Haifa’s Rambam Hospital is set to trial speech-based analysis technology by Israeli company Cordio Medical to remotely monitor and diagnose the status of COVID-19 patients based on samples obtained via a smartphone app.
The Or Yehuda-based Cordio Medical developed the HearO technology to monitor patients suffering from heart failure and the platform could prove useful for the sophisticated management of COVID-19 patients who have received a diagnosis and are in isolation but have yet to be hospitalized, the hospital said.
HearO analyzes recordings of a patient’s speech and can sense fluid accumulation and forewarn of an anticipated deterioration, the company says, adding that the tech has already been proven in clinical trials at 10 medical institutions in Israel including Rambam, the Beilinson-Rabin Medical Center, and Barzilai Hospital.
The system provided a pre-deterioration notice on an average of ten days prior to an acute heart failure hospitalization with an accuracy rate of over 80 percent, Cordio says.
The HearO system is in the process of seeking FDA approval as a means to monitor heart-failure patients.
Cordio’s solution for COVID-19 patients is based on adapting HearO to the deterioration process associated with the disease characterized by bilateral pneumonia with edema in the lungs.
The system would pick up on a deterioration based on slight changes in lung fluids and the onset of inflammation before the patient can feel it, Cordio explained.
The system will also be used for remotely monitoring patients who have recovered and are at home.
The trial will monitor “5,000 patients ongoing in all stages: isolated (no symptoms), mild, moderate and severe conditions,” for 18 months, Cordio Medical CEO Tamir Tal tells NoCamels via email.
A production version of the system will be available six weeks into the trial, he adds.
Cordio is concentrating on the lung disease-related symptoms that are not easily trackable. Currently, all other symptoms are easily discovered — for example cough and fever, Tal tells NoCamels
The system works by establishing a baseline for each patient after which they will record their voices daily “and we analyze it in our cloud-based system. Each day the system is not alerting, the baseline is adapting and adjusting utilizing the new info,” he says.
Tal explains that if successful, the HearO system will be able to alert the medical team about the start of a lung condition 10-48 hours before the patient’s condition becomes moderate or severe due to the infection.
As of April 1, Israel has over 6,000 confirmed cases of COVID-19, including over 200 recoveries and 25 deaths. Most of those diagnosed with the disease are in isolation at home or in hotels with a minority in hospital.
“Cordio’s novel solution is designed to make remote monitoring of COVID-19 patients more efficient, to identify with the use of a user-friendly mobile app those whose condition is deteriorating and need to be hospitalized and to reduce the pressure on hospitals,” the company said.
The clinical trial will be led by Dr. Gidon Berger and Professor Zaher Azzam from Rambam Hospital and overseen by Professor Ilan Shallom, CTO at Cordio.
“The solution for remotely diagnosing the condition of COVID-19 patients is based on a proven technology that can be quickly adapted for the new pandemic,” said Professor Chaim Lotan, medical director at Cordio, and the former director of the Heart Institute at Hadassah Hospital.
Professor Lotan said that within a few weeks, expected statistical projections will see Israel with tens of thousands of coronavirus patients.
“If we succeed through the assistance of Rambam Hospital in proving the efficiency of the system, we will be able to establish a fully operational network within about two months,” he said.
Tal said in a hospital statement that the “pandemic requires the efficient managing of the resources of the health system in order for it to handle the thousands of patients whose number is only increasing and to hospitalize only those with the highest risk.”
“The trial will lead to a turning point in the ability of the health system to efficiently manage the struggle against the pandemic,” he added.
Dr. Michael Halbertal, the director-general of Rambam Hospital said that the previous successful trial conducted with Cordio “gives us hope that we will be able to successfully adapt a similar system for identifying the deterioration in COVID-19 patients who have not yet been hospitalized or who have been discharged to their homes suffering with only a minor case or after complete recovery.
Rambam, like other health institutions, is quickly adapting telemedicine and digital health technologies capable of diagnosing diseases, conserving resources, and reducing risk and the pressure on medical teams.”
