Keywords:
Performed at one institution, Diagnostic or prognostic study, Retrospective, Osteoporosis, Metabolic disorders, Demineralisation-Bone, Experimental investigations, Computer Applications-Detection, diagnosis, CAD, Digital radiography, Absorptiometry / Bone densitometry, Musculoskeletal bone, Computer applications, Artificial Intelligence, Information Technology
Authors:
A. Creeden; Coventry/UK
DOI:
10.26044/ecr2020/C-02138
Purpose
Osteoporosis is very common, affecting over two million people in the UK [1]. It is the largest risk factor for fragility fractures, which are associated with high morbidity and mortality as well as significant care costs for health systems.
Osteoporosis is usually identified using a fracture risk assessment tool, in association with Dual Energy X-ray Absorptiometry (DXA) for equivocal cases. However osteoporosis remains vastly underdiagnosed [2].
Changes in bone density can also be identified on standard radiographs. Twenty million radiographs were taken in the UK in 2016/17 [3] and it is possible that these radiographs might be leveraged to identify individuals with previously unsuspected osteoporosis. However on plain film radiographs up to 40% of skeletal calcium may be lost before osteoporosis can be visualised by observers [4].
Textural Analysis software can be used to identify textural changes on digital radiographs which are imperceptible to human observers. Textural analysis of pelvic radiographs has previously shown potential for identifying individuals at risk of osteoporosis under controlled conditions [5]. This study investigated whether textural analysis of pelvic radiographs obtained in routine clinical practice combined with machine learning might allow the DXA result of a patient to be predicted.