Bone marrow hyperattenuation on CT for detection of hip fractures without cortical disruption and assessment of intertrochanteric extension in greater trochanter fractures

Abstract Purpose To evaluate the diagnostic performance of bone marrow hyperattenuation (BMH) on CT for detecting fractures in patients without definite cortical disruption on CT and assessing the extent of intertrochanteric (IT) fracture extension in patients with suspected isolated greater trochanter fractures (IGTFs). Materials and methods This retrospective study included two groups of patients aged ≥ 65 years undergoing CT for suspected hip fractures: the cortical disruption–negative group (without definite cortical disruption on CT) and the suspected IGTF group (with a cortical fracture confined to the greater trochanter on CT). BMH was defined as a visually appreciable, localized increase in bone marrow attenuation on CT. Three radiologists independently evaluated CT images. In the cortical disruption–negative group, BMH was visually assessed for fracture detection, using MRI or clinical follow-up as the reference standard. In the suspected IGTF group, visual assessment evaluated whether BMH extended ≥ 50% of the IT line, using MRI as the reference standard. Patients with bilaterally evaluable femora underwent ROI-based quantitative assessment. Diagnostic performance, agreement, ROC analysis, and reproducibility were analyzed. Results In the cortical disruption–negative group ( n = 57), visual BMH assessment showed median sensitivity and specificity of 90.5% and 97.2% for fracture detection, with almost perfect agreement (κ = 0.90). In the suspected IGTF group ( n = 25), visual assessment showed median sensitivity and specificity of 86.7% and 80.0% for detecting IT extension ≥ 50%, with substantial agreement (κ = 0.72). In ROI assessment, performed in 50 and 16 patients, reader-specific AUCs were 0.720–0.935 for fracture detection and 0.945–1.000 for IT extension assessment; corresponding ICCs were 0.70 and 0.86. Conclusion BMH on CT allows fracture detection without definite cortical disruption and assessment of IT extension in suspected IGTFs, supporting its role as an adjunctive CT finding. Quantitative assessment may provide complementary information in selected patients.