Go to contents

Prosecution develops brain wave analysis method using AI

Prosecution develops brain wave analysis method using AI

Posted December. 26, 2023 08:04,   

Updated December. 26, 2023 08:04

한국어

The South Korean prosecution has developed a new technique for analyzing brain waves using artificial intelligence (AI) machine learning technology and has decided to incorporate it into investigations of violent crimes. This marks the first instance of the prosecution applying AI technology to an investigation.

According to The Dong-A Ilbo's report on Monday, the Supreme Prosecutor's Office's Forensic Investigation Department has recently collaborated with Professor Lim Chang-hwan's team at the Department of Brain Engineering at Hanyang University to develop a brain wave analysis technique using machine learning.

The prosecution's current EEG analysis technique involves the utilization of specific EEG signals, such as the P300 (EEG that appears 0.3 seconds after stimulus presentation), triggered when a familiar stimulus is presented to the brain. This method aims to measure changes in brain waves after presenting details of the location and instrument involved in the crime, assuming that such information is recorded in the suspect's brain. This technique was traditionally employed in cases lacking sufficient evidence or when a suspect denied the charges. However, its effectiveness as evidence was limited due to accuracy concerns.

The new technique developed by the prosecution in collaboration with Professor Lim's team is distinguished by its capacity to analyze the P300 response, the connection strength, and the number of changes in signals exchanged between different brain regions. For instance, when a suspect, provided with details such as the weapon or location of the crime, undergoes brain examination, activation in the parietal lobe - responsible for processing sensory information like vision and touch - is observed. Furthermore, an increase in connectivity with the decision-making part is noted, and the brain's overall 'network flow' is comprehensively analyzed. Based on these findings, machine learning technology is employed to teach artificial intelligence (AI) how to differentiate between criminals, references, and ordinary witnesses.


jej@donga.com