Viranos

Project introduction

At least 30% of people diagnosed with COVID-19 continue to have symptoms more than 3 weeks after initial onset (known as “long COVID”). Our project is a platform for patients, clinicians, and researchers - patients can keep track of symptoms and receive support, clinicians can use the data to help them manage their patient’s condition, and researchers can help us better understand the long term effects of COVID-19. Our software is open-source and can also be extended to other conditions as well.

Issue/Inspiration

Little is known about the long term effects of COVID-19 on patients’ physical and mental health. We were inspired to build a platform to help clinicians and researchers study these effects and provide resources for patients suffering with them.

What it does

We built a mobile app that allows patients to report their symptoms via periodic surveys of their physical and mental health, and share data from wearables and medical records via HealthKit. We parse free-text symptoms (e.g. “I have a lack of energy”) to standardized terminologies (e.g. ICD-10 codes) for clinical and research purposes. We also built a web dashboard which, integrated with an Electronic Health Record system, allows clinicians and researchers to visualize symptom data.

How did your project evolve with the support of the COVID-19 hackathon fund by Google Cloud?

We owe gratitude to our Google Mentor, Shyam Sabhaya, who helped us plan our software architecture and was an amazing resource to us when we ran into stumbling blocks during the development process. We have used several Google Cloud services in our project (listed below) which would not have been possible without the support of the COVID-19 hackathon fund. We are also grateful for this opportunity to demonstrate our work to organizations and individuals who would be interested in implementing it!

How you built it

Our project was initially created in the Stanford Byers Center for Biodesign CardinalKit Buildathon (supported by Google Cloud) where our concept was awarded Best COVID-19 Project. Building on the Stanford CardinalKit framework allowed us to leverage Apple HealthKit, ResearchKit, and CareKit in addition to Google Cloud’s Firebase services. In addition, we built a clinician & researcher dashboard powered by Google Cloud’s App Engine, BigQuery, and Healthcare API. We also utilized the SMART-on-FHIR framework developed at Boston Children’s Hospital to securely exchange data between our app and an electronic health record system.

Challenges you ran into

Integrating our dashboard with an electronic health record system (EHR) was a significant challenge, especially figuring out the technical details of secure data exchange between our system and the EHR. Analyzing free-text survey responses and accurately mapping symptoms described by patients to standardized clinical terminology was also challenging and was facilitated by use of Google Cloud’s Healthcare Natural Language API. Fortunately, this API became available to the public during November 2020.

Accomplishments you are proud of

We are proud of having built an end-to-end solution for researching long term effects of COVID-19 including both a patient-facing mobile app and a clinician/researcher dashboard that can exchange data with an Electronic Health Record system. We are excited to contribute back to the Stanford CardinalKit community and release this project as open-source code for researchers around the world to use and build their own custom solutions from.

What you learned

We learned a lot about managing privacy and security, which is paramount when it comes to dealing with patients’ personal health information. We learned about standardized and validated techniques for assessing patients’ physical and mental health through the National Institutes of Health, and these questionnaires are employed in our app. The COVID-19 hackathon fund in particular helped us to get experienced with using Google Cloud services.

What’s next for your project

We plan to validate the symptom mapping of the Healthcare Natural Language API to ensure that patients’ free-text symptoms are mapped correctly to a clinical terminology. We also hope to deploy our project at a long COVID clinic to assess how usable and feasible the app and dashboard are for patients and their clinicians. Beyond the COVID-19 pandemic, we would like to expand to other use cases! Project introduction: Viranos is a platform for studying the long-term effects of COVID-19 and supporting patients suffering with long COVID symptoms, powered by Google Cloud, Stanford Biodesign CardinalKit, and SMART-on-FHIR.

What Google Cloud products did you use to build your project?

  • Firebase
  • Cloud Firestore
  • Cloud Functions
  • BigQuery
  • App Engine
  • Cloud Healthcare API
  • Healthcare Natural Language API
Profile

Ashley Griffin

University of North Carolina, Chapel Hill

Profile

Vishnu Ravi

Cornell University