MENU

Indietro
Jacopo De Angelis1, Paola Ricciardelli1, Enzo Grossi2
Pain Anticipation and Empathy for Pain Towards Social and Non-Social Agents in Adults with ASD: A Novel Psychophysiology Artificial Neural Networks Based Study (2021)
Accepted at INSAR 2021

Abstract: Background: 
Although abnormal pain sensitivity has become one of the diagnostic criteria for Autism Spectrum Disorder (ASD), this area is still under-investigated. Speci cally, a very few studies focused on the anticipatory components characterizing pain experience related to self and others in ASD. 

Objectives: 
The present study was aimed at investigating psychophysiological responses in adults with High Functioning Autism (HFA) when observing a painful stimulus approaching (pain anticipation) their own skin (self-condition), another participant skin (other social agent condition) or a robotic hand (other non-social agent condition). 

Methods: 
Skin Conductance Responses (SCR) were measured in a group of adult participants with ASD diagnosis and a control group of participants without any psychiatric or neurological diagnosis. The experimental sessions consisted of a naturalistic paradigm where the participants observed a painful (sterilized needle) or a neutral (cotton bad) stimulus that could approach their own skin, the skin of another person or the hand of a robot (Figure 1). Data were analysed by implementing a new Machine Learning approach combining predictive modelling (i.e. Arti cial Neural Networks) and Auto-Contractive Maps, that is an emerging data mining approach able to detect connectivity associations among variables through an arti cial adaptive system. 

Results: 
The predictive model was 91% accurate and able to differentiate ASD participants from control group participants. Two distinct patterns were found for each group. Specifically, ASD individuals exhibited a lower physiological activation than controls when the noxious stimulus (needle) approached their skin after repeated stimulations. Interestingly, ASD participants showed a higher response when the robotic hand was approached by the painful stimulus, whereas no differences related to the human social agent between the two groups emerged. 

Conclusions: 
These results can contribute to clarifying the psychophysiological mechanisms underlying pain experience in ASD individuals. Implications for the study of the self-injury phenomenon, often observed in ASD, will be discussed.

Notes:

1 - University of Milano - Bicocca, Italy
2 - Villa Santa Maria Foundation