Ehab Khatab

Ehab Khatab is affiliated with the School of Mathematical and Computer Sciences at Heriot-Watt University Dubai, United Arab Emirates. Based on the publication record, this researcher appears to be an early-career academic working at the intersection of computer science, computer vision, and emerging technologies. Previous work includes research on computer vision algorithms in drone surveillance and high-speed edge detection implementation, suggesting a background in image processing and computational systems before entering the polygraph research domain.

Khatab's 2025 contribution to polygraph science involved presenting a new multimodal dataset with physiological data (heart rate, galvanic skin response, and body temperature), in addition to demographic data, collected from 49 unique subjects, and presenting a polygraph-based lie detection system utilizing multimodal sensor fusion. The study employed machine learning optimization for deception detection, with Random Forest achieving an accuracy of 97%, outperforming Logistic Regression, Support Vector Machine, and k-Nearest Neighbor algorithms. A notable finding was that demographic factors, such as weight and height, contributed more to the model's predictions than physiological signals. The research team also publicly released the multimodal polygraph dataset, contributing an open resource to the field.

Khatab's polygraph research was conducted as part of a collaborative team including Omar Osama Mohamed Shalash (an Assistant Professor at the Arab Academy for Science, Technology, and Maritime Transport with a Ph.D. in Biomedical Engineering from the University of Strathclyde, UK), Ahmed Métwalli (Assistant Lecturer at AASTMT specializing in AI and Data Science with 17+ publications spanning multidisciplinary AI applications), and Mohamed Hany Sallam (also affiliated with AASTMT). This multi-institutional collaboration bridged expertise from the United Arab Emirates and Egypt, reflecting growing regional interest in applying artificial intelligence to forensic psychophysiology.

The work represents an entry point into the emerging field of AI-enhanced polygraph systems, focusing specifically on multimodal sensor fusion and machine learning classification for credibility assessment. While the sample size was modest, the research contributes to the broader trend of applying computational methods to traditional psychophysiological detection, an approach that remains scientifically controversial but technologically active in the mid-2020s.