“School of Cognitive Sciences”

Back to Papers Home
Back to Papers of School of Cognitive Sciences

Paper   IPM / Cognitive Sciences / 14230
School of Cognitive Sciences
  Title:   Long-Range Reduced Predictive Information Transfers of Autistic Youths in EEG Sensor-Space during Face Processing
  Author(s): 
1.  A. Khadem
2.  GA. Hossein-Zadeh
3.  A. Khorrami
  Status:   In Proceedings
  Proceeding: 4th basic and clinical neuroscience congress 2015
  Year:  2015
  Supported by:  IPM
  Abstract:
Autism Spectrum Disorder (ASD) affects the information processing in the brain by changing the coupling structures among neurons and their synapses. The common and essential sign of ASD is the disordered social interactions and communications. The majority of functional/effective connectivity studies conducted on the ASD patients has converged to the underconnectivity theory of ASD: ‘‘long-range underconnectivity and sometimes short-rang overconnectivity’’. Recently, we conducted a study to compare the total (linear and nonlinear) predictive information transfers (PITs) of autistic and typically developing healthy youths during human face processing by using EEG data. The ERPs of 12 autistic youths and 19 age-matched healthy control (HC) subjects were recorded while they were watching upright and inverted human face images. The PITs among EEG channels were quantified using transfer entropy with self-prediction optimality (TESPO). Afterwards, the directed differential connectivity graphs (dDCGs) were constructed to characterize the significant changes in the estimated PITs of autistic subjects compared with HC subjects. By using TESPO, long-range reduction of PITs of autistic subjects during face processing was revealed (particularly from frontal channels to right temporal channels). These long-range reduced PITs may be hypothesized to underlie the difficulties of autistic subjects in perception of human faces. Also, it seemed the orientation of face images (upright or upside down) did not modulate the binary pattern of PIT-based dDCGssignificantly.

Download TeX format
back to top
scroll left or right