Ultrafast CT scoring system for assessing bronchopulmonary dysplasia: reproducibility and clinical correlation.
PURPOSE: To evaluate the reproducibility of the Ultrafast CT (UFCT) scoring system and assess its usefulness in monitoring clinical severity in infants with bronchopulmonary dysplasia (BPD). MATERIALS AND METHODS: UFCT scoring was done in 22 infants (15 boys and 7 girls aged 1 to 37 months) with BPD. A total of 258 lung fields were evaluated for the presence of hyperaeration, linear opacities, triangular subpleural opacities, and bronchovascular bundle distortion or thickening, and UFCT scores were given. Intraobserver and interobserver agreement and reproducibility of UFCT scores were statistically analyzed. In 12 patients, UFCT scores were linearly correlated with clinical severity scores based on respiratory dysfunction and complexity of care. RESULTS: "Hyperaeration," which was the most frequent (18 of 22, 81.8%) finding, showed high concordance (kappa = 0.73, p < 0.001, kappa = 0.59, p < 0.001), and its UFCT scores significantly correlated with intraobserver and interobserver analyses (r = 0.94, p < 0.001, r = 0.82, p < 0.001, respectively). UFCT scores for hyperaeration significantly correlated with clinical scores (r = 0.75, p < 0.01), whereas those for the others did not. CONCLUSION: UFCT is useful for assessing BPD. Hyperaeration was the most common and reproducible finding, and its extent significantly correlated with clinical severity.