“Solutions like those of Cordio are especially critical for dealing with the coronavirus which confronts health institutions with complex and large-scale challenges that have not been witnessed before,” indicated Dr. Halbertal.
Voice-based analysis for coronavirus
Separately, Israeli startup Vocalis Health announced last week that it was launching a voice-based test study to triage, screen and monitor for COVID-19 symptoms.
Vocalis developed an AI-based platform that uses voice recordings to detect and monitor health and will now use all its resources “to try to find a unique vocal ‘fingerprint’ that will be able to detect a deterioration in the health of coronavirus patients as well as flag potential symptoms,” the company said.
COVID-19 symptoms include fever, a cough, and shortness of breath and Vocalis believes these symptoms “can be logged through voice tasks like phonation (‘ahhh’) and free speech (e.g. ‘describe your day from start to finish’) – and could potentially be used to screen for COVID-19 symptoms early. ”
Vocalis says that “by analyzing the voice beyond what the human ear can hear, we can unveil dedicated vocal biomarkers that will enable the healthcare community to get insights on the symptoms and hopefully the onset of the COVID-19 virus.”
One part of the study will be led by the Israeli Defense Ministry where Vocalis will work with hospitals and academic institutions to sample the voices of confirmed coronavirus patients through a mobile app.
Another part of the study will collect voice samples from the general population at home and the company asks that people donate their voice.
Vocalis hopes to enroll at least 100,000 patients to collect as much data as possible. Participants will be tracked weekly for a month.
Both parts of the study began last week and will end in December.
Vocalis researchers hope to rapidly build a point-of-care diagnostic to save lives and prioritize testing and hospitalized treatment.
Monitoring, surveillance, detection and prevention of COVID-19
First, the IoT provides a platform that allows public-health agencies access to data for monitoring the COVID-19 pandemic.
For example, the ‘Worldometer’ provides a real-time update on the actual number of people known to have COVID-19 worldwide, including daily new cases of the disease, disease distribution by countries and severity of disease (recovered, critical condition or death) (https://www.worldometers.info/coronavirus/).
Johns Hopkins University’s Center for Systems Science and Engineering has also developed a real-time tracking map for following cases of COVID-19 across the world, using the data collected from US Centers for Disease Control and Prevention (CDC), the World Health Organization (WHO), the European Center for Disease Prevention and Control, the Chinese Center for Disease Control and Prevention (China CDC) and the Chinese website DXY, which aggregates data from China’s National Health Commission and the China CDC (https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6).
Second, big data also provides opportunities for performing modeling studies of viral activity and for guiding individual country healthcare policymakers to enhance preparation for the outbreak.
Using three global databases―the Official Aviation Guide, the location-based services of the Tencent (Shenzhen, China), and the Wuhan Municipal Transportation Management Bureau―Wu et al. performed a modeled study of ‘nowcasting’ and forecasting COVID-19 disease activity within and outside China that could be used by the health authorities for public-health planning and control worldwide8.
Similarly, using the WHO International Health Regulations, the State Parties Self-Assessment Annual Reporting Tool, Joint External Evaluation reports and the Infectious Disease Vulnerability Index, Gilbert et al. assessed the preparedness and vulnerability of African countries in battling against COVID-19; this would help raise awareness of the respective health authorities in Africa to better prepare for the viral outbreak9.
Third, digital technology can enhance public-health education and communication. In Singapore, the government has partnered with WhatsApp (owned by Facebook) to allow the public to receive accurate information about COVID-19 and government initiatives (https://www.form.gov.sg/#!/5e33fa3709f80b00113b6891).
Multiple social-media platforms (e.g., Facebook and Twitter) are currently used by healthcare agencies to provide ‘real-time’ updates and clarify uncertainties with the public. Additionally, some facial-recognition companies (e.g., SenseTime and Sunell) have adopted the thermal imaging–enabled facial recognition to identify people with an elevated temperature at various screening points in China (https://apnews.com/PR%20Newswire/354aae0738073bc95331ee72a458cb50).
