Related Experiment Video
Updated: Aug 28, 2026
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
Diagnostic heterogeneity in granulomatous inflammation: High-resolution computed tomography imaging patterns and
Jawad Ali Memon1, Mohammad Sibtain Shah2, Zubair Ali Memon3
1Department of Diagnostic Radiology, Peoples University of Medical and Health Sciences For Women (PUMHSW), Nawabshah, Pakistan.
Background:
Granulomatous inflammation presents substantial diagnostic complexity in tuberculosis-endemic regions. Radiological-histopathological concordance for distinguishing granulomatous diseases in resource-limited South Asian healthcare settings remains poorly characterized. Whether specific HRCT phenotypes reliably predict causative etiologies has not been systematically evaluated.
Methods:
Cross-sectional diagnostic study enrolling 300 participants with suspected granulomatous inflammation at a tertiary care center. Participants underwent HRCT chest imaging and tissue sampling with histopathological confirmation. Primary outcome was radiological-histopathological concordance assessed by Spearman's correlation and weighted kappa. Secondary outcomes included diagnostic accuracy by etiology, inter-rater reliability for radiologists and pathologists, and specific HRCT predictor features. All analyses reported sensitivity, specificity, and 95% confidence intervals.
Results:
Of 300 enrolled participants (mean age 48.2 years; 68.0% male), radiological-histopathological concordance was strong: Spearman's ρ = 0.72 (95% CI 0.64-0.79; p<0.001), weighted kappa κ_w = 0.68 (95% CI 0.61-0.75). Sensitivity 81.3% (95% CI 75.2%-86.8%), specificity 78.6% (95% CI 71.4%-85.1%). Diagnostic accuracy: tuberculosis 88.5%, sarcoidosis 79.2%, malignancy 81.8%, fungal disease 72.2%. Upper lobe predominance (adjusted OR 6.42) and cavitation (OR 5.24) strongly predicted tuberculosis.
Conclusions:
Specific HRCT features demonstrated strong concordance with histopathological diagnosis and high accuracy for granulomatous etiologies, particularly tuberculosis. However, diagnostic performance varied substantially by etiology and radiological complexity. System-level barriers to biopsy access remain critical limitations. These findings support evidence-based radiological-pathological diagnostic algorithms in resource-limited settings.
