EarliPoint System

The EarliPoint System is a diagnostic aid for Autism Spectrum Disorder in children aged 16-30 months, utilizing eye-tracking technology to record visual responses to video-based social stimuli. Artificial intelligence software analyzes these responses to generate an ASD diagnosis and severity indices correlated with established clinical assessment tools.

The EarliPoint System is a medical device designed to assist clinicians in the diagnosis and assessment of Autism Spectrum Disorder (ASD) in children aged 16 to 30 months. The system functions by using an eye-tracking device to capture a child's visual responses while they watch a series of age-appropriate videos containing social information. This eye-tracking data is then analyzed by artificial intelligence (AI) software, which provides a diagnostic output for ASD. The system comprises an EarliPoint WebPortal for patient data management and results access, the EarliPoint Device with its eye-tracking capability, and the AI software. In addition to a diagnostic determination, the software also generates three EarliPoint Severity Indices: a Social Disability Index that correlates with ADOS-2 scores, and Verbal and Non-verbal Ability Indices that correlate with the Mullen Verbal and Non-verbal Ability scores, respectively. A pivotal clinical study involving 475 evaluable patients demonstrated the EarliPoint System's performance by comparing its diagnoses to those of expert clinicians. In the modified intent-to-diagnose population, the system achieved a sensitivity of 71% and a specificity of 80.7%. Performance was higher in a subgroup where clinicians were more certain of their diagnosis (78.0% sensitivity, 85.4% specificity). The study also indicated good repeatability and reproducibility for the device's diagnostic output. Notably, performance differences were observed across subjects based on sex and race/ethnicity. Unlike a predicate device which often provides indeterminate results, the EarliPoint System is designed to output a definite diagnosis of positive or negative ASD for each patient. The eye-tracking technology uses near-infrared spectrum light, similar to established video-oculography systems, and operates with a sampling rate of 120 Hz.

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 were the 'age-appropriate videos' presenting social information developed and validated to ensure consistent and relevant stimuli for the 16-30 month age range?; What are the specific factors that contribute to a clinician's diagnosis being 'UncertainDx', and how does the EarliPoint System's performance in this subgroup inform its utility in complex cases?; The document mentions 'performance differences across subjects based on sex, race/ethnicity'. What is the magnitude of these differences, and what strategies are being considered to mitigate potential biases or improve performance across diverse demographic groups?; Could the EarliPoint Severity Indices (Social Disability, Verbal Ability, Non-verbal Ability) be used to track developmental changes or response to interventions over time, beyond their initial diagnostic correlation with ADOS-2 and Mullen scores?; Given the 0.5° accuracy/precision for gaze position, how does this level of precision impact the AI's ability to differentiate subtle visual response patterns indicative of ASD, especially when compared to normative data?