van Eijk, L., Seidel, M., Pannek, K., George, J.M., Fiori, S., Guzzetta, A., Coulthard, A., Bursle, J., Ware, R.S., Bradford, D., Rose, S., Colditz, P.B., Boyd, R.N., and Fripp, J. (2021) Automating quantitative measures of an established conventional MRI scoring system for preterm-born infants scanned between 29 to 47 weeks' postmenstrual age. American Journal of Neuroradiology, 42 (10). pp. 1870-1877.

PDF (Published Version) - Published Version Restricted to Repository staff only

 

2

Abstract

Background and Purpose: Conventional MRI scoring is a valuable tool for risk stratification and prognostication of outcomes, but manual scoring is time-consuming, operator-dependent, and requires high-level expertise. This study aims to automate the regional measurements of an established brain MRI scoring system for preterm neonates scanned between 29-47 weeks post-menstrual age (PMA).

Materials and Methods: This study used T2WI from the longitudinal Prediction of PREterm Motor Outcomes cohort study (PPREMO) and developing Human Connectome Project. Measures of biparietal width, interhemispheric distance, callosal thickness, transcerebellar diameter, lateral ventricular diameter, and deep grey matter area were extracted manually (PPREMO only) and automatically. Scans with poor quality, failure of automated analysis, or severe pathology were excluded. Agreement, reliability, and associations between manual and automated measures were assessed, and compared against statistics for manual measures. Associations between measures with PMA, gestational age at birth (GA), and birth weight were examined (Pearson’s correlation) in both cohorts.

Results: 652 MRIs (86%) were suitable for analysis. Automated measures showed good to excellent agreement and good reliability with manual measures, except for interhemispheric distance at early MRI (scanned between 29-35 weeks PMA; in line with poor manual reliability) and callosal thickness measures. All measures were positively associated with PMA (r=0.11-0.94; R2=0.01-0.89). Negative and positive associations were found with GA (r=-0.26-0.71; R2=0.05-0.52), and birth weight (r=-0.25-0.75; R2=0.06-0.56). Automated measures were successfully extracted for 80-99% of suitable scans.

Conclusion: Measures of brain injury and impaired brain growth can be automatically extracted from neonatal MRI, which could assist with clinical reporting.

Actions (Repository Staff Only)

Item Control Page