EarliPoint

The EarliPoint System aims to provide an objective aid for the diagnosis and assessment of Autism Spectrum Disorder in young children by utilizing eye-tracking technology to analyze their looking behavior in response to video-based social stimuli. The system's science relies on an artificial intelligence algorithm to interpret these gaze patterns, generating a diagnostic output and severity indices correlated with established clinical measures.

The EarliPoint System is designed as an objective diagnostic aid for Autism Spectrum Disorder (ASD) in children aged 16 to 30 months, intended for use in specialized developmental disabilities centers. Its methodology involves an eye-tracking module that records a child's visual responses—specifically their looking behavior—to a series of age-appropriate videos presenting social information. This data is then securely transmitted to a web portal where artificial intelligence software analyzes it. The system, which has been modified for improved usability and portability with separate patient eye-tracking and operator modules connected via a web-portal, processes this data to provide an ASD diagnosis. The scientific basis of the EarliPoint System lies in its proprietary algorithm that analyzes eye-tracking data to determine ASD status and provides three "EarliPoint Severity Indices": Social Disability, Verbal Ability, and Non-verbal Ability, which correlate with established clinical instruments like ADOS-2 and Mullen scales. A pivotal multi-center study involving 475 evaluable patients demonstrated the system's performance against expert clinical diagnosis. For the overall evaluable population, the device showed a sensitivity of 71% and specificity of 80.7%. In a sub-group where clinicians were certain of their diagnosis (N=335), sensitivity increased to 78.0% and specificity to 85.4%, with no serious adverse events reported, indicating its utility as a supportive tool in ASD assessment.

AI Regulatory Concerns

Powered by

Jev reads the full FDA letter and returns calibrated concern probabilities per regulatory category. These are AI estimates, not regulatory advice.

• Q1. How does the algorithm differentiate between typical and atypical looking behaviors in response to social video stimuli, and what specific visual patterns are most indicative of ASD according to this system?; What is the inter-rater reliability of the EarliPoint Severity Indices when compared to traditional ADOS-2 and Mullen assessments administered by different clinicians?; Given the approximately 15-minute duration of use, how robust is the system to variations in child attentiveness, movement, or non-compliance during the eye-tracking session, and how are these factors accounted for in the analysis?; What is the longitudinal validity of the EarliPoint System's diagnosis and severity indices; specifically, how well do early diagnoses made with this tool predict later developmental trajectories and outcomes?; Considering the specified age range of 16-30 months, what are the performance characteristics (sensitivity/specificity) at the lower and upper bounds of this range, and are there plans to evaluate its utility in younger or older pediatric populations?