MRI-based morphometric analysis of posterior cranial fossa in the diagnosis of chiari malformation type I

Aintzane Urbizu, Maria Antonia Poca, Xavier Vidal, Alex Rovira, Juan Sahuquillo, Alfons Macaya

Research output: Contribution to journalArticleResearchpeer-review

51 Citations (Scopus)


BACKGROUND AND PURPOSE: The diagnosis of Chiari malformation type I (CMI) relies on MRI identification of a tonsillar descent (TD) through the foramen magnum, reflecting the overcrowding of an underdeveloped posterior cranial fossa (PCF). However, TD occurs in some patients with normal-sized PCF and, conversely, some patients with borderline or no TD have small PCF. We thus sought to identify a set of prototypic PCF measures for the diagnosis of CMI. METHODS: We performed nineteen measurements of the PCF on sagittal MRI of 100 cases with cerebellar TD ≥5 mm and 50 control individuals, compared the average values in both cohorts and used logistic regression to devise a probability model to predict CMI status. RESULTS: Significant decrements were detected for several PCF-related measures in the patients' cohort. We developed a probability model that combined seven of these parameters to predict diagnosis with 93% sensitivity and 92% specificity. CONCLUSIONS: The addition of simple morphometric measurements in the diagnostic work-up of patients with suspected CMI may facilitate radiological diagnosis. Moreover, identification of the subset of CMI that arise from basichondrocranium underdevelopment is important for both, selection of the most appropriate therapeutic approach as well as proper CMI categorization in research studies. © 2013 by the American Society of Neuroimaging.
Original languageEnglish
Pages (from-to)250-256
JournalJournal of Neuroimaging
Issue number3
Publication statusPublished - 1 Jan 2014


  • Chiari malformation type I
  • Morphometry
  • MRI
  • ROC


Dive into the research topics of 'MRI-based morphometric analysis of posterior cranial fossa in the diagnosis of chiari malformation type I'. Together they form a unique fingerprint.

Cite this