Brain-based concealed memory detection is driven mainly by orientation to salient items. (March 2021)

• Record Type: Journal Article
• Authors: klein Selle, Nathalie; Gueta, Chen; Harpaz, Yuval; Deouell, Leon Y.; Ben-Shakhar, Gershon
• Is Part Of: Cortex. Volume 136(2021)
• Page Start: 41
• Page End: 55
• Publication Date: 2021-03
• Subjects: Concealed Information Test (CIT) -- Orienting response -- Arousal inhibition -- Event-related potentials -- Machine learning; Neuropsychology -- Periodicals; Nervous system -- Periodicals; Neurology -- Periodicals; Psychophysiology -- Periodicals; Behavior -- Periodicals; Neurology -- Periodicals; 612.825
• DOI: 10.1016/j.cortex.2020.12.010
• Languages: English
• ISSNs: 0010-9452

Abstract

Abstract: In the pursuit of new methods for concealed memory detection, event-related potential components (ERP) have been placed at the forefront of research. No method, however, is scientifically complete without a theory and the present study therefore aimed to unravel the cognitive processes underlying these ERPs (i.e., orienting and arousal inhibition). This was accomplished by using a Concealed Information Test (CIT) in which participants were once motivated to conceal and once motivated to reveal their identity. The results showed a similarly strong P3 CIT effect in the two motivational conditions, which was enhanced for high salience compared to low salience identity items. Similar results were observed when using a multivariate machine-learning algorithm – suggesting that brain-based concealed memory detection is driven mainly by orientation to salient stimuli, rather than by arousal inhibition. In addition, the algorithm, trained and tested on the ERPs of different identity items, achieved detection rates exceeding those achieved by the P3. This implies that CIT researchers and practitioners could potentially rely on the entire ERP waveform instead of a-priori selecting separate components. Together these results enrich current understanding of the mechanisms underlying neurophysiological responding to concealed information and pave the way for novel and powerful algorithms which could be used in real-life forensic investigations. Highlights: We examined the underlying processes of brain-based concealed memory detection. Univariate (P3) responses and multivariate patterns (MVP) were used. MVP machine-learning algorithm achieved detection rates exceeding those of P3. P3 and MVP are affected by salience, not motivation to conceal information. Brain responses to concealed information are driven by orienting, not inhibition.