The current COVID-19 containment rules call for: a) home isolation for people in contact with infected individuals, for healthy carriers and for paucisymptomatic patients; b) hospitalisation for patients with breathing difficulties and for serious cases requiring intensive care. The goal is to provide an automatic monitoring system for the national health service at a low cost, which is: scalable to COVID centres (both hubs and spokes) and hospitals not specialised in infectious diseases; quick to install and activate; capable of providing rapid responses; reliable; safe in terms of protecting and securing data; able to be converted into monitoring systems for other medical specialisations; trained based on clinical knowledge and decisions; trainable using artificial intelligence techniques; economically sustainable; and with a high social/psychological support impact for citizens/patients.
The automatic detection of COVID-19, the reduction of human-to-human transmission and the prevention of the spread of the pandemic are supported by the following technology: 1) Fingertip pulse oximeter with Bluetooth communication to iOS/Android devices and automatic data storage on a web server; 2) Wristband thermometer with Bluetooth communication to iOS/Android devices and automatic data storage on a web server; 3) Oral flow meter with Bluetooth communication to iOS/Android devices and automatic data storage on a web server; 4) Smartphone voice recorder with automatic data storage on a web server.
The smartphone application “TechnoScience Voice App” allows users to intuitively capture voice samples and stores the data on a web server with a unique ID: the data is only accessible to medical and research personnel and is used to train the supervised machine learning model. Sneezing, coughing, clearing of the throat, shortness of breath and respiratory fatigue are just some of the symptoms detectable by the artificial intelligence model, which monitors the status of the subject’s lungs thanks to its recognition of digital biomarkers (features such as pitch, jitter, shimmer, etc.).
The data from each sensor unit associated with the patient can be accessed with login credentials, allowing specialists to monitor physiological parameters (HR, SpO2, T, peak flow, etc.). An automatic system issues an alert when the physiological values exceed or fall below those for the laboratory and/or diagnostic test.
Together with the clinical indications, the IgG/IgM analysis kit allows for the supervision of the machine learning model. The system is modular in that it can receive inputs from N sensors but produce a single output to support the specialist’s clinical choice.
The wearable system allows for automatic and non-invasive 24-hour monitoring at low costs and ensures that the data remains secure and communicable (cybersecurity along the lines of Cisco Systems and Vodafone). The system has CE certification with medical device marking, can be used at home, and does not need to be modified regardless of the user’s age, gender or technological skills. The data complies with the GDPR.
The system provides a record over time that can be interpreted by clinicians, who have a real-time tool to help them make preventive medical choices. The system provides a greater feeling of safety and monitors parameters of interest for the citizen/patient with regard to COVID-19. The system is supervised by an analysis kit based on IgG/IgM which has a high degree of accuracy, a low cost (tens of euros), extensive availability, fast responses (a few minutes), and a high implementation capacity across the national health service during the COVID-19 emergency. The monitoring system, based on the finite-state automata model, is able to define the staging of the disease in different states, each with a defined aetiology and the implementation of medical treatment. The system allows the gradual progress of the disease to be monitored, not only on the days when it intensifies, but also on previous days with the detail of specific samples from those times: the personalised COVID-19 medical history of each patient can be reconstructed to optimise the intervention strategy, both for an individual patient and for an entire population of homogeneous patients. The system makes it possible to predict when the national health service will become overloaded based on medical staff and the ventilators available.