During epidemics and at other times, it is important for public health officials, individuals, and other entities to make health-related decisions based on sound epidemiological statistics. Such statistics are today generated only when individuals voluntarily seek medical treatment. Diseases that are contagious but initially asymptomatic can spread rapidly through vulnerable populations with little warning. Due to the initial mildness of symptoms, individuals may not realize that they have a dangerous infection and may fail to approach medical professionals for treatment or diagnosis.
This disclosure describes techniques that can leverage sensors, networks, and consumer electronic devices to determine a fine-grained estimate of epidemiological statistics. With user permission, ambient sounds can be analyzed to detect audio symptoms of disease, e.g., cough, changes in voice, etc.; to classify the symptom by disease; and to geographically aggregate disease detections to automatically construct real-time epidemiological maps. Such epidemiological maps can enable accurate public-health decisions and efficient healthcare delivery.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Narayanan, Ajit, "Using Audio Processing To Determine Disease Spread", Technical Disclosure Commons, (July 16, 2020)