Until now, for children with cerebral palsy (CP) , diagnostic and some prognostic predictive machine learning studies have been conducted, but prognostic studies targeted specific milestone according to specific gross motor function measure (GMFCS) levels; such as walking and running predictors at GMFCS II and III and GMFCS II respectively, and not covered specific types of cerebral palsy. Predictions studies were limited by the lack of specificity of child and family characteristics was not taken into the account prospectively.

It is therefore the utmost need to support clinical decision making by predicting the recovery in spastic cerebral palsy. Recovery predictive factors can play an important role for this purpose. Thus, this study aims to predict the recovery in spastic cerebral palsy according to all GMFCS level by means of a prediction index/model.

Spasticity is often considered as the main cause of functional limitation in cerebral palsy (CP) children. The main feature of cerebral palsy is the impaired development of gross motor functions in children. Gross motor functions are considered as an indicator of the overall prognosis of cerebral palsy as these are closely associated with other impairments in the cerebral palsy child such as auditory, cognitive or visual impairments. The gross motor function measure (GMFM) tool is most widely used to assess motor function, severity and treatment response of children with cerebral palsy. The five levels of GMFCS have been widely employed in cerebral palsy children less than 12 years of age with the focus on sitting and walking abilities of the child. Literature confirmed the importance of addressing the gross and fine motor skills in cerebral palsy children. Childhood factors that predict the participation of young adults with cerebral palsy in domestic life include; intellectual disability, low manual ability, limited motor capacity and epilepsy. Moreover, CP child primary and secondary impairments, co-morbidities, their adaptive behaviour, family, rehabilitation services all are determinants of changes in the gross motor ability of the child and their participation in daily routine activities. Thus, all these determinants need to be considered while planning the intervention for a cerebral palsy child and at the time outcome evaluation as well. Good prognostic predictors for ambulation in cerebral palsy children were identified through meta-analysis of observational studies which includes; independent sitting at 2 years of age, epilepsy, absence of intellectual disability and visual impairment. Machine learning (ML) approaches have been increasingly used in cerebral palsy research. Jing Zhang et al identified GMFCS and intellectual capacity as associated factors of self-care activity development, it was also mentioned that GMFCS has a role in mobility activities development. A predictive machine learning model was developed to highlight the factors associated with intellectual disability in the cerebral palsy population of the teenager, with the sensitivity, specificity and average accuracy of 78%. The result of this model confirmed the significant association of gross motor function, poor manual abilities and epilepsy with profound intellectual disability.

Inclusion Criteria:

• Cerebral palsy children with positive score 1 on items 3 and 4 of Hypertonia Assessment Tool (HAT).
• Spastic cerebral palsy children with any functional ability level(Gross Motor Function Classification System (GMFCS) levels I-V)

Exclusion Criteria:

• Children of any other type of cerebral palsy
• Not willing to participate

Spastic cerebral palsy children