Fourth, AI and deep learning can enhance the detection and diagnosis of COVID-19. The need to provide access to accurate and low-cost tests for the diagnosis of COVID-19 is a challenge. Many peripheral hospitals in China and other developing countries in Asia, the Middle-East and Africa do not have the tests or resources to accurately distinguish COVID-19 from the ‘common flu’.
In Indonesia, which has only two reported case thus far, despite substantial exposure to Chinese tourists (Bali had 1.2 million Chinese tourists in 2019), health authorities decided against testing the 243 returning but asymptomatic citizens from Wuhan because of cost of the test (the reagent was quoted as costing nearly US$75,000).
Alternative diagnostic and screening tests for COVID-19 will be extremely useful. In this context, China has large datasets of cases positive for COVID-19 (>70,000 cases). These are ideal datasets for deep AI and deep learning (https://www.wired.com/story/chinese-hospitals-deploy-ai-help-diagnose-covid-19/).
Such AI algorithms can then be used as an initial screening tool for suspected cases (e.g., travel history to China, Iran or South Korea, or exposure to confirmed cases) so that patients at higher risk could have confirmatory laboratory-based tests or be isolated.
Although most patients have mild cases of COVID-19, physicians have to apply the same level of intensive methods to isolate, treat and monitor all patients. AI algorithms could be developed to help physicians triage patients with COVID-19 into potentially three groups: the 80% who have mild disease; the 15% who have moderate disease; and the 5% who have severe disease, including those at high risk of mortality. Finally, AI can also facilitate the discovery of novel drugs with which to treat COVID-19.
Mitigation of COVID-19’s impact
Although the focus of tackling the direct impact of COVID-19 is important, in many healthcare settings, it is important to maintain core and critical clinical service. The initial reaction in many countries is for healthcare facilities to reduce or even cease many clinical services, including closure of clinics and postponement of medical appointments or elective surgeries. However, such strategies cannot be sustained indefinitely if the COVID-19 pandemic extends beyond 6 months.
Healthcare systems should plan to use digital technology. For example, ‘virtual clinics’ could be set up through the use of tele-medicine consultations with imaging data (e.g., chest X-ray and/or CT of the thorax) uploaded from peripheral sites and interpreted remotely.
This would ensure that patients continue to receive standard clinical care while reducing physical crowding of patients into hospital premises. For other key hospital activities (e.g., research and education), virtual e-learning platforms are increasingly being explored to eliminate physical meetings.
Second, the utilization of various AI-based triage systems could potentially alleviate the clinical load of physicians. An online medical ‘chat bot’ could help patients recognize early symptoms, educate people on the importance of hand hygiene and refer people for medical treatment should symptoms worsen.
Additionally, phone-based software that detects and records patients’ data (e.g., daily temperature and symptoms) may prevent unnecessary hospital consultations for patients with mild flu-like symptoms. These data could also be developed into AI algorithms for the detection of COVID-19.
Third, many hospitals in China are collaborating with blockchain companies and pharmacies to deliver patients’ medication to their doorsteps. Through the use of blockchain, hospitals could ensure timely delivery of medications with accurate tracking.
In summary, while the world continues to rely on classic public-health measures for tackling the COVID-19 pandemic, in 2020, there is now a wide range of digital technology that can be used to augment and enhance these public-health strategies (https://www.vox.com/recode/2020/2/27/21156358/surveillance-tech-coronavirus-china-facial-recognition).
There is also a longer-term goal. The immediate use and successful application of digital technology to tackle a major, global public-health challenge in 2020 will probably increase the public and governmental acceptance of such technologies for other areas of healthcare, including chronic disease in the future. As the saying goes, ‘a crisis provides an opportunity’; this first great crisis of 2020 provides a great opportunity for digital technology.