Search for a command to run...
Abstract This study addresses the need for objective, real-time assessment of emotional responsiveness and coping strategies in individuals with Amyotrophic Lateral Sclerosis (ALS) to support personalized care. We are using non-invasive speech analysis and data science methods on an expanded cohort comprising 28 ALS patient visits. We first demonstrate that commonly available artificial intelligence tools, including current-generation large language models (LLMs), such as ChatGPT, Gemini and Claude, do not provide reliable or reproducible assessments of patient concern levels in the absence of expert clinical supervision. Further, we observe a discrepancy between subjective and objective metrics such as the forced vital capacity for breathing. We introduce a novel functional classification system that contextualizes clinician-rated emotional concern relative to the patient’s functional impairment as measured by the ALS Functional Rating Scale (ALS-FRS). Patient responses are categorized as: Congruent: Emotional responsiveness is proportional to functional impairment. Muted: Emotional response is lower than expected given functional impairment. Excessive: Emotional response exceeds that expected given functional impairment Key Findings Emotional Drivers and Acoustic Filtering We subsequently identified objective digital biomarkers, independent of LLMs, that robustly distinguish the muted and excessive response groups. Our data analysis revealed that the excessive response group was predominantly male and required significantly more clinician time, suggesting that this type of classification is primarily driven by psychosocial coping mechanisms such as anxiety related to functional dependency. Conversely, the muted response group was predominantly female. Most significantly, we discovered a direct link between this behavioral classification and the patient’s speech acoustics, demonstrating that dysarthria may act as an acoustic filter for emotional expression: Patients with muted response exhibited a voice profile characterized by high acoustic loudness, high sharpness, low roughness, and low fluctuation consistent with attenuation of emotional expression. Patients with excessive response presented with the opposing profile: low loudness, low sharpness, high roughness, and high fluctuation consistent with amplification of emotional expression. However, a significant diagnostic overlap exists between these biomarkers of anxiety and the early manifestations of Amyotrophic Lateral Sclerosis, spastic dysarthria for the former, flaccid dysarthria for the second. Our work enables non-invasive assessment of both emotional coping strategies and the dominant underlying neuromuscular mechanisms. These insights support the early identification of patients who may benefit from targeted, proactive interventions, and represent a significant step toward personalized ALS management.