Beyond clinical observations: a scoping review of AI-detectable observable cues in borderline personality disorder

Borderline Personality Disorder (BPD), impacting approximately 2% of adults worldwide, presents a formidable challenge in psychiatric diagnostics. Often underdiagnosed or misdiagnosed, BPD is associated with high morbidity and mortality. This scoping review embarks on a comprehensive exploration of observable cues in BPD, encompassing language patterns, speech nuances, facial expressions, nonverbal communication, and physiological measurements. The findings unveil distinctive features within the BPD population, including language patterns emphasizing external viewpoints and future tense, specific linguistic characteristics, and unique nonverbal behaviors. Physiological measurements contribute to this exploration, shedding light on emotional responses and physiological arousal in individuals with BPD. These cues offer the potential to enhance diagnostic accuracy and complement existing diagnostic methods, enabling early identification and management in response to the urgent need for precise psychiatric care in the digital era. By serving as possible digital biomarkers, they could provide objective, accessible, and stress-reducing assessments, representing a significant leap towards improved psychiatric assessments and an invaluable contribution to the field of precision psychiatry.