P1707 Can 3 Dimensional Motion Analysis and Fuzzy Entropy detect movement differences in General Movement Assessment Categories in the normative infant population?
Mrs Michelle McGrath1,2,3, Prof Ian Turner1,3, Dr Hongbo Xie1,3, Prof Mark Pearcy4, Dr Robyn Grote1,2, Prof Paul Colditz2,5
1Queenland University of Technology, Brisbane, Australia. 2Queensland Health, Brisbane, Australia. 3ARC Centre of Excellence for Mathematical & Statistical Frontiers (ACEMS), Brisbane, Australia. 4Queensland University of Technology, Brisbane, Australia. 5University of Queensland, Brisbane, Australia


Introduction: Prechtl’s Method on the Qualitative Assessment of General Movement (GMsA) of infants (Darsaklis et al., 2011, Einspieler, 2004) is one method of early prediction of neurodevelopmental outcomes in infants. This movement assessment uses video recordings and the naked eye of the assessor and has established 2 distinct movement classifications (Writhing and Fidgeting) which occur in healthy infants aged from term to 1 month and 3 months, respectively.

GMsA relies on qualitative classification, resulting in potential intra-, inter- assessor reliability and naked eye errors.  A review from Darsaklis et al. (2011) found that there was conflicting evidence that GMsA could accurately predict neurodevelopmental outcomes particularly in cases where the neurological dysfunction was mild. Three Dimensional Motion Analysis (3DMA) may be a useful tool in measuring small movements that maybe missed in predicting neurodevelopmental outcomes using GMsA’s naked eye approach.

Compared to GMsA’s, 3DMA gives insight in all three planes (sagittal, coronal & transverse) of infant movement without “naked eye” error.  Complemented by force plate data, enhanced understanding of the infant postural behaviour can be determined.

Method: This novel preliminary study developed an infant 3DMA technical protocol to collect infant movement using the latest 3D motion analysis technology. The collected data was used to quantify GMsA movement classifications using a mathematical pattern recognition technique Fuzzy Entropy, in a small cohort of healthy full term infants.

Results and Discussion: During the development of the Infant 3DMA protocol, it was found that the optimum placement of markers for the infant biomechanical model, with the least amount of interference from the infant and obstruction from the cameras, were located on the thigh, shin, foot, upper arm, forearm, chest and head segments.

Fuzzy entropy was found to be a useful mathematical pattern recognition technique to separate movement patterns in the writing and fidgeting datasets especially in the force datasets. It was found that there was a marked decrease in force fuzzy entropy between the writhing and fidgeting data. It is believed this decrease in entropy is due to the infants’ postural stability and control development.

This normative-reference data, along with the developed infant technical protocol and fuzzy entropy, could be used to differentiate between potential normal and abnormal child development and the need for targeted early interventions.


This work was supported by Robert Bird Group and Royal Brisbane and Women’s Hospital foundation.




DARSAKLIS, V., SNIDER, L. M., MAJNEMER, A. & MAZER, B. 2011. Predictive validity of Prechtl’s method on the qualitative assessment of general movements: a systematic review of the evidence. Developmental Medicine & Child Neurology, 53, 896-906.

EINSPIELER, C. 2004. Prechtl's method on the qualitative assessment of general movements in preterm, term and young infants, Mac Keith Press London, UK.