Hand Recognition Obtained by Simulation of Hand Regard

Area of Science:

• Developmental psychology
• Computational neuroscience
• Infant motor development

Background:

• Infant eye-hand coordination develops early, with hand regard appearing around two months.
• The mechanism by which infants recognize their own hands during hand regard remains unclear.
• Understanding this process is crucial for comprehending early self-recognition and sensorimotor integration.

Purpose of the Study:

• To simulate and elucidate the process of infant hand recognition during hand regard.
• To investigate how infants distinguish their hands from other objects.
• To model the neural mechanisms underlying early self-recognition.

Main Methods:

• A modified real-time recurrent learning (RTRL) algorithm was employed to train a simple neural network.
• The simulation focused on processing time-varying input and output data characteristic of hand regard.
• The model aimed to replicate the developmental stages of hand recognition.

Main Results:

• Simulation results indicated that information for hand recognition is stored in self-organized cell assemblies.
• These cell assemblies emerged during U-shaped developmental phases of hand regard.
• The configuration of cell assemblies dynamically changed with each developmental phase, correlating with general movements (GMs).

Conclusions:

• Self-organized cell assemblies play a key role in storing and processing hand recognition information in infants.
• The dynamic changes in cell assemblies reflect the developmental trajectory of hand regard.
• The findings provide insights into the neural basis of early self-perception and sensorimotor learning.