Diagnostic performance of dual-energy CT for the detection of bone marrow oedema: a systematic review and meta-analysis

The aim of this systematic review and meta-analysis was to assess the sensitivity and specificity of dual-energy CT (DECT) for the detection of bone marrow oedema (BME).

An electronic search of the PubMed and EMBASE databases was conducted. Bivariate modelling and hierarchical summary receiver-operating characteristic modelling were performed to evaluate the overall diagnostic performance of DECT for BME. Subgroup analysis was performed according to the assessment type (qualitative vs. quantitative) and anatomical location (spine vs. appendicular skeleton). Meta-regression analyses were performed according to the subject, study, and DECT characteristics.

Twelve eligible studies (1901 lesions, 450 patients) were included. DECT exhibited a pooled sensitivity of 0.85 [95% confidence interval (CI): 0.78–0.90] and a pooled specificity of 0.97 (95% CI: 0.92–0.98) for BME detection. In addition, the diagnostic performance of qualitative assessment (sensitivity, 0.85; specificity, 0.97) was higher than that of quantitative assessment (sensitivity, 0.84; specificity, 0.88) of DECT findings. The diagnostic performance of DECT for the spine (sensitivity, 0.84; specificity, 0.98) and appendicular skeleton (sensitivity, 0.84; specificity, 0.93) were excellent. According to meta-regression analysis, the use of a tin filter, ≥ 2 image planes, and a slice thickness < 1 mm tended to exhibit higher sensitivity and hyperacute stage BME (< 24 h) tended to exhibit lower sensitivity.

These findings indicate that DECT has excellent sensitivity and specificity for BME detection. Qualitative assessment of DECT findings obtained using a tin filter, ≥ 2 image planes, and a 0.5-1-mm slice thickness in the acute stage BME (≥24 h) is recommended for more sensitive diagnosis.