 Bytes

Deep-learning-for-automatic-diagnosis-of-lung-diseases

This project consists of detecting lung diseases from 3D CT Scans. Different architectures of Neural Networks have been trained and tested to make the predictions.

The architectures developed include :

  • RNN (LSTM, GRU)
  • 2D CNN + LSTM
  • Convolutional LSTM
  • 3D CNN

The best model has an accuracy of 96%, based on a test set of 200 CT Scans.

Technologies : Python, TensorFlow, Keras, OpenCV