/bone_fracture_detection

Detection of Fractures from Xray Plates

Primary LanguageJupyter Notebook

bone_fracture_detection

Detection of Fractures from X-ray Plates using Deep CNN

Overview

This repository contains a Convolutional Neural Network (CNN) model for detecting fractures in X-ray images. The model is trained on the Fracture-Multi-Region-X-Ray-Data dataset, aiming to classify X-ray plates into two categories: fracture and no fracture.

Model Architecture

The model is built using a deep CNN architecture with the following layers:

Convolutional Layers:

5 Convolutional Layers with ReLU activation and Batch Normalization. MaxPooling layers to reduce spatial dimensions.

Fully Connected Layers:

2 Fully Connected Layers with dropout for regularization. Output layer for binary classification.

Training

Loss Function: Binary Cross Entropy with Logits (nn.BCEWithLogitsLoss)

Optimizer: Adam with learning rate of 0.0001

Epochs: 15 but also configurable in the training script

Performance Metrics

Accuracy: 96%

Precision: 0.97

Recall: 0.97

F1 Score: 0.97

These metrics indicate a highly effective model for identifying fractures in X-ray images.