Building for Digital Health
Overview
| Date/Time |
Tuesdays and Thursdays at 5:30PM - 7:00PM
2022 - 2023 Winter Quarter |
| Location | MSOB 303 + Zoom |
| Units | 3 Ltr (CR/NC and Med option available) |
| Instructors |
Oliver Aalami, M.D. (aalami@stanford.edu)
Carlos Guestrin, Ph.D. (guestrin@stanford.edu) Vishnu Ravi, M.D. (vishnur@stanford.edu) Paul Schmiedmayer, Ph.D. (schmiedmayer@stanford.edu) Ashley Griffin, Ph.D. (griffina@stanford.edu) |
| Project Coaches |
Raghav Samavedam (raghavsa@stanford.edu) Varun Shenoy (vnshenoy@stanford.edu) |
| Office hours |
Oliver Aalami - Mondays 2-3pm via Zoom Vishnu Ravi - Fridays 4-5pm via Zoom Paul Schmiedmayer - Wednesdays 10-11am at Biodesign Office |
| GitHub | http://github.com/cs342 |
| Syllabus | View Course Syllabus on Google Docs |
Introduction
As our world becomes more and more digitized, patients and their devices are generating streams of valuable data that can provide meaningful clinical insights. This digital health revolution provides great opportunities to design and validate new digital health concepts. Many groups within Stanford Medicine have promising ideas that are ripe for development, however, they lack the software engineering and healthcare compliance know-how to take them forward.
Building for Digital Health is a Biodesign course sponsored by the Stanford School of Medicine (SoM) and Stanford's Computer Science (CS) department. Its goal is to provide CS students with the opportunity to apply their skills to real-world health technology development projects, while enabling SoM faculty to leverage these talented individuals to help advance their technology concepts toward patients. Both audiences will learn a repeatable approach for developing new digital health technologies and preparing to launch them in the market.
Over the course of ten weeks, students and faculty will work together to tackle a project and launch an app-enabled solution for research use. Every week, students will learn about app-development, sensor technologies, privacy, security, and more. In the final week of class, teams will present their final project (app) to a panel of digital health experts.
Goals
- Gain exposure to unmet needs in the healthcare field and learn to design and build secure digital solutions to address them.
- Get experience working with technical and multidisciplinary teams with the help of external mentors and industry experts who can help guide career paths in health tech innovation.
- Become familiar with the iOS platform, the Swift programming language, the use of CardinalKit and other open-source frameworks, and secure data collection practices.
Projects
Students will be able to work on one of five projects for the Winter of 2023. The participating projects will present their challenges during the first lecture. Everyone will have an opportunity to submit a form of interest with their preference during the first day of class. We will then create balanced teams based on this input.
Vascular Surgery Patient Reported Outcomes
Mentors
Benjamin Brooke, MD, PhD, Professor and Chief of Vascular Surgery
Teryn Holeman, MS, MD/PhD Candidate
Julie Hales, MSN, RN, CCRC, Research Associate
University of Utah, Division of Vascular Surgery
The University of Utah provides care for a wide region (Mountain West) covering many remote areas making access to care and coordination of care challenging. The system is government funded and has a history of collecting patient reported outcomes (PROs) in person to assess quality of life and functional capacity measures. The Division of Vascular Surgery is interested in expanding this program via a mobile application to achieve broader adoption, increase accessibility and to gain improved longitudinal data. This project will include building the application in iOS and will include 1) consenting, 2) collecting HealthKit data metrics (including passive activity), 3) Deploying PRO surveys, 4) Deploying active tasks such as the remote 6-minute-walk test, among other features.
Pediatric Apple Watch Study
Mentors
Scott Ceresnak, MD, Professor of Pediatric Cardiology
Aydin Zahevidash, MD, MBA, Pediatric Resident Physician
Stanford University, Pediatric Cardiology
Arrhythmia symptoms are frequent complaints in children and often require a pediatric cardiology referral. Wearable technologies, such as the Apple Watch®, have been studied in adults but there are limited data on the utility of these wearable technologies in children. We hypothesized that an Apple Watch can be used to identify clinically significant arrhythmias in children. Since 2018, the Stanford group has had 41 pediatric patients detect arrhythmias on an Apple Watch which were later confirmed through traditional clinical methods. Dr. Ceresnak’s group would like to develop an iOS research application which leverages the Apple watch to collect ECG tracings in pediatric patients. Their application will have the standard 1) onboarding consent, 2) PRO/survey collection, 3) Apple Watch ECG Active task, among other features.
