Abstract: Background 

Complex neurodevelopmental disorders like ASD may be characterized by subtle brain function signatures which can be measured using advanced machine learning tools. 
Sound postural control system requires the interaction of sensory, motor and integration 
systems. There is evidence that individuals with Autism Spectrum Disorder (ASD) present postural 
control deficits but is not known if the complexity of sensorimotor contribution can play a role. 
Objective 
The aim of the study is to compare the degree of complexity of sensory and motor functions in a group of subjects with autism with a matched group of Typically developing children and adolescents. 
Methods 
The study sample included 18 children and adolescents with ASD and 18 age-matched typically-developing (TD) individuals, (age range 6-18 years; 16 males for both groups). 
Participants completed a Physiological Profile Assessment, which includes 17 tests : Visual acuity with high- and low-contrast, Edge contrast sensitivity, Depth perception, Touch sensitivity, Tactile lower limb Proprioception, Simple reaction time of hand and foot, Isometric knee extensor and flexor muscle strength, Ankle isometric dorsiflexion strength, Postural sway with eyes open and closed, first on a firm surface, then on a medium density foam rubber mattress, Maximum balance range and Coordinated stability. 
The complexity of interaction among these physiological functions has been explored with Maximally Regular Graph (MRG) theory applied to the Semantic Connectivity map produced by a fourth generation artificial neural network (Auto-CM system; Semeion, Rome, Italy). 
MRG is the graph whose hubness function attains the highest value among all the graphs generated by adding back to the original Minimum Spanning Tree (MST), one by one, the connections previously skipped during the computation of the MST itself. In other words, starting from the MST, the MRG, presenting the highest number of regular microstructures, highlight the most important connections of the dataset. The resulting “diamond” expresses the complexity core of the system and in our specific case, the core of the physiological system. 
Results 
The map relative to TD group showed a coordinating role of Maximal Balance Range with respect to physiological functions on study. The MRG was generated by the addition of twelve links to MST involving eight Nodes (figure 1). At variance, the map relative to ASD group showed a coordinating role of touch sensitivity. The MRG was characterized by only six additional links involving five nodes ( figure 2). 
Conclusion 
This proof-of-principle study suggests that ASD postural physiology is characterized by a loss of complexity in its component interaction which in the future might become a useful biomarker for early detection of risk for ASD. 

Notes: 

(1)Autism Research Unit, Villa Santa Maria Foundation, Tavernerio, Italy
(2)School of Medicine and Surgery, University of Milan-Bicocca, Milano, Italy