Hanu

Hanu is a two-pronged hardware and software solution to provide users with an inexpensive, accessible and reliable device that can assess respiratory health.

Hanu is a complete hardware-and software solution that can (a) obtain a respiratory sound signal via lung auscultation, using the Hanu low-cost digital stethoscope, and (b) subsequently analyze that sound signal using the Hanu mobile app, backed by a multi-model machine learning algorithm, to provide users with a comprehensive assessment of their respiratory health.

According to the World Health Organization, there are 4 million premature deaths each year, resulting from chronic respiratory disease. Despite the severe and widespread nature of chronic respiratory illnesses, many cases are left misdiagnosed, or worse, undiagnosed. These wrongful diagnoses come with serious adverse effects on the patient. Not only do they pose the risk of improper/delayed treatment, but they also increase the likelihood of future exacerbations of the patient’s condition. This effect is further worsened during the COVID-19 pandemic, where it can be difficult for individuals suffering from respiratory illnesses to seek in-person medical attention, thus posing the challenge of creating easily accessible methods for lung auscultation and analysis at-home. In addition, recent studies have indicated that patients of SARS-CoV-2 are at a higher risk of developing Pneumonia, due to damage of the alveoli and surrounding tissues. Moreover, there is evidence that individuals with pre-existing respiratory health conditions are predisposed to severe illness from COVID-19. With such risks comes a greater need for active monitoring and early diagnosis of respiratory complications. Current market costs for digital stethoscopes can range from $75-400, and existing approaches for the diagnosis of lung conditions are often inaccessible due to their high costs and lack of availability in low-resource areas. Therefore, an effective, reliable, and accessible remote screening mechanism for the diagnosis of chronic respiratory illnesses is highly desired. 

Hanu can be, and is intended to be used by any individual who is interested in monitoring their respiratory health. Hanu has been designed with accessibility and convenience in mind; and is suitable for use by a wide variety of individuals. Further, the chest piece used in Hanu's low-cost digital stethoscope includes two interchangeable diaphragms and three open bells, making it effective for everyone - ranging from infants to the elderly. However, my primary focus is to get Hanu in the hands of individuals in under-resourced areas, who lack proper access to healthcare.

I conducted a survey using Google Forms, to gain a better understanding of the preferences and needs of my target market. Questions were asked regarding the struggles faced with current approaches to diagnose respiratory health, what users would find helpful, how prevalent respiratory illnesses are in their families, and how easy they feel it is to obtain sufficient medical attention for respiratory conditions in their communities. Data from the studies showed that the vast majority (~80%) of subjects indicated that they do not believe they have sufficient means for diagnosis of respiratory illnesses at-home. Statistics from the survey can be found summarized below.

Currently, there exists a fully functional proof-of-concept Hanu mobile app, and low-cost digital stethoscope. Hanu has undergone performance testing in the phases described below, and is under further development to optimize performance, utility, and efficiency. 

Phase 1: Data Exploration and Preliminary Review in Create ML 

Phase 2: User Interface and Software Performance Testing in Xcode 

Phase 3: Hardware-Software Integration

Software: Hanu Mobile App

Consists of an iOS app backed by a multi-model machine learning algorithm. Prior to developing the machine learning algorithm, I performed data pre-processing using a Python script that I wrote in Jupyter Notebook. I developed the multi-model machine learning algorithm using Apple's machine learning framework in Swift; Create ML. Then, the 3 trained machine learning models were integrated into an iOS app using Swift programming in Xcode. The Hanu mobile app provides users with 3 key pieces of information: 

• Patient Diagnosis (COPD, URTI, Bronchiectasis, Bronchiolitis, Pneumonia, Healthy cases) 
• Presence/Absence of Wheezes
• Presence/Absence of Crackles

Each output is displayed with its corresponding confidence level in %, which indicates how confident the app is, in the prediction it is making.

While Hanu hasn't yet been launched, we plan to begin piloting Hanu with our local community and through publication of the Hanu mobile app on the App Store. We hope that the Hanu mobile app will reach a minimum of 500 impressions by the end of 2022, and hope to have at least 20 prototypes of the Hanu low-cost digital stethoscope distributed as part of the pilot program. 

I hope to begin distributing free prototypes of the hardware and mobile app installations to pilot in small local clinics, such as One Medical and Sutter Health, and gradually expand to pilot my prototype into more communities before commercialization in larger retailers such as Target. I am in the process of preparing to submit the Hanu mobile app for approval in the App Store and learning about the process of approval for commercialization of medical devices for the hardware prototype. I am also using my knowledge of HTML, CSS and JavaScript to develop a website for Hanu, which I will share with my friends, family, and peers to bring more awareness to Hanu. Although I have already captured a general sense for user needs through my consumer research phase, I plan to integrate a user feedback form in my website, where those who have participated in the pilot program for Hanu, or independently utilized the Hanu app will be able to provide input on how to continue improving Hanu further. Once Hanu is available on the App Store, I plan to market the mobile app using Google Ads.

I hope to measure my progress towards the impact goals I have set for Hanu, through quantitative means, such as the number of hardware prototypes distributed, and number of app impressions, as well as through qualitative means, such as patient reviews and testimonies about user experiences with Hanu. 

Currently, I am the developer and sole team member at Hanu. I foresee challenges in single-handedly deploying the entire Hanu prototype, while working on both backend and frontend. I hope to combat this challenge by actively seeking and promoting Hanu in my community, to recruit additional team members to make Hanu's impact goals more accessible. 

I'm Ashwika Agrawal, a junior at Cupertino High School. I started Hanu after losing my paternal grandfather to COPD. I'm passionate about increasing accessibility to healthcare, and believe that Hanu has the power to make a difference. I hope to see Hanu in the hands of users all around the globe.

Hanu does not currently have any partnerships, however, in the near future, we intend to seek out partnerships with local non-profit organizations focused on increasing healthcare accessibility. 

I am a girl between the ages of 13-18 and my solution is aiming to increase accessibility to a healthcare resource. I would use the funding from the HP Girls Save the World Prize to help advance Hanu by satisfying expenditures and operating costs, and beginning deployment phases. 

I am a youth innovator aiming to improve the lives of all individuals suffering from chronic respiratory illnesses. I would use the Pozen Social Innovation Prize to to help advance Hanu by satisfying expenditures and operating costs, and beginning deployment phases.