Balance (Management of Eating Disorders)
Mentors
James Lock, MD, Professor and Vice Chair, Department of Psychiatry
Brittany Matheson, PhD, Clinical Instructor, Psychiatry and Behavioral Sciences
Nandini Datta, PhD, Post-Doctoral Scholar, Eating Disorders Research Program
Stanford University School of Medicine
Eating disorders are a common condition many teenagers and young adults suffer from. Sometimes the condition requires a hospital admission to help manage this condition. Dr. Locke’s group specializes in the treatment of anorexia and bulimia. In particular he studies dialectical behavioral therapy treatments and is looking to build a companion application to support patients in-between therapy sessions- which is most of the time! For this project you will build a new iOS app which will include, 1) Onboarding/Consenting, 2) Survey assessments/PROs, and several others which could include 3) setting up an “SOS” button, 4) a “Distract Me” feature, 5) a “Let’s Chill Out” feature, 6) a “Feeling Learning” feature, or 7) a Diary feature.
GetMoovin
Mentors
HealthSpan Heroes Team (BIOE273 NEXT Award Winners)
Jack Keene, MD
Ank Agarwal
Ben Randoing
Natasha Kacharia
GetMoovin is an app/digital system focused on motivating seniors to engage in physical activity to improve their healthspan. The aging process decreases one’s muscle strength, agility, and cardiovascular fitness leading to frailty and chronic disease. GetMoovin’s free app engages families and friends in synchronous photo curation of physical activities. The subscription side provides customers personalized plans to mitigate the decline of aging so they will be able to do the things that really matter to them as they age. We want to give people the Gift of a Better Healthspan.
Remote Assessment of Contact Allergies
Mentors
Golara Honari, MD, Professor of Dermatology
Ali Mottaghi, PhD Candidate in Electrical Engineering
Serena Yeung, PhD, Assistant Professor of Biomedical Data Science
Haiwen Gui, MD Candidate
Stanford University School of Medicine
Eczematous dermatoses are a group of common inflammatory skin disorders, also known as dermatitis. Currently, patch testing is only offered in specialty clinics, and the procedure typically requires three visits to the specialty clinic within a week. The limited number of specialty clinics nationwide and the fact that specialty clinics are mostly concentrated in large cities and academic centers make access to care very limited. Only a small portion of patients who benefit from patch testing have access to this test. A cost analysis study of patients with eczema shows that among 2.6 million patients with eczema and continuous enrollment at the IBM market scan dataset between 2014-2016 only about 2% had patch testing. Patient with patch testing had an average 50% reduction in direct cost of care in the year following patch testing compared to the prior year. This project aims to facilitate access to care through a multimodal digital health solution for patients with chronic eczema, facilitate patch testing from a remote location, and interpretation of patch test results through AI augmented analysis of sequential photos from the allergen application site.
Schedule
Week 1 (1/10 & 1/12)
- Syllabus overview, course website, grading, attendance, and guest lectures.
- Discussion of topics covered, assignments, Health Insurance Portability and Accountability Act (HIPAA) training, and delivers.
- Introducing our projects for this quarter (need statement presentation).
- Xcode, the Swift programming language, and Swift UI
- Swift Language Guide
- SwiftUI Tutorials
Week 2 (1/17 & 1/19)
- Data privacy, personally identifiable information (PII), and data-risk assessments (DRA).
- HIPPA-compliance, Institutional Review Board (IRB) protocols, and the consent process.
- What does it mean to develop reusable software?
- Testing & Linting, Code Quality
- How do you develop reusable applications and interfaces?
- Mobile application architectures
Week 3 (1/24 & 1/26)
- CardinalKit architecture and application programming interface (API)
- Use your first CardinalKit modules
- Combining modules to build software components
- How do you develop your own modules?
- How do you create a well-architected mobile application with CardinalKit?
Week 4 (1/31 & 2/2)
- Accessing health record data
- Guest lecture - Ricky Bloomfield, Apple Health
- Learn how to architect a cloud backend for your digital health app
Week 5 (2/7 & 2/9)
- Accessing health and fitness data
- Health management, symptom tracking, and more.
- Guest workshop - Pariece McKinney & Gavi Rawson, Apple
- Share your plan and software architecture for your application
- Demonstrate the API and design of your first modules
Week 6 (2/14 - 2/16)
- In-class feedback and development.
- Tools for analyzing healthcare data.
- Guest lecture - Ryan Brush, Google
Week 7 (2/21 & 02/23)
- Guest Lecture - Marius Mainz / Daphne Petrich, HelloBetter
- In-class development and feedback.
- The United States Core Data for Interoperability (USCDI) and the Argonaut project.
- If time permits: in-class development and feedback.
Week 8 (2/28 & 3/2)
- Learn how the HL7 FHIR data standard works for interoperability
- Guest lecture - Vivian Neilley, Google Cloud
- Guest lecture - Garrick Olson & Lei Wang, Stanford Research IT
- Guest lecture - Ron Li, Stanford Hospital SEAL Team
Week 9 (3/7 & 3/9)
- Trust in Machine Learning for Healthcare - Carlos Guestrin, Stanford University, Computer Science Department
- In-class feedback and development.
- Guest lecture - Steve Hershman, Biofourmis
- In-class feedback and development.
