Feelix - Detecting Diseases with AI

Sonavi Labs creates sound-based medical devices and software that harness the power of machine learning to transform the way diseases are detected and managed. Feelix, a suite of screening and monitoring devices, supportive apps and a cloud platform, developed over 4 years of research at Johns Hopkins University, puts the accuracy of a physician in the hands of anyone by deploying advanced algorithms onto a low-cost, user-friendly device. Diseases like asthma and COPD affect hundreds of millions of people and cost families billions of dollars to manage. Pneumonia, a treatable condition, is the leading cause of death in children under 5; the primary reason being a shortage of trained physicians, particularly in emerging markets. Feelix will connect busy patients with the care and education they need, in addition to supporting healthcare professionals in their efforts to reduce costs while expanding access to patients.

Respiratory diseases affect hundreds of millions of people around the world, and create a significant burden on people in densely populated urban centers, as well as rural communities around the world in a variety of ways. Asthma for example is the most common chronic condition of childhood, affecting 14% of children, according to WHO data, over 230 million people around the world. COPD is one of the leading causes of death in adults, and impacts more than 250 million people. Treating and managing these conditions is made more challenging by the global shortage of physicians and healthcare workers. Patients must take off time from work, and some travel great distances for in-person doctor’s visits, only to spend 15 minutes collecting vital signs and engaging with their physician. Increasing access to healthcare services will only be made possible by incorporating smart, connected devices that provide physicians and patients with timely information. Feelix can be scaled across any healthcare system to optimize workflow efficiency and improve patient experiences. When provided to patients, Feelix will reduce readmissions and the need for unnecessary travel, encourage people to be more proactive in monitoring their health, all while improving communication between patients and physicians.

There are hundreds of millions of people who currently lack access to a trained physician. The global shortage of medical professionals creates a significant burden on patients who desperately need adequate and timely care, but also on the overworked, under-resourced professionals. 

We are working with some of the largest hospital systems in the US to develop and validate our technology. Our goal is to ensure that our hardware, in the hands of healthcare professionals, is a practical and useful tool that makes it easier to engage with patients, and create complete, robust electronic health records. We have provided devices to physicians in multiple countries, working with patients experiencing a wide spectrum of conditions, to gather data on the functionality of the technology. We have also sought the feedback of a variety of healthcare professionals, such as physician’s assistants, nurses and respiratory therapists, to determine how to best respond to their disparate needs. 

We continue to conduct clinical studies to develop future iterations of our devices for home users. We will be engaging in considerable dialogue with primary care physicians, patients, and their families to best address the needs of the entire healthcare system.

Sonavi Labs has created Feelix, a solution that harnesses the power of machine learning, applied to body-sound analysis, to detect and manage diseases.

Developed over 4 years of research at Johns Hopkins University, Feelix deploys advanced AI onto a low-cost, user-friendly device that anyone, anywhere can use at any time to detect an abnormality. Feelix uses active noise cancellation to deliver high-fidelity sound quality in any environment, which is especially useful in research settings, but most importantly in obtaining clear screenings that can be included in patient records. We have also deployed a classification algorithm onto the devices that works in real-time to detect abnormalities in body sounds. The classification algorithm has, so far, been validated to have an overall accuracy of 91%, which is the same as that of a trained physician. 

The Feelix hardware was designed to be highly functional with or without power or internet connectivity and can support healthcare professionals and patients in the most remote locations. Because the analytics are completed on the device, the technology can be used anywhere. The hardware was originally intended to provide diagnostic decision support for healthcare professionals and community health workers to help reduce the unnecessary administration of antibiotics.

The devices feature two distinct sensors: an array of microphones and an accelerometer that collect a significant amount of data to help providers develop a full perspective of the patients health. 

The apps and cloud platform we’ve built to support the Feelix hardware offer a host of services and features to all users, including healthcare professionals and patients. Physicians can use the patient management dashboard to monitor treatment efficacy and track longitudinal trends. Patients will also benefit from the insights gained from in-depth trend analysis, in addition to receiving educational content and advice on how to live healthier lives.

We have incorporated Bluetooth to update the software as we continue to enhance the capabilities of our algorithms. The devices can be charged quickly using our portable (wired or battery-operated) base. The charging base also functions as a modem that instantly pulls screenings from the device and incorporates them into electronic health records.

The Feelix platform will help to reduce the administrative burden on healthcare professionals and the financial burden on patients. The Feelix platform was created to benefit those most in need and connect physicians and patients like never before.

While the digital stethoscope is not a novel invention, Feelix is far more superior than any device like it. We have deployed multiple advanced algorithms onto the device, making it the smartest digital screening device in the world.

