An Enquiry into Application-Based Diagnostic Support System for Distal Humerus Fracture using Deep Learning

Abstract

Abstract Background: Recently, the development of deep learning has been remarkable and the accuracy of the image recognition has improved. Distal humerus fracture treatment is a complex injury requiring good diagnostic judgment and management protocol. There have been no attempts toward development of artificial intelligence (AI) in treating distal humerus. Objectives: To find out whether deep learning (AI) gives reasonable image-based recommendation for the surgical procedure of distal end humerus fractures. Methods: The image of the distal end humerus will be sent to application (app)-based deep learning software. The result of the deep learning software diagnosis and intervention will be compared with two senior orthopedic surgeons’ opinion. Intra- and inter-observer reliability will be studied, and Student’s t-test will be used for identifying the P value of the objective. Expected Outcome: We will identify the layers required for the deep learning software to process the distal humerus fractures and reach a conclusion. The app will help in identification and planning the management of the fracture. Conclusion: The Application based diagnostic support system for the distal humerus fracture is an initial investigation for the usage of AI in peripheral centres and can provided further diagnostic support.