Design Day 2024 Roundup

On May 1, The Whiting School of Engineering held its annual Design Day to showcase the innovative design projects conceived, designed, and built by students during the academic year. The Department of Biomedical Engineering was represented by over 50 teams of undergraduate and graduate students from six design courses, including the undergraduate Design Team program, CBID master’s program, Precision Care Medicine, and Biomedical Data Design.

To see all of our student projects, visit the BME Design Day website. Below is a selection of student innovations:

OTSync: The future of personalized OTC hearing care

Team Members: Bryan Sabogal, Yin Xin (Bobby) Ni, Elena Porras, Yagmur Ozturk, Brian Zhou, Nadia Momtaz, Daniel Yao, and Amos Li

In 2022, the Food and Drug Administration developed regulations for hearing aids to be sold over the counter. While the regulations were intended to increase accessibility to hearing health care, it’s still difficult for the average consumer to choose the right hearing aids on their own. Sponsored by ARTIS Ventures and mentored by Nicholas Reed, a Johns Hopkins audiologist who helped develop the FDA regulations, the OTSync Design Team created an app that matches patients with the best hearing aid for their needs. As part of the ARTIS Ventures sponsorship, two OTSync team members completed a summer internship with Reed to conduct background research before starting the design process the following fall.

“We continue to be impressed by the entire OTSync team. Their efforts over the past 18 months to identify a gap in clinical care and innovate a novel solution is a testament to their entrepreneurial spirit and dedication to improving the lives of patients around the world experiencing hearing loss. We could not have asked for better partners for the project and give special kudos to the team’s faculty mentors lead by Professors Elizabeth Logsdon, Michelle Zwernemann, and Nicholas Reed in guiding the students every step of the way. We are excited to continue working with the team as they progress towards a launch,” said Henry Klingenstein, partner and Omair Khan, associate at ARTIS Ventures.

Pinnacle: Your Prostate Biopsy, Optimized

Team Members: Dylan Zhu, Devashree Gupta, Benjamin Wen, Grace Noh, Kedar Krishnan, Wendy Yang, Kenzi Griffith, and Mahmoud Radwan

Sponsored by medical technology company Mendaera, Team Pinnacle focused on the optimization of prostate biopsies; specifically, allowing clinicians to take multiple biopsy samples without having to insert and remove their biopsy needle from the patient’s body multiple times. Their solution aims to reduce patient discomfort and make the task simpler for clinicians.

“Working with Mendaera on this project has been very valuable to [our] overall learning of the design process and how it relates to industry practices. Our mentors from Mendaera were very open with discussing the techniques they employ in their actual practice, giving valuable advice on everything from how we might differently structure our interview questions to whether a design requirement we had written was poorly formulated,” said Dylan Zhu, third-year biomedical engineering undergraduate student and design team leader.

Sensovate: An Early-Stage User Experience Testing System for Oral Healthcare Devices

Team Members: Amy Zhang, Alan Mao, Neeti Prasad, Mili Ramani, Eileen Stiles, Justin Rosman, Megh Tank, and Avery Ye

The Sensovate Design Team, sponsored by Philips Oral Healthcare, is developing a way to measure how people use and experience the company’s oral health care device prototypes — before the company tests the devices with actual users. The team members were mentored by testing engineers in Research & Development at Philips, allowing them to align their design with user needs early in the product development process. The team used modeling to map the physical movements and forces produced by the device to predict user comfort levels, considering variations in user sensitivity and perceptions to identify optimal ranges.

During spring break, a portion of this team traveled to Philips’ Seattle headquarters to observe manufacturing and present their project findings for user feedback.

Precision Medicine aims to make healthcare more efficient, affordable, and accessible by using data and analytical insights to improve clinicians’ decision-making processes. Precision Medicine BME teams used this model to solve critical patient problems, including:

Monitoring and Prediction of Cardiac Arrest in Pediatric ICU Patients with Machine Learning

Team Members: April Yujie Yan, Sukrit Treewaree, Jiahui Yao, Jiwoo Noh, Sheel Tanna, and Tamara Orduna

About 40% of pediatric cardiac arrests in the United States every year occur in pediatric intensive care units (PICU), making PICUs a leading location for these life-threatening events. To enable early detection of children at risk, a team led by BME graduate student April Yujie Yan is using machine learning techniques to predict in-hospital cardiac arrests (IHCA) up to five hours in advance.

Their project demonstrates promising results in machine learning-based prediction of pediatric IHCA by leveraging real-time monitored patient data (e.g., ECG, PPG, physiological time series) and static data (e.g., demographics). The developed machine learning models achieve and demonstrate actionable early warning of impending IHCA in pediatric patients by using patient data routinely collected in the PICU, without further invasive tests.

Pre-Surgical Risk Stratification using Deep Learning on 12-lead ECGs for Non-Cardiac Populations

Team Members: Carl Harris, Anway Pimpalkar, Ataes Aggarwal, Patrick Yang, and Xiaojian Chen

Faculty Mentors: Robert D. Stevens, Casey Overby Taylor, and Joseph L. Greenstein

Recent studies indicate that about 3% of non-cardiac surgeries lead to major adverse cardiovascular and cerebrovascular events (MACCE), translating to roughly 1.5 million in the United States over a decade, with nonfatal heart attacks and strokes being the most common. Though electrocardiography (ECG) is commonly used in clinical settings, its effectiveness at predicting the risk of MACCEE during surgery is unclear.

Using pre-operative 12-lead diagnostic ECG data from 28,661 adult patients in the MIMIC-IV dataset who underwent major non-cardiac surgery, a student team led by BME graduate student Carl Harris trained a series of convolutional neural network (CNN)-based models to predict major post-surgical adverse outcomes including myocardial infarction (MI), stroke, and death. Their project aims to help clinicians more accurately assess a patient’s risk of MACCE before non-cardiac surgery so that alternative treatment options can be found for patients identified as high-risk.