Feelix is equipped with active noise cancellation that works in real-time to filter out external noise from patient screenings. The noise-cancellation algorithm delivers high-fidelity sound quality in any environment. While other stethoscopes feature passive filtering, which only removes sounds based on a standardized frequency scale, our device actively listens and identifies noises from the external environment, removing them so as not to corrupt the screening.

The classification algorithm that has been deployed onto the device also works in real-time to identify abnormalities in the lungs. No other device has been equipped with on-board analytic capabilities. Many companies are attempting to use AI to analyze patient screenings, however, they all rely on an app to perform the analysis separate from the device. Because we have deployed the algorithms onto our device, it is able to perform complex computations without the use of internet connectivity or its supportive app. This ensures that even in extremely remote locations, or in the event of a major emergency, the devices will perform at the same capacity.

Enabling access to underserved populations requires technology to be adaptive, highly intuitive and, most importantly, functional in any environment.

The Feelix system harnesses the power of artificial intelligence to transform the way diseases are detected and managed.

Our technology incorporates highly sophisticated machine learning algorithms to not only provide high quality sound and instant classification, but we also provide trend analysis and predictive forecasts to help patients manage behaviors to avoid acute incidents. We aim to use AI to help patients and care providers be more proactive in managing chronic conditions in an effort to reduce the likelihood of an emergency.

Additionally, we are building supportive apps that address the needs of our varied customers. The physician’s app will include a dashboard to visualize patient data and quickly identify the most high-risk cases. There will also be a version of the app to provide diagnostic decision support for community health workers (CHW’s) in addition to a version for patients and chronic disease patients that will provide educational content, medication reminders, and notifications of potentially impactful environmental conditions.

By improving the lines of communication between care providers, patients, and family members, we can strengthen communities for generations to come. Not only is the platform intended to support current patients, but we’ve built the entire system to seamlessly receive updates as we improve the technology through further research.

As we continue to collect data from around the world, we will be use our smart cloud platform to develop insights into regional health challenges, helping hospital systems, government agencies and humanitarian organizations allocate resources more appropriately to address specific challenges.

Our solution addresses the needs and concerns of multiple stakeholders across the healthcare continuum, which is why we expect it to be widely adopted.

The Feelix system supports initiatives in hospital systems to reduce spending, increase efficiency and create a better experience for healthcare professionals to reduce burnout.

We are offering healthcare professionals a device that increases accuracy while making it easier to create fuller health records, which will ultimately create a better experience for patients visiting the hospital. Telemedicine programs are rapidly gaining adoption around the world, and many physicians believe that it could help to create a better work-life balance, which could lead to a reduction in burnout.

Payer organizations will also benefit from lower reimbursement costs, which should facilitate wide adoption across health systems. A typical in-person doctor’s visit may cost between $125 - $250, while the typical tele-health visit costs $40 - $50 and saves the patient time in the waiting and in transit. Tracking patients over time also provides more longitudinal data that can help to provide predictive trend analysis, which can potentially reduce the likelihood of an emergency visit. 

We believe the biggest benefit will come to the end user, the patients who will have more information, access, and confidence in their ability to manage chronic conditions from home. Early results from developing telehealth initiatives show promising signs of increased engagement and positive impacts across multiple sectors to include increasing access for patients in rural communities.

With current pilot programs in countries like Peru, Bangladesh, Malawi, and the United States, our devices are being used to collect data on thousands of patients. To validate the classification algorithm embedded on our hardware, we created the world’s largest database of high-fidelity pediatric lung sounds, and collected data on over 3,000 pediatric patients.

In the next year, we plan to launch our physician-support platform in the US, EU, and African markets. With a conservative estimate of 1,000 devices distributed in our first year to physicians and community health worker programs, we will be able to impact the screening of over 2,000,000 patients, assuming the average physician panel contains approximately 2,000 patients.

As we expand distribution and develop additional partnerships with hospitals, global aid organizations and ministries of health we hope to increase our footprint by a magnitude of 30 over the next 5 years, with over 30,000 devices in circulation.

Modernizing urban centers to support population growth will require the use of smart, connected devices. Over 100 million people are driven into poverty every year because of associated healthcare costs. We can support the current efforts of healthcare providers to reduce costs and use our expertise in data management to prepare for the future of care delivery.

Over the next year, we plan to achieve multiple critical milestones that will position the company for long-term success over the following five years.

Our primary milestones for the next year will be to complete the FDA and CE regulatory approval processes and conduct additional clinical trials along with impact studies.

Completing the regulatory process will trigger our ability to begin selling the product globally and by conducting additional clinical trials and impact studies we will be able to significantly increase our value proposition along with quantifying our impact on improving workflow efficiency and patient outcomes.

