JJEM: Special Edition 2 (December 2024)
JJEM: Special Edition 2 (December 2024)
2024-12-08
A Novel Method for Detecting Bone Fractures via Deep Supervised Learning from Radiological Pictures.
Sadiya Fathima, Dr. Raghavendra S P
The bone fracture detection project uses deep learning techniques with the focus on employing or using the CNN architecture InceptionV3 to detect and determine the type and existence of different bone fractures from the X-ray images of the limbs. Using a total of 1440 samples over 12 types of fractures and utilizing all the tests the system accurately identified them with a success rate of up to 98%. The model is carefully trained with TensorFlow and Keras frameworks and to make sure the model robust, the image data augmentation techniques combined with the dataset. One of the key characteristics of the current work is the connection to a simple Flask web GUI that enables simple user-interaction in the form of an X-ray image upload for subsequent fracture identification. The application maintains a secure and effective mechanism of managing the users and the information they input behind the help of the MySQL database for users authentication. This solution not only helps save a considerable amount of time for diagnosing and can decrease the probability of errors, but also tries to assist the medical staff in providing the patients with efficient and adequate treatment which can raise the success rates
Deep Learning, CNN, Image Processing.