Heartune

Supplementary tool for cardiologists and music therapists by passive music therapy tuned to the heartbeat of an individual.

Heartune is intended to be used as a supplementary tool to cardiologists and music therapists. It can also be used as an individual tool for mapping music to specific activities and tracking based on heart rate. Its function lies in mapping specifically-curated music (pre-processed) to the heartbeat intervals of the user. This is done through a combination of technologies and algorithms, namely a Joint Classification Regression Neural Network, connection to APIs, data security methods, and database operations. The application will be accessible across platforms, such as Android and iOS, as well as a Web App. The app will be integrated with wearables compatible with Android and iOS devices. The sending and syncing of data for Android devices will be facilitated through the Wearable Data Layer API, the Google Play services. Similarly, for iOS devices, the Healthkit, provided by Apple, will be used to access the heartbeat data collected and fed to the algorithm. The application, across all platforms, requires internet access for access to the extensive music repository. The repository is then processed through the classification algorithm, sorted by genre and activity, and the results are provided to the regression algorithm. (Further technical described in the following questions.)

The problem that Heartune aims to solve is the severity of the impact of several diseases on the lives of patients suffering from a given set of diseases. The technology we are trying to implement utilizes the patient's heartbeat and maps it onto the Fourier domain to obtain its frequency and intensity. Our application uses this technology alongside music files that possess a specific frequency known as 'beats per minute', which synchronizes with the listener's heartbeat when played. When coupled with passive music therapy, the application becomes a successful method to provide the supplementary treatment of several diseases, including but not limited to Alzheimer's, Autism, ADHD, Mild Depression, and Developmental disorders. With over 100 million people facing these diseases and the rate of developmental disorders increasing rapidly within the Indian rural population, our solution aims to better their lives. The consequences of the disorders are the impact on the patient's lives that remain forever impacted and their family and community that may be adversely affected physically and emotionally. The solution aids in reducing the severity of these disorders by serving as a supplementary tool to pre-existing treatment. With treatment with a high success rate for improving its patients' well-being, passive music therapy, when administered with Heartune, ensures that the treatment is complete and thriving based on the algorithms encoded within.

Our MedTech solution is an application that is open to all users. Our target audience would be the patients currently suffering from ADHD, Alzheimer’s, Autism, Developmental Disorders, and Mild Depression, and their family members, friends, or caretakers. As the application poses to provide supplementary treatment to this group of people, the solution proposed will look to support the lives of hundreds of millions of patients worldwide. The target demographic would thereby benefit from this supplemental treatment which is said to increase creativity, lighten moods, increase mental capacity and have several other health benefits otherwise unattainable easily, as per published scientific research. By using algorithms to understand the various types of music that work to reduce the severity of several diseases, we would not only be able to provide supplementary treatment to the various patients but also use the data that is collected within the algorithm of the application to generate a more efficient solution. 

The current population that faces diseases such as developmental disorders and ADHD are severely underrepresented within India especially within the rural areas. Thereby, the application would be providing a standardized supplemental treatment to those patients within India.

Having worked alongside a specialist doctor, we understood the different aspects and applications of Heartune. We reached out to potential users of the application and received positive feedback on their future utilization. Alongside this work, we have also written an extensive whitepaper that is a technical overview and evaluation of the application; this paper analyses over 30 research studies conducted that administer passive music therapy to different groups of people. Moreover, the report aims to verify the feasibility of utilizing the novel application Heartune as a supplemental treatment for Autism, Alzheimer's, Depression/anxiety, and developmental delays using passive music therapy. The studies and meta-analyses discussed within the whitepaper illustrate the ability of music therapy in general to benefit the well-being and behavioral tendencies of those diagnosed with the aforementioned illnesses. For the design and development of the application, we have currently designed the UI/UX of the application and are working on the backend of the application. Our advisory panel includes a doctor who has worked closely with Autism, Alzheimer’s, and ADHD patients. We are currently talking to several developmental pediatricians to onboard them onto the application.

At present, the application is currently ready for Pilot testing with a blind group of volunteers but is pending Institutional Review Board (IRB) approval to conduct testing. Upon receiving approval, the Pilot will be conducted with two sets of twelve individuals to determine the efficacy of the application. The first investigation will be conducted with individuals who do not have developmental disorders and will simply be monitoring visible responses to changes in the music based on the user’s heartbeat. The aim of the first investigation is to test the working of the product and the general user response. Plans for secondary and tertiary investigations are in place to test the use of the application as a supplementary tool for music therapists and cardiologists and will be done under the supervision of a panel of experts. The results of the investigation will be documented and published for public access. A whitepaper detailing the purpose, usage, technology, and potential effects of the application has been submitted for publishing in the International Journal of Research in Medical Sciences and Technology. 

The functionality of the Heartune App is powered by a combination of technologies and algorithms, namely a unique neural-network, series of APIs, differential privacy protocols and standard AES encryption. The application is powered by a model that utilizes two classification algorithms and a regression model that performs the functions sequentially. The model utilizes deep learning and libraries such as TensorFlow and Keras. The prediction problems surfaced by the application require both numeric and class label values for each input. 

The essential features of the model are (A) Classification of ECG-based heartbeat data wherein data received from the wearables are classified into normal and abnormal ranges as per the standards outlined by the Association for the Advancement of Medical Instrumentation (AAMI). Given that this requires individualized data from the user, this is a local task and is unique to the device. (B) Classification of Music Dataset to Genre/Activity using Deep Learning, Transfer Learning and TensorFlow. Through the combination of pooling layers and 1D convolutional layers, key features can be extracted from the raw audio, and the genre and mood can be classified. The genre and mood are run through a separate classification algorithm to map it to the activity. This is run as a global task and is not unique for the user or device. (C) Regression of results from classification wherein a Deep Neural Network was utilized to map the pre-processed music dataset to the segmented heartbeat intervals and monitored at fixed intervals.