We have begun building relationships with domestic and international partners that will enable us to rapidly scale our solution across hospital systems in the developed world and in emerging markets. We focused on creating a pathway for the company to generate significant revenue from product sales in developed markets to support the deployment of the technology across emerging markets through community health worker programs and other outreach initiatives.

Quantifying our impact on clinical workflow is also a major milestone we hope to accomplish through multiple impact studies over the next 5 years. These studies will seek to validate our assertions that providing clinical diagnostic decision support will expedite patient treatment, while increasing the accuracy and confidence of a physicians diagnosis.

When we achieve these milestones, we anticipate significant market penetration across frontier and developed markets.

Currently, not yet having received CE marking or FDA approval presents a significant barrier for our company because we are not able to generate revenue from sales. Once we have completed these milestones we will be able to distribute our technology in markets around the world.

Gaining significant market adoption is also going to be a barrier that we will need to overcome once we have received our CE mark and FDA approval. In order for our company to continue developing new technologies, we will need to gain support of the medical community.

Entering multiple international markets also presents a significant barrier when addressing cultural, political and economic differences. We have recognized numerous regions that would significantly benefit from a profusion of our technology but will need to navigate the various nuances that exist when serving these populations.

Developing educational content to support community health worker programs across multiple countries is an example of this.

We are currently preparing our regulatory submission and feel very confident in our pathway and the predicate devices that we have identified. In advance of receiving regulatory approval, we are generating revenue through grants and we have also completed a fundraising round.

We are also building relationships with numerous hospital systems, government agencies and global aid organizations to support our entry into developing and frontier markets.

Working with hospital systems allows us to gain the support of the professional healthcare community. By conducting impact studies, we can prove our impact on workflow efficiency and patient outcomes. The validation from the medical community ensures the wider adoption of our technology for the companies we seek to partner with to help emerging markets.

By working with organizations that are already deployed across frontier markets, we will be able to lean on their experience in these markets to mitigate the cultural nuances that exist when providing care and services.

There are 5 full time employees, including the CEO, CTO, CIO, CFO (also serves as Chief Data Officer) and Executive Director of Business. We have 3 vendors providing contracted engineering, software development, and production services.

Feelix is the brain child of the Whiting School of Engineering , and the Bloomberg School of Public Health, proving that diverse academic teams can spark worldwide change. Our leadership team has come together to harness over 50 years of research, engineering, and business experience to commercialize the technology with a focus on making it accessible to populations most in need.

Ian McLane, our Chief Technology Officer, is one of the principal inventors of the AI at the core of Feelix. His doctoral studies at Johns Hopkins focused on electrical engineering, computer science, artificial intelligence, and signal processing.

Our Chief Innovation Officer, Dr. Ilene Busch-Vishniac, is an expert in acoustics, and is the former dean of the Whiting School of Engineering. She has received numerous patents for her work and continues to deliver talks around the world.

Ellington West, the CEO of Sonavi Labs, brings over a decade of experience in market research, and is a former sales director for a national healthcare organization.

We are building an advisory board, which includes physicians, pediatricians, respiratory specialists, world-renowned acousticians, philanthropists and business leaders.

The team is as unique and diverse as the populations Sonavi seeks to serve.  With a commitment to preserving the original mission that lead to the creation of Feelix, a challenge by the Gates Foundation to reduce pediatric mortality from pneumonia, the Sonavi Labs team is breaking the mold for health technology companies.

Sonavi Labs has multiple partnerships in order to facilitate the continued development of our proprietary classification algorithm and commercialization of the device.

• Children’s Hospital of Los Angeles has partnered with us to provide research support and clinical validation
• Consortium for Technology Innovation in Pediatrics (CTIP) has provided support for our regulatory process
• Johnson and Johnson, we are an affiliate of the Philadelphia Innovation Lab, JPOD, which will help facilitate our entry into developed markets
• FIND is supporting additional research into Tuberculosis diagnostics

The goals of these partnerships are:

• Increase capabilities of core technology
• Refine and enhance functionality to suite nuanced challenges.
• Advise future product development and technology enhancements.
• Facilitate new partnerships and collaborative opportunities.
• Facilitate relationships with Community Health Worker organizations
• Evangelism, communications and marketing support

We have a multi-pronged business model to provide significant value to our customers, which will fit into different sectors. Our business consists of generating revenue through direct hardware sales, subscription software services, and providing data analytics.

Initial Customers include physicians and researchers in the medical community, who will benefit from our devices superior sound quality and data collection capabilities. Providing diagnostic decision support helps to improve efficiency and accuracy within the treatment facility and reduce administrative burdens by allowing physicians to upload audio files directly into a patient record. This will also be a significant value to researchers programs and data collection initiatives. 

Devices will also be made available to patients managing chronic conditions through insurance programs as well as commercially thorough e-commerce platforms. Payer organizations are recognizing the value remote monitoring has on patient experiences and the reducing the overall costs of care.

