Health Condition 1: M20-M25- Other joint disorders
Conditions
Interventions
Intervention1: nil: nil
Sponsors
DR P EVANGELINE CHRISTINA
Eligibility
Inclusion criteria
Inclusion criteria: Presenting Complaint: Patients with a primary complaint of low back ache lasting for more than 6 weeks. Imaging Requirement: Patients who underwent pelvic or lumbosacral imaging (MRI, CT, or X-ray) for evaluation of sacroiliitis. Clinical Suspicion of Sacroiliitis: Referred for imaging due to clinical signs of sacroiliitis (e.g., morning stiffness, sacral tenderness, inflammatory back pain). Availability of Radiologist Report: Radiologist-generated reports must be available for comparison. AI-Readable Images: Imaging data must be compatible with AI software for automated analysis. Consent: Informed consent provided for participation in the study and use of medical data.
Exclusion criteria
Exclusion criteria: History of Trauma or Surgery: Patients with prior pelvic or spinal fractures, surgery, or prosthetic implants in the pelvic region. Known Autoimmune Disorders: Patients with pre-diagnosed autoimmune conditions other than those typically associated with sacroiliitis (e.g., advanced rheumatoid arthritis or lupus). Inadequate Imaging Quality: Poor-quality scans that are not interpretable by AI systems or radiologists. Acute Infectious Sacroiliitis: Cases with infection-induced sacroiliitis diagnosed clinically or microbiologically. Pregnancy: Pregnant patients due to ethical concerns and differences in diagnostic approach. Inaccessible Radiology Data: Lack of radiologist reports or failure to retrieve complete imaging datasets. Non-cooperative Patients: Individuals unable or unwilling to complete clinical evaluations or imaging procedures.
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| The primary outcome is to assess the diagnostic accuracy of the AI system in detecting sacroiliitis, measured by sensitivity, specificity, PPV, and NPV. Secondary outcomes include overall accuracy, AUC (if applicable), interpretation time (AI vs. radiologist), inter-observer agreement (Cohen s Kappa), and the AI system s impact on clinical decision-making and radiologist confidence.Timepoint: All outcomes will be assessed at a single time point immediately after image upload and interpretation. Diagnostic metrics and interpretation times will be recorded during the evaluation, and feedback on clinical utility will be obtained post-diagnosis. | — |
Secondary
| Measure | Time frame |
|---|---|
| Reporting Time: Comparison of the time taken by AI(artifical intelligence) vs radiologists to generate reports. Inter-observer Agreement: Degree of agreement between AI (artifical intelligence )output & radiologist findings using metrics like Kappa or Intraclass Correlation Coefficient (ICC). Clinical Impact: Influence of AI(artifical intelligence)-supported diagnosis on clinical decision-making (e.g-initiation of specific treatments). Patient Outcomes: Changes in pain levels, mobility, or inflammatory markers (e.g-ESR(erythrocyte sedimentation rate, CRP(c- reactive protein)) after appropriate treatment based on diagnostic findings.Timepoint: Reporting Time: Recorded during the immediate reporting phase Inter-observer Agreement: Assessed within days by an independent review.Clinical Impact: Evaluated at 6-week follow-up after the initial diagnosis. Patient Outcomes: Pain & functional status assessed at 6 weeks & 12 weeks post-diagnosis to evaluate the effectiveness of the management strategy. | — |
Countries
India
Contacts
Public ContactDR.P.EVANGELINE CHRISTINA
Saveetha Medical College Hospital
Outcome results
None listed