<aside>

trained to be clinical EEG expert: scoreAI (autoScore) → integrates with targeted activity and tablet device BLE applications ← instead of trying to be a diagnostic tool for Alzheimer's or dementia, we propose fine tuning clinically approved AI model to become a specialist for explaining and assisting quality of life improving applications

real human brain activity → EEG headbands on pre-dementia care users and long-term dementia care or nursing home residents → our devices with AI → rich training data → precise medicine and personalized wellness plan

</aside>

We mention possible stakeholders like domain-specific AI models, such as AutoScore from Holberg EEG have clinical validation as well and can be used in such non-diagnostic use cases.

Sensor → standardized EEG representation → domain-specific models → activity-specific applications.

Meaning Persimmon Quest does not necessarily have to invent every EEG model itself.

The tablet could eventually act as the orchestration layer through which validated or research-grade algorithms from specialist partners operate, provided those models are validated for the actual hardware, population and intended use.

We will eventually have (on our proposal, we will have to avoid naming hardware models):

qBand

collects longitudinal wearable EEG (need to show why we don’t need a research-grade EEG cap, and a consumer-grade device can effectively work with specific activities via qPad)

↓

PQ Workflow Layer (these are algorithms for light processing before feeding to AI)

signal quality, artifact removal, embeddings, personal baseline, activity context

↓

Specialist models

sleep

relaxation/arousal

attention/engagement research

EEG representation models

future clinically validated modules

↓

qPad experiences