See WhatObservationAlone Can't.
NeuroPatch continuously captures five physiological signals from a flexible patch worn on the upper back, providing clinicians with objective sensor data for continuous bipolar disorder monitoring and mood state detection.
Five Signals. One Patch.
NeuroPatch simultaneously captures five physiological signals: EDA, EMG, skin temperature, accelerometry, and cortisol. These streams are processed by a BiLSTM model to detect deviations from each patient's individual baseline.
Built for Clinical Reality
Current bipolar disorder management relies on episodic clinical assessments and self-reported mood diaries. NeuroPatch provides continuous physiological data to supplement these evaluations with objective, longitudinal signal records.
Continuous Passive Monitoring
NeuroPatch samples five physiological channels 24/7 without patient interaction. Data is collected during uncontrolled daily activity, eliminating the gaps inherent to episodic clinical evaluations and self-reported diaries.
Personalized Baseline Modeling
The BiLSTM model builds a rolling physiological baseline per patient from historical sensor data. Deviation scores are computed relative to each patient's own baseline, not population-level thresholds.
Defined Clinical Outputs
The system produces risk scores, baseline deviation metrics, and temporal signal trends. Each output is tied directly to specific sensor inputs and surfaced through a structured clinician portal.
HIPAA-First Architecture
On-device preprocessing limits PHI exposure to derived signal features only. Bluetooth LE transmission is encrypted. The portal enforces role-based access, audit logging, and PHI decoupling at every layer.
Ready to Learn More
About NeuroPatch?
Connect with our team to discuss the system architecture, sensor specifications, or research collaboration opportunities.