Rising to the Challenge
From the fabrication of personal protective equipment (PPE), to machine learning models that assist with patient triage, to mobile tools for monitoring symptoms of disease, Allen School researchers are rising to the challenge and developing innovative, real-world solutions to aid health care providers, patients, and public health officials in responding to COVID-19.
Read on to discover how our faculty, students, and staff are applying their expertise to support communities here in Washington and around the world in their fight against the pandemic.
Medical Equipment Fabrication
Protecting front-line health care workers and patients
Protecting front-line health care workers and patients
Shortages of personal protective equipment (PPE) such as masks, face shields, and other items for medical personnel are well-documented in multiple communities responding to the current pandemic. Without this critical equipment, individuals who are caring for COVID-19 patients risk contracting the disease themselves and/or passing it on to others. UW Medicine recently launched an effort to harness the university’s 3D printing and other fabrication expertise to address the shortage of PPE. As part of a campus-wide initiative, Allen School faculty and staff led by professors Jennifer Mankoff and Adriana Schulz are working alongside UW Medicine clinicians and other partners to rapidly design, fabricate, and test safe and effective PPE that can be locally produced and put in the hands (and on the faces) of our front-line health care workers.
Allen School Maker in Residence Colleen Carroll, who also serves as associate director of the Center for Digital Fabrication (DFab) at UW, has focused on streamlining the production process across various fabrication labs, material sources, and newly created assembly stations in order to scale up the initial effort into a truly campus-wide operation. The team commenced production of its first project, a face shield designed in collaboration with Tim Prestero of Redmond-based Design that Matters, in late March after UW Medicine obtained design approval from the National Institutes of Health (NIH). The face shield is one of, if not the only, DIY designs recommended by the NIH so far for use in clinical settings. After making more than 1,000 face shields and bridging the gap in supply for UW Medicine until its vendor could resume filling orders, the team is planning to supply 300 shields adapted for field conditions to first responders at King County Medic One. The group is also exploring the design and fabrication of additional equipment, such as gowns, reusable respirator masks, and aerosol boxes, to aid in the pandemic response. Learn more in this UW News story.
Preventing the spread of disease
A shortage of ventilators is a pressing concern for facilities serving patients with COVID-19 and other respiratory diseases who require assistance to breathe. Where patients have access to ventilators, another looming shortage — that of ventilator filters — could also endanger the people around them. Without adequate filtration, ventilators can discharge humid air full of contagions into places like nursing homes that are full of already vulnerable people, increasing the risk of additional infections. Also, many people in nursing homes, which have been hit hard in the pandemic, use ventilator-like equipment such as CPAP/BiPAP machines. To cut off this avenue of potential disease spread, a team of researchers in the Allen School’s Programming Languages & Software Engineering group (PLSE) and the Graphics & Imaging Laboratory (GRAIL) led by professors Zachary Tatlock and Adriana Schulz is working with a team from University of California, San Diego that includes professor Stefan Savage (Ph.D., '02) of the Department of Computer Science and Engineering and emergency room physician Dr. Christian Dameff, M.D., on a project to enable 3D-printing of replacement ventilator filters. The team is assisting front-line health care providers to rapidly prototype and manufacture a 3D-printed enclosure that can be easily assembled to serve as an emergency filter when regular filter supplies are depleted. While initially focusing on ventilators, the team plans to make their design customizable for use on a range of devices used to assist respiration.
Supporting rapid and accurate screening at the point of care
Rapid diagnostic tests (RDTs) are used to quickly screen for infectious diseases without the need for expensive laboratory equipment, making them ideal for point-of-care health monitoring. While RDTs are easy to use, people can sometimes misinterpret the test results due to human errors. Researchers in the UbiComp Lab led by Allen School and Department of Electrical & Computer Engineering professor Shwetak Patel developed a smartphone-based system called RDTScan that uses computer vision techniques to assist the user in capturing a high-quality image of their completed test, combined with an algorithm that automatically analyzes the result. This improves the accuracy and consistency of test results while also enabling researchers and public health officials to gather data about community-level disease prevalence. After initially focusing on infectious diseases such as malaria and influenza, the team is currently building out a new open-source RDTScan library that will enable developers to extend these capabilities to new COVID-19 RDTs as they come onto the market. The development team can be reached at RDTScan(at)cs.washington.edu.
Improving patient triage and health system capacity with machine learning
Health care providers use clinical risk scores generated by machine learning models to triage patients and maximize clinical resources. These risk scores are based on an analysis of multiple factors; in the absence of sufficient COVID-19 data, providers rely on data for pneumonia patients to serve as a proxy. A team led by Allen School professor Su-In Lee is applying transfer learning to improve the generalizability of these scores to patients with COVID-19 and building sophisticated, nonlinear machine learning models that will more accurately classify patient risk compared to simpler, linear models. The researchers are incorporating feature selection and optimization methods to achieve accurate predictions with data that can be gathered more easily than laboratory results and x-ray studies. The project has the potential to save lives by reducing incidents of misdiagnosis, avoiding unnecessary hospitalizations, and improving the health care system’s ability to cope with the COVID-19 pandemic.
