Abstract
PURPOSE: Numerous classification systems of nonsyndromic sagittal craniosynostosis (NSC) are applied but none has gained a wide acceptance, since each classification is focused on distinct aspects of cranial dysmorphology. The goal of this study was to depict the most common combinations of radiomorphologic characteristics of NSC and to separate groups where the patients were morphologically similar to one another and at the same time significantly different from others. METHODS: The study was conducted on anonymized thin-cut CT scans of 131 children with NSC aged 1-12 months (mean age 5.42 months). The type of cranial dysmorphology was assessed using four criteria: skull shape, pattern of sagittal suture fusion, morphologic features and cerebrospinal fluid (CSF) spaces alterations. After assigning the categories, an unsupervised k-modes clustering algorithm was applied to identify distinct patients clusters representing radiomorphologic profiles determined by investigated characteristics. RESULTS: Cluster analysis revealed three distinct radiomorphologic profiles including the most common combinations of features. The profiles were not influenced by sex nor age but were significantly determined by skull shape (V = 0.58, P < 0.0001), morphologic features (V = 0.50, P < 0.0001) and pattern of sagittal suture fusion (V = 0.47, P < 0.0001). CSF alterations did not significantly correlate with the profiles (P = 0.3585). CONCLUSION: NSC is a mosaic of radiologic and morphologic features. The internal diversity of NSC results in dissimilar groups of patients defined by unique combinations of radiomorphologic characteristics, from which the skull shape is the most differentiating factor. Radiomorphologic profiles support the idea of clinical trials targeted at more selective outcomes assessment.