With the solution being extremely beneficial to over 100 million people worldwide, our projections estimate that we would be able to capture a small fragment of the market whilst working alongside developmental pediatricians, music therapists, and other doctors. Through recommendations from clinics and hospitals, our application would gain traction as an application usable for medical purposes and recreational purposes. As this administration of passive music therapy requires a high level of medical expertise, our solution will work with the necessary professionals. Therefore, we estimate our reach to be approximately 50,000 to 75,000 users within the next year. As the application is within the MedTech field, we would also have to ensure that there are no flaws within the algorithms and obtain the required certification before administering passive music therapy to patients that require it to significantly reduce the impact of the disease currently battling. Our prediction for the total number of people arises from the results and responses we are currently receiving from our Beta testing and based on results from other med-tech startups within the same community that received recognition and funding from various institutions. Moreover, due to the versatile nature of the passive music therapy treatment, there are several patients that we can impact and their family members, caretakers, and friends that would be better off emotionally.

Our primary impact goal is to better the lives of patients who have ADHD, Alzheimer’s, Autism, Developmental Disorders, and Mild Depression by boosting their creativity and elevating their mood through the daily provision of passive music therapy. As evidenced by the studies previously conducted, the transcendental impact that passive music therapy has had on the lives of the test group has proven the ability of the supplementary treatment to alter people’s lives. From this impact goal, which would be achieved once the application is available to the market, the Heartune team can generate its subsidiary impact goals:

1. To publish results of this impact that the application would have on the lives of the aforementioned patients. By collecting data for over 75% of our users and running the given data through several layers of data privacy, we would publish the essence of our findings without providing any localized or specified information to any of the patients. This can be done through Differential Privacy. 
2. To change how all users listen to music and their experience of hearing their favorite songs. This minute altercation would make the experience more satisfying as the heartbeat of the listeners would be immediately synchronized with the music being played. Furthermore, the user would be able to regulate the user’s heartbeat by controlling the BPM of the music played to a given limit, thereby allowing them to elevate their mood with slight alterations.

The central United Nations Sustainable development goals that we would aim to accomplish would be the betterment of life to promote good health and wellbeing. The impact that we would have on patients' and non-patients lives would thereby be measured using a set of surveys and collecting data about a user's navigation of the application and their habits. We would use a set of privacy protocols incorporating AES encryption, Differential Privacy, and Secure Multi-Party Computation to comply with standard HIPAA data privacy laws. We would then analyze the aggregated and anonymized data to find pertinent trends and patterns for further analysis. We would successfully achieve our goal if we saw positive indicators that the application benefits patients' lives in the longer run. Not only would this indicate success in the initial rounds of the therapy, but it would also indicate the application's ability to better the lives of its users.

For recreational users, metrics such as average time spent on the application, average listening session, average heart rate of users, and more would help understand the usage patterns and efficacy of the application beyond the medical scope. We would also put the aforementioned data through a similar differential privacy algorithm to make the data entirely anonymous. This would help us measure our solution's progress in bettering their music experiences.  We are working towards obtaining legal approval for the administration of this solution on a broader scale. 

The significant barrier in our way would be the regulatory barriers for any medical supplementary treatments. This legal barrier would pose the most significant threat as we enter the market and based on the different fields for which we receive approval, the impact of the application would change. Another legal barrier would be the publication of anonymous data that has gone through several layers of data privacy. Moreover, as the current startup has minimal seed capital, finances may eventually also become a problem when expansion begins to occur due to the high amount of cloud storage that would be required and higher computing power that would be necessary. The application would also have to have a higher level of overall safety. Information about the application would have to be spread through credible sources to overcome any cultural or market barriers that may be in the way of our growth. Thus, the impact of Heartune would depend on the various aforementioned factors. We are currently working on minimizing the effects of these factors on our development and have begun associating with doctors to prevent market and cultural barriers from arising.

Ishaan Singh is a 17-year-old high school student in Mumbai with advanced knowledge and experience in the MedTech field. Having developed his own mental health startup, Inaya Connect, his experience proved crucial in navigating the challenges of running a MedTech startup. In the past, he has worked and partnered with organizations such as MIT PathCheck Foundation (being a finalist in the MIT Solve Health Securities and Pandemics Challenge), IEEE, Linux Foundation, and more. He is also the founder and youngest member of a blockchain organization, funded by the Ocean Protocol, that specializes in developing decentralized apps for the MedTech industry and is currently incubated by Deep Medicine Labs. 

Siddhant Sukhani is an 18-year-old student specializing in mathematics and medical knowledge. Whilst partnering with a clinic within Mumbai and 4 developmental pediatricians, he has garnered experiences within the MedTech field as he previously developed an application for the promotion of early intervention within the Indian subcontinent. Having dealt with cases of Autism, ADHD, and other developmental disorders that arise as a result of developmental delays, he is equipped with the expertise to understand which treatments prove to have maximum efficacy for these given groups. Moreover, his experience with developing 3 Medtech devices currently in the prototyping alongside a Ph.D. student for the treatment of blindness, sleep apnea, and type 1 diabetes further qualifies him to develop an application of this caliber. Both lead members of the team have received recognition not only within their school but also through several other awarding organizations for their ability to develop startup organizations.

We are not currently affiliated with any organizations but are currently speaking with several clinics and treatment centers to partner up.

N/A

N/A