Health Informatics - Fall 2023

Course Instructor.

Course Description

This course provides an overview of fundamental concepts of information management including the information that healthcare organizations generate and use, the importance of data quality, and external factors that impact healthcare data management. The curriculum will also examine information systems (i.e., the arrangement and interaction of information, processes, people, and technology) as well as system standards and security issues. Health Informatics aims to improve our health and well-being by utilizing data provided by technologies such as wearables, IoT devices, and mobile phones. Students will also learn the essential steps needed to analyze sensor data focusing on health applications through assigned reading, group discussions, and hands-on projects. Since Health Informatics is an interdisciplinary field, we will cover a range of literature on human-computer interaction, wearable systems, machine learning, embedded systems, and health. Readings and in-class activities will prepare students for their final creative assignment that involves writing and presenting a research proposal for a research idea.

By the end of this course, students should be able to:

• Recognize, analyze, and pose solutions for information processing challenges as faced by health professionals managing select patients, populations, or groups.
• Identify the current state of development of the major areas within health informatics including the following: Standards and terminologies, Decision making and decision support, Clinical information systems (institution, enterprise, public health, and personal health records), Systems life cycle, with particular attention to systems selection and implementation, Knowledge management and information retrieved, Information retrieval and human-computer interaction.
• Demonstrate the conceptual ability and technical skills to develop and execute an evaluation plan involving data extraction from practice information systems and databases. Understand the process of creating and evaluating a digital health solution
• Devise and evaluate biomedical informatics solutions to the information processing challenges identified, consistent with ethical, legal, and social considerations.
• Learn about developing an embedded system for digital health applications using Arduino
• Analyze different types of data obtained from mobile health (mHealth) systems, including but not limited to the following sensors: ECG, PPG, accel/gyro, video, audio, proximity, ambient light, piezoelectric
• Write a proposal for a digital health solution

The grading breakdown is as follows:

• Class Participation (5%)
• Lead Class Discussions (15%)
• Paper Reviews (15%)
• Projects (40%)
• Research Proposal (25%)

Class Participation (5%)

Students are expected to attend all class sessions (unless students are granted prior permission to be excused). Additionally, students are expected to engage in class discussions by referencing relevant points from the readings, asking questions, and providing feedback to their peers. Students also participate by collectively annotating the reading assignments on Perusall.

Lead Class Discussions (15%)

Health Informatics is an interdisciplinary field that encompasses people from different backgrounds. Every week, a group of students will lead the class discussion by presenting a topic not covered in the class related to digital health. For example, if you are interested in diabetes management, you can choose to present the topic and engage your peers in a discussion. Another example is to utilize a skill that you have. For instance, if you are an expert in 3D printing, you can demonstrate your skills to your peers. You will get the chance to choose a slot in the first week of class. This activity can be performed in a group of 1-4 students.

Paper Reviews (15%)

The paper reviews aim to encourage students to critically read the assigned readings to help students write their research proposals. The reviews should mainly consist of the following sections: 1) paper summary, 2) things that you liked about the paper, and 3) areas that you believe need more improvement or clarification. In addition, there will be a prompt specific to the topic of the week or previously covered topics in some review assignments. Reviews are due at 9 a.m. on the assigned due date to allow the instructors to read over the reviews and incorporate them into class discussions.

Projects (40%)

We will become familiar with developing and evaluating digital health systems. Basically, we can employ embedded systems knowledge to develop mHealth systems that include microcontrollers, sensors, actuators, communication modules, learning algorithms, and firmware. As a result, we will learn about the embedded systems development process. We use an Arduino development board to develop our systems for real-time data processing of health-related signals. There will be 3-4 small projects that will help us learn about the mHealth systems development process.

Research Proposal (25%)

Students will integrate in-class discussion and reading knowledge to create a digital health research proposal. The proposal is divided into seven connected stages (S) that will prepare the students for their final submission. The stages are the following:

S1 - Focus area and idea generation and peer review of ideas (6%): There will be two separate steps for S1 including:

• Focus area: In this stage, students will identify an area in health that they are interested in and would like to improve. Students also determine what type of digital health contribution they want to make (e.g., building a system, analyzing data, understanding a population/process, etc.). This stage aims to help students understand what has already been investigated and identify gaps that can be filled. This stage’s deliverables will be a 1-2 pages single-spaced document (excluding references) that describes the chosen health focus area and the current digital health research in that area. The aim is to provide the reader with the big picture of what has been done and is unanswered. You do not need to go into detail here. Most of this information can be extracted from the abstract and the paper’s title. Google Scholar is a suitable source for searching for research articles. Once you find a paper, you can also check the citations of these papers, which is a possible way to find other related work.
• Idea generation and peer review of ideas: In this stage, you will produce three ideas related to the focused health area. The deliverables will be three short paragraphs introducing the problem, what has been done, and what you plan to do. In addition, you will be randomly assigned to review six pitches from your peers. This review aims to provide constructive feedback to help make the ideas as solid as possible. Good feedback will involve questions, identifying opportunities and risks, and suggesting modifications or extensions.

S2 - Team formation and Introduction and related work of the final proposal idea (8%): Students will be able to discuss their ideas in class, which will help them form project groups for the final proposal. The group should decide the final ideas they want to pursue and submit a 1-page proposal abstract or summary. A good proposal abstract should summarize the significance of or the need for the proposed work, the project’s main objective/contribution or hypothesis, the method used to achieve the objective or test the hypothesis, and finally, the impact of the work. In addition, students will present more details about their ideas by supporting their work’s significance and positioning it within the existing literature. This will require the team to read and summarize papers related to their research question to help demonstrate their work’s significance and novelty.

S3 - Methods and presentation of results (2%): In this stage, students will provide more details on how they can achieve their objectives. For example, groups proposing a system should describe how they will build such a system in style and explain how they can prove or show that their system archives their objective. Another example is a group that wants to interview a particular population to identify design opportunities; in this case, the group should provide details on how they would reach such communities, the questions they intend to ask in the interview, and why these questions are essential. Students also describe how they intend to present or communicate their results. The usage of graphics, tables, or charts is highly encouraged. Students are expected to get any results; however, planning how results should be analyzed to help find gaps in the experiment design or system evaluation method.

S4 - Final proposal document (3%) and presentation (6%) During finals week, students will get 10 minutes to present their proposal, followed by 3 minutes for Q&A. In the final proposal document, students must incorporate feedback from S2 and S3.

Main Textbooks:

Students do not need to purchase textbooks for this class, but we will cover the following books in this course.

1. TinyML: Machine Learning with TensorFlow Lite on Arduino and Ultra-Low-Power Microcontrollers 1st Edition by Pete Warden & Daniel Situnayake, ISBN-13: 978-1492052043.
2. Health Informatics: Practical Guide Seventh Edition by William R. Hersh & Robert E. Hoyt, ISBN-13: 978-1387642410

Sensors and components to be used in this course:

You can find the list of the required components here.

• IMU (inertial measurement unit), including accelerometer and gyroscope
• Microphone
• Optical Sensor for PPG Monitoring
• Heart Rate Monitor (ECG)
• GSR (galvanic skin response) Sensor
• RGB Camera
• Low-Resolution Thermal Camera

Projects:

• Physical activity recognition using IMU
• Keyword spotting for anxiety level monitoring using audio
• Stress detection using a GSR sensor
• Heart rate monitoring for anomaly detection using PPG and ECG signals
• Food detection using RGB camera

Class Schedule