CoughAPP offers a wearable hardware module equipped with a high‑sensitivity microphone array and an on‑device AI accelerator that detects, classifies (dry vs. wet), and quantifies cough events in real time. All audio processing occurs locally for privacy, with encrypted logs and optional cloud sync via FHIR‑compatible APIs, allowing clinicians to monitor respiratory conditions continuously through integrated EHR and tele‑medicine platforms.
Funding
Funding not disclosed
Founders
Product
Problem
Patients with chronic respiratory conditions or limited mobility often lack objective, continuous monitoring of cough, leading to delayed diagnosis of infections or disease exacerbations. Traditional cough assessment relies on sporadic clinical visits and subjective reporting, which can miss early warning signs.
Solution
CoughAPP combines a dedicated hardware module—high‑sensitivity microphone array and an on‑device AI accelerator—with edge‑optimized deep‑learning models to detect, classify (dry vs. wet) and quantify cough events in real time. All audio processing occurs locally, preserving patient privacy while delivering instantaneous feedback through a mobile UI. The system logs each event, generates trend analytics, and securely syncs aggregated data to a cloud service for clinician review. Integrated APIs expose FHIR‑compatible endpoints, enabling seamless incorporation of cough metrics into electronic health records and tele‑medicine platforms. By providing continuous, objective cough biomarkers, CoughAPP supports early intervention for asthma, COPD, pneumonia and other respiratory disorders.
Target Audience
Primary users are individuals with asthma, COPD, or other chronic respiratory diseases—including seniors—and the healthcare providers (primary‑care physicians, pulmonologists, tele‑health services, and long‑term care facilities) who require objective cough data for diagnosis and treatment monitoring.
Features
- On‑device convolutional neural network inference delivering < 100 ms latency per cough event
- Dual‑microphone acoustic front‑end with noise‑cancellation tuned for speech‑range frequencies
- AI accelerator (edge‑optimized chip) enabling low‑power operation and battery life > 12 hours
- Real‑time cough classification (dry, wet, severity) with 84 % accuracy on clinical validation set
- Local data encryption and end‑to‑end TLS for optional cloud sync, ensuring GDPR‑compliant privacy
- Mobile application UI with customizable alert thresholds, multilingual support, and accessibility modes
- Longitudinal cough event history, trend charts, and export of CSV/JSON reports via FHIR‑compatible API
- Energy‑management firmware that dynamically scales AI workload based on activity level