Radiologist Diagnostic Performance Versus Artificial Intelligence (AI)

University Hospital, Montpellier1,000 enrolled

Overview

In France, femoral neck fracture is mainly detected with interpretation of pelvis/hip X-ray imaging (French Health Authority recommandation). However, up to 10% of fractures are not identified or misdiagnosed, especially in patients admitted to the emergency department. Indeed, radiologists may be subject to excessive work, wich cause the risk of inaccurate on X-rays diagnosis. The Artificial intelligence (AI) begins study the detection of fratures on medical imaging. In this retropective study, this technology developed by GLEAMER company is tested to evaluate the detection rate of hip fracture and specifically femoral neck fracture, compared to the radiologist diagnostic, in eldery patients admitted in emergency department. AI could optimize the diagnostic performance of radiologists (increase of confidence level) and improve the efficiency of suspected fractures sorting from emergency department.

Study Type

OBSERVATIONAL

Enrollment

1,000

Conditions

Painful Hip Fonctional Impotence After Minor Trauma (Fall)

Interventions

hip standard radiography/ CT scan/ MRI

Eligibility

Sex: ALLMin age: 60 Years

Medical Language ↔ Plain English

Inclusion criteria: * Patients aged 60 and older * Minor trauma * Admitted in emergency department for painful hip/fonctional impotence after minor trauma * Takes at least a hip radiography Exclusion criteria: * Painful before the minor trauma * Important trauma

Outcomes

Primary Outcomes

Detection rate of femoral neck fracture

Time frame: 1 day

Secondary Outcomes

Detection rate of other hip fracture