Bone Imaging Statistical Learning

REC REF:  18/LO/1746

Active and recruiting.

The aim of this study is to develop automated data driven software to evaluate a large dataset of medical images from varying populations with different clinical findings to aid in the diagnosis of disorders in joint extremities. Furthermore, the team aims to develop the software model as a learning tool for junior doctors, as well as an assisting tool for pre-surgical planning.

STUDY OBJECTIVE

The objective of the study is to build a data driven software algorithm, based on statistical learning methods and machine learning, that can automatically locate features, analyse bone, joint, and tissue health, and predict clinical outcomes from a medical image.

OVERVIEW

This study involves collecting retrospective data from 2200 patients who have had medical image examinations as part of a diagnosis or care referred by a consultant or local GP. The chief investigator and a member of the study team will identify, anonymise, and harvest the datasets to develop the software model as a learning tool for junior doctors as well as an assisting tool for pre-surgical planning.