Idea Name: |
MURA Classification |
Slogan: |
Abnormality detection in musculoskeletal radiograph |
Supervisor Name: |
Muhmmad Bilal Khan |
Supervisor Designation: |
Lecturer |
Supervisor School: |
SEECS |
Supervisor Department: |
Faculty of Information Technology (FOIT) |
Contact number: |
03214269282 |
Email ID: |
[email protected] |
Abstract: |
Musculoskeletal conditions will increase as the population of the world ages. For radiologists to obtain second opinions and to facilitate cases where radiologists are not available, an automated system capable of analyzing x-ray images is necessary.Using deep learning techniques and the MURA dataset, this project intends to serve this need by binary classifying X-ray images as normal or abnormal. |
What is the unmet need in society that your idea will fulfill ? |
Musculoskeletal conditions affect more than 1.7 billion people worldwide, with 30 million emergency department visits annually and increasing. The shortage of radiologists, as well as the increase in the number of musculoskeletal conditions, necessitate the development of an automated system . |
Who needs it ? How many would benefit ? |
It is useful for radiologists and medical professionals who need to accurately diagnose and assess musculoskeletal conditions. It is also useful for researchers and developers in the medical imaging field who are looking for a reliable and accurate way to classify medical images. |
How will the solution works |
The acquired Mura dataset will help in Convolutional Neural Network (CNN) model training. Once a final Convolutional Neural Network (CNN) model with better performance is ready, it will be deployed into an application that will enable a user to input an x-ray image and receive a prediction. The x-ray image must be one of the seven types of bone included in the Mura Dataset. Additionally, the model will be of the Federated Learning type. |
Who are your competitors ? How is your solution different |
Our competitors are researchers such as the Stanford Machine Learning Group, who have attempted Mura Classification using different Artificial Intelligence models. We intend to build a model to achieve better accuracy and further, create an application to provide this facility. |
Status: |
new |
Entry Date & Time: |
2022-12-30 (1804) |