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

Abstract

Background and Objective(s): Previous studies propose that elucidation of brain injury from structural Magnetic Resonance Images (sMRI) for the clinical assessment of children with cerebral palsy (CP). Distinct aetiologies, including cortical maldevelopments, white and grey matter lesions and ventricular enlargement, have been categorised in a qualitative fashion. This study aimed to use several automated and validated MRI techniques to automatically quantify these brain lesions. Study Design: Measurement validation in a test cohort. Study Participants & Setting: 139 children aged 5-17 years, including 95 children diagnosed with unilateral CP (50 male, mean age of 11y 5mo [SD 3y 1mo], Manual Ability Classification System level I=38, II=57) and 44 children with healthy development (15 male, mean age of 11y 9mo [SD 2y 6mo]) were included. Materials/Methods: Structural MRI was performed at 3T. Clinical function was assessed for manual ability (AHA), motor planning (AMPs), cognition (IQ, EF), communication and vision (TVPS). The Least Absolute Shrinkage and Selection Operator (LASSO) feature selection approach was applied on a training data set (n=97), to find severity of specific injury biomarkers that were predictive of clinical functions (motor, cognitive, communicative and visual function). These biomarkers include cortical shape, regional lesion burden and volume of ventricular enlargement. The findings from this model were then validated to an unseen test set (n=42). Results: The measured prevalence of injury, including ventricular enlargement (70%), white and grey matter lesions (55%) and cortical malformations (30%), were similar to the prevalence observed in other cohorts of children with unilateral CP. Cortical shape (thickness, sulcal depth) and regional lesion burden were most often found to be associated with clinical functions. Test set correlation values ranged between 0.545 and 0.795 (p<0.008), indicating significant associations with clinical function. Conclusions/Significance: These findings support the early characterisation of injury from sMRI into previously defined aetiologies as part of standard clinical assessment. Furthermore, the strong and significant association between quantifications of injury observed on structural MRI and multiple clinical scores accord with empirically established structure-function relationships. (Figure Presented).L6184696992017-09-29 <br />