Subscription services will offer smaller, private practices and larger hospitals management tools that will help to sustain quality communication with their patients. Longitudinal trend analysis also has the potential to reduce the likelihood of emergency incidents by providing more in-depth insights into treatment efficacy.

Providing analytics services to hospital systems, payer organizations and government organizations will also generate revenue to support further research and development and aid with resource allocation decisions.

Sonavi Labs will generate funding through multiple service models, additional grant funding to support further research and an additional capital raise to scale our business.

Capital will be raised through device sales, subscription services and providing analytics as a service. Providing ongoing data management services to hospitals and patient engagement dashboards for physicians will yield enough revenue to support overhead and significantly contribute to research and development of new value propositions.

Additional grants will be sought to advance research initiatives. We will also collaborate with organizations around the world to participate in studies.

To increase scale and capacity of the company, we anticipate completing a Series A funding round after we have completed the regulatory approval process and launched our product. 

The access Solve at MIT program offers would be extremely helpful to our budding startup. Inclusion is one of our company pillars because we know that creating the solutions for the world’s challenges relies on customers collaboration and diverse perspectives. By facilitating introductions to world-class organizations in the MIT network, we will be able to reach a wide audience and get our technology into the hands of leaders and change-makers.

Access is another pillar of the company that drives our efforts everyday. We are seeking access to world

We’re an enthusiastic, diverse team and want to contribute to the SOLVE community. We know that participation in this program doesn’t just benefit our company but we can offer new insights, versatile solutions, and thoughtful collaboration. Because of the range of companies that have been selected in previous cohorts, we are excited about the opportunity to grow alongside the next generation of innovators.

Lastly, as a burgeoning startup, any capital we can raise, and notable partnerships we can develop will significantly improve our efforts to grow our business and solutions.

We will be seeking additional partnerships to deploy our technologies into under-resourced communities and accumulate impact data across health systems 

We would like to partner with hospital systems that connect numerous networks of healthcare professionals. Larger systems, such as Providence/St. Joseph’s, and INOVA Hospitals serve a wide range of patients, in densely populated urban centers and rural communities. These relationships will enable us to service chronic disease patients and members of the provider community, while gaining significant insight into the nuanced challenges of these communities.

We would like to partner with global health organizations and humanitarian initiatives to facilitate our entry into emerging markets. These include:

• Philips has created a Community Life Center (CLC) program to deploy modular clinics across frontier markets. We would like to work with them to provide disease intervention, data collection and analysis support.
• FIND partners with organizations to deploy novel diagnostic solutions across frontier markets and works to fund research initiatives like ours to enhance diagnostic capabilities
• USAID & UNICEF deploys humanitarian assistance programs all over the world.

Sonavi Labs defines as a company that creates solutions, rooted in AI. The stethoscope was chosen as the original method of delivery because it is one of the most widely used medical instrument on the planet and relies on the subjective interpretation of human analysis.

Healthy cities of the future are going to rely on AI to monitor the health and well-being of their citizens.

We recognized that AI would provide the vehicle for objective analysis and enable physicians to deliver more accurate, efficient and effective care. 

We are developing tools that will put AI in the hands of the highly skilled and minimally trained because it provides the consistency and accuracy needed to reduce healthcare costs, while simultaneously increasing access and improving quality. 

The advanced machine learning algorithms deployed into the Feelix platform are designed to be enhanced as we further our research. The funding provided in this prize will be used for those enhancements of our software capabilities and to optimize the supportive apps for providers and patients. The apps, connected to our smart cloud platform, will harness AI to offer a range of functions including medication reminders, environmental trigger warnings, and predictive trend analysis.

The Feelix system was developed to increase access to quality screenings to patients across the world. The technology that resulted has the potential to revolutionize communication the way that the telephone did in the late 19th century. We have worked with physicians around the world, and engaged with business and community leaders to better understand the nuanced challenges of each region struggling to manage respiratory disease.

We first realized that research was significantly stunted because the lack of quality screenings. As a result, we worked with physicians in crowded community clinics to refine and validate the noise-cancelling algorithm on our devices. Now, quality screenings can be easily incorporated into electronic health records or shared with consulting physicians for additional confirmation.

By harnessing the power of cloud and Bluetooth technology, Feelix offers patients the ability to be more proactive in managing their health, and gives physicians the ability to remotely manage patients who are currently struggling to balance the costs of healthcare.

We don’t believe that acute incidents occur without warning signs, so we are going to use AI to develop better understandings of disease trends, environmental triggers, and provide impactful data. Organizations and government agencies that provide the resources, infrastructure support require smart technologies like ours to better manage resources for future generations to come.

We aim to facilitate healthier communities through improved communication and better analysis of the problems at hand.