The MRI classification system (MRICS) for Cerebral Palsy (CP) proposed by the SCPE network provides the framework for understanding etiology and timing of underlying brain disorders and aims to overcome variations in diagnosis in different countries by applying simple and  straightforward criteria.

Methods

Post-neonatal MRIs of 301 patients from the Greek CP registry (born 1999-2006), were classified according to MRICS. Statistical processing and correlation analysis were performed.

Results

MRICS classification: A) Maldevelopments 20/301 (6.7%), B) Predominant white matter injury 164/301 (55%), C) Predominant grey matter injury 43/301 (14.3%), D) Miscellaneous 51/301 (16.5%), E) Normal 23/301 (7.5%). Spastic CP was the commonest type (245/301, 81.4%), with bilateral distribution (BSCP) in 80% of cases, followed by dyskinetic (16/301, 5.3%), ataxic (14/301, 4.7%) and non-classifiable (26/301, 8.6%) CP. Correlation analysis revealed that BSCP was mostly related with MRICS type B, hemiplegia with type C, ataxic with types D and E lesions (P<0.001). Gestational age (GA) <32 weeks was significantly associated with MRICS type B lesions; term pregnancies were more likely to exhibit MRICS type C, D or E pathology (both groups P<0.001). Caesarean section was the principal mode of delivery (70%), for which an increased possibility for MRICS type B lesions was observed (P<0.001).

Conclusion

Specific MRI patterns were associated with CP subtype, GA, delivery mode and other categorical variables. Significant correlations were extracted for MRICS categories and CP type, verifying white matter injury as the main finding. The MRICS by SCPE provides a valid system for understanding pathophysiology and clinical presentations in CP.