AURA — Adaptive Understanding of Response to Adversity

Neurodivergent Ability Profile Framework

Objective

Develop a multidimensional ability profile for 2e‑ADHD+ASD individuals, capturing dynamic cognitive strengths and contextual variability, using AI to integrate multimodal data.

Components

• Cognitive Domains: Verbal comprehension, visual‑spatial reasoning, divergent thinking, working memory, hyperfocus.
• Contextual Variability: Low vs high sensory load, timed vs untimed, interest‑aligned tasks.
• Behavioral/Emotional Correlates: Self‑regulation (latency & error), emotional state (self‑report & physiology).
• Qualitative Insights: Structured interviews; open‑ended strengths/challenges.

Data Collection

• Cognitive Tasks: Matrix reasoning, story creation, gamified memory (PsychoPy/jsPsych).
• Behavioral Data: Video/audio for engagement (eye‑tracking, vocal tone).
• Physiological Data: Wearables (e.g., Empatica E4) for HRV & GSR.
• Self‑Reports: Cognitive experiences + context.
• PCL‑5 Integration: Correlate PTSD symptoms with cognitive variability.

AI Methodology

• Supervised: Random Forest (predict ability/PTSD scores from multimodal features).
• Unsupervised: K‑means for 2e‑specific profiles.
• NLP: BERT‑based analysis for qualitative responses.
• Output: Percentiles + variability index, plus narrative summary.

Validation

• Compare profiles with WAIS/Stanford‑Binet.
• Correlate with functional outcomes (e.g., academics, creative output).
• Validate via clinician feedback and participant self‑reports.

Ethical Considerations

• Tailored consent for neurodivergent communication.
• Diverse datasets to minimise bias.
• Secure, compliant data storage.
• Co‑design with the 2e community.

Implementation Plan

1. Pilot with 10–15 participants; test tasks & models.
2. Refine based on feedback (accessibility & relevance).
3. Scale to 50–100 for robust AI training.
4. Integrate with AI‑enhanced PCL‑5 for holistic assessment.

PTSD

2e

ADHD

ASD

AI