Optimization of MRI-based scoring scales of brain injury severity in children with unilateral cerebral palsy

Pagnozzi, Alex M., Fiori, Simona, Boyd, Roslyn N., Guzzetta, Andrea, Doecke, James, Gal, Yaniv, Rose, Stephen and Dowson, Nicholas (2015) Optimization of MRI-based scoring scales of brain injury severity in children with unilateral cerebral palsy. Pediatric Radiology, 46 2: 270-279. doi:10.1007/s00247-015-3473-y

Author Pagnozzi, Alex M.
Fiori, Simona
Boyd, Roslyn N.
Guzzetta, Andrea
Doecke, James
Gal, Yaniv
Rose, Stephen
Dowson, Nicholas
Title Optimization of MRI-based scoring scales of brain injury severity in children with unilateral cerebral palsy
Journal name Pediatric Radiology   Check publisher's open access policy
ISSN 1432-1998
Publication date 2015-11-10
Sub-type Article (original research)
DOI 10.1007/s00247-015-3473-y
Open Access Status Not Open Access
Volume 46
Issue 2
Start page 270
End page 279
Total pages 10
Place of publication Heidelberg, Germany
Publisher Springer Verlag
Language eng
Formatted abstract
Background: Several scoring systems for measuring brain injury severity have been developed to standardize the classification of MRI results, which allows for the prediction of functional outcomes to help plan effective interventions for children with cerebral palsy.

Objective: The aim of this study is to use statistical techniques to optimize the clinical utility of a recently proposed template-based scoring method by weighting individual anatomical scores of injury, while maintaining its simplicity by retaining only a subset of scored anatomical regions.

Materials and methods: Seventy-six children with unilateral cerebral palsy were evaluated in terms of upper limb motor function using the Assisting Hand Assessment measure and injuries visible on MRI using a semiquantitative approach. This cohort included 52 children with periventricular white matter injury and 24 with cortical and deep gray matter injuries. A subset of the template-derived cerebral regions was selected using a data-driven region selection algorithm. Linear regression was performed using this subset, with interaction effects excluded.

Results: Linear regression improved multiple correlations between MRI-based and Assisting Hand Assessment scores for both periventricular white matter (R squared increased to 0.45 from 0, P < 0.0001) and cortical and deep gray matter (0.84 from 0.44, P < 0.0001) cohorts. In both cohorts, the data-driven approach retained fewer than 8 of the 40 template-derived anatomical regions.

Conclusion: The equal or better prediction of the clinically meaningful Assisting Hand Assessment measure using fewer anatomical regions highlights the potential of these developments to enable enhanced quantification of injury and prediction of patient motor outcome, while maintaining the clinical expediency of the scoring approach.
Keyword Assisting Hand Assessment
Brain injury
Cerebral palsy
Magnetic resonance imaging
Structural assessment
Q-Index Code C1
Q-Index Status Confirmed Code
Institutional Status UQ

Document type: Journal Article
Sub-type: Article (original research)
Collections: Official 2016 Collection
School of Information Technology and Electrical Engineering Publications
School of Medicine Publications
Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 1 times in Thomson Reuters Web of Science Article | Citations
Scopus Citation Count Cited 2 times in Scopus Article | Citations
Google Scholar Search Google Scholar
Created: Tue, 24 Nov 2015, 10:26:06 EST by System User on behalf of Scholarly Communication and Digitisation Service