Week 10 (3/14 & 3/16)
- In-class feedback and development.
- Teams will share their final apps
Student Info
Skills Survey
Please fill out the the skills survey when you sign up for the class, so we are aware of your current skill level at the beginning of the course.
Responsibilities
AttendanceNo more than one unexcused absence is permitted. A second absence may be permitted with written justification by the student and fulfillment of a make-up assignment. Two or more unexcused absences will lead to a No CR grade.
Amount of Work Expected
Teams can expect up to 7-9 hours of work each week beyond classroom hours. A majority of the effort is spent researching needs, meeting need coaches and mentors, interviewing stakeholders, and coding solutions as a team.
Deliverables
Teams will complete a midterm and final presentation, discussing progress and functionality. The final deliverable will be a research-ready iOS mobile application hosted on GitHub. Code and IP created is governed by Stanford's standard IP procedures, and will be used to inform University research.
Trainings and Certifications
Health research requires adherence to strict security and privacy protocols, and Stanford requires researchers working in this area to obtain specific certifications. Specifically, a Stanford HIPAA certificate is required, and will be obtained through starsexpress.stanford.edu. This certification will enable students to work on other IRB-approved research, and is a requirement for most research within the School of Medicine. This training is offered through the STARS system (use the link above).
Coding and Requirements
Teams will be coding throughout the quarter. We require that you have taken CS106A/B or equivalent. We highly recommend some experience with mobile platforms (CS193P/A, CS47), OR web platforms (CS142). Since we are working primarily on iOS, you will need access to a macOS computer running Xcode. The Macs in residential clusters and at Lathrop come with Xcode pre-installed. You can checkout Macbook Airs from the Tech Desk for up to 4 weeks through this link. See list of computers on-campus with Xcode here or here.
Topics Covered
Open Source + the iPhone as a platform for digital healthLearn about what makes the iPhone a powerful platform to develop clinical apps. Overview of how some open source frameworks leverage iOS to accelerate medical research.
Using GitHub (https://github.com/CS342)
Become familiar with using GitHub to manage software development projects.
Using CardinalKit (https://cardinalkit.stanford.edu)
Developed with the goal of providing a suite of tools to build and accelerate your digital health experience from the ground up — from an app itself to storing collected data in the cloud.
Using ResearchKit (https://www.apple.com/researchkit/)
Developed with the goal of streamlining the process of gathering robust and meaningful data. Used to document patient-reported outcomes (surveys) and track a person’s reaction time, balance, gait, mobility, among many other things.
Using CareKit (https://www.apple.com/carekit/)
Developed with the goal of helping track the development of symptoms and adherence to medication over a period of time. Often used to help patients manage a particular health condition.
Using HealthKit (https://developer.apple.com/healthkit/)
We will use HealthKit to read health and activity data from the iPhone. This enables providers to track a patient’s step count, heart rate, flights climbed, calories burned, and other similar metrics over time. HealthKit also allows us to download electronic health records (EHR) to your iPhone using the Fast Healthcare Interoperability Resources (FHIR) specification.
Using iPhone Sensors (https://developer.apple.com/documentation/coremotion)
We will use CoreMotion to track accelerometer, gyroscope, and pedometer data in real-time.
Backend for Data Visualization and Analysis on Google Cloud
Introduction to internet-enabled dynamic apps. We will build a web dashboard hosted on Google Cloud to conceptualize how data can travel from a patient to a provider's hands in real-time. We will explore how data can be analyzed using BigQuery.
Launching your first health app
We will discuss data privacy and HIPAA compliance in-depth. We will discuss IRB approval for those interested in clinical-trial applications.
Grading
3 Units – CR/NC or Letter GradeStudents may elect to attend the course for CR/NC or for a Letter Grade. In addition to attending classes, completing readings, and participating actively in the classroom environment, students are required to complete a class project exploring a need in digital mobile health. Projects are chosen in advance from a list of need areas, sourced by faculty.
Grading will be based on:
- 10% Attendance
- 10% Individual and team class participation; project teams must share presentation responsibilities.
- 30% Assignments
- 20% Midterm Presentation. All team members must be present.
- 30% Final Presentation and Deliverables (10-15 slides). All team members must be present. The slide deck (powerpoint or pdf) must be submitted in advance and no later than 2:00 PM on March 16th. Code must be submitted on the last day of class via GitHub. (The HIPAA training certificate is also part of the final presentation grade!)
Readings
Students with Documented Disabilities: Students who may need academic accommodation based on the impact of a disability must initiate the request with the Office of Accessible Education (OAE). Professional staff will evaluate the request with required documentation, recommend reasonable accommodations, and prepare an Accommodation Letter for faculty dated in the current quarter in which the request is made. Students should contact the OAE as soon as possible since timely notice is needed to coordinate accommodations. The OAE is located at 563 Salvatierra Walk (phone: 723-1066, URL: http://studentaffairs.stanford.edu/oae).