Disease Monitoring & Contact Tracing
Monitoring cough to aid patient recovery and new case detection
Members of the UbiComp Lab led by Allen School and Department of Electrical & Computer Engineering professor Shwetak Patel are developing a smartphone-based app for monitoring patient coughs — a frequent symptom of COVID-19 — to aid patient recovery and to identify potential new cases of the disease. The project, which is a collaboration with clinicians at UW Medicine and Seattle Children's Hospital and funded by the Bill & Melinda Gates Foundation and National Institutes of Health (NIH), builds on the lab’s previous work on CoughSense and will employ a combination of the phone’s microphone and an algorithm capable of distinguishing coughs from other sounds to measure their frequency. The team is in discussions with the U.S. Centers for Disease Control about the project, which aims to alleviate capacity pressures on the health care system by enabling providers to remotely monitor the condition of patients recovering at home. The app could also be used to monitor the condition of persons under investigation (PUI) who have had contact with confirmed cases but have not yet begun displaying symptoms. The researchers are currently inviting people to participate in an online study collecting human coughs and other vocalizations to refine the algorithm that powers the app. For more information, see coverage by UW News, Slate, PBS NewsHour, and KING5 News.
Contactless respiratory monitoring to gauge the onset of respiratory infection
Changes in a person’s breathing rate can be a useful indicator of the onset of viral respiratory infections such as COVID-19. These changes can occur at any time, including while a person is asleep. A team of researchers led by Allen School professor Shyam Gollakota and Dr. Jacob Sunshine of UW Medicine is developing a method for passive sensing of a person’s breathing overnight with the aid of a smartphone or smart speaker placed near their bed. Their approach relies on sonar — inaudible acoustic signals such as those a bat or dolphin emits to navigate and avoid objects in its path. The system would emit high-frequency tones above the range of human hearing, which would be reflected back to the device’s built-in microphone. Those reflections will change based on the motion of a person’s chest while breathing, enabling the system to extract information about their breathing rate and detect changes that could indicate the onset of respiratory infection. The team is currently validating their approach in collaboration with the Seattle Flu Study. The researchers wrote about the project in a recent New York Times op-ed.
Privacy-preserving PACT for mobile contact tracing
A team of UW and Microsoft researchers produced a white paper in which they propose a framework for the use of technology to halt the spread of COVID-19 while protecting civil liberties. PACT, which stands for “Privacy Sensitive Protocols and Mechanisms for Mobile Contact Tracing,” would augment conventional contact tracing methods to support public health objectives while preserving the privacy of people affected by a disease outbreak. The paper offers a comprehensive set of principles to guide the development of mobile tools for contact tracing and population-level disease tracking. These principles rest on a “third party free” approach that incorporates strict user privacy and anonymity standards underpinned by a decentralized approach to data storage and collection to mitigate security and privacy concerns. To illustrate, the team outlines the features of a potential smartphone app for conducting “privacy-sensitive” mobile contact tracing that relies on Bluetooth-based proximity detection to identify instances of co-location (two phones in proximity) via pseudonyms to determine which individuals may be at risk. PACT can also be applied to enable mobile-assisted contact tracing interviews and the sending of narrowcast messages by public health officials — all while affording robust privacy protections to the user.
The team was led by Allen School professors Stefano Tessaro, Sham Kakade (who holds a joint appointment in the UW Department of Statistics), Shyam Gollakota, and Tadayoshi Kohno, along with John Langford, Eric Horvitz, and Jonathan Larson at Microsoft. Read more in this Allen School blog post and the team’s white paper on arXiv.org.
CovidSafe contact tracing app
The team behind the PACT framework subsequently teamed up with Dr. Jacob Sunshine of UW Medicine and volunteers at Microsoft to produce a new mobile contact tracing app, CovidSafe, that combines the robust privacy-preserving protocols of PACT with tools for tracking disease symptoms and transmitting public health messages to users. The researchers are working on versions of the app, currently in the demo stage, for the Android and iOS mobile operating systems. They also publicly released the code to enable public health agencies and other organizations to customize CovidSafe’s features for their own use. Learn more in the UW News release and a related GeekWire article.
Crowdsourcing a potential cure
Fold-it, an online protein-folding puzzle game that is a collaboration between the Allen School’s Center for Game Science and the UW Institute for Protein Design, is engaging citizen scientists in some friendly competition to fight COVID-19. The scientists behind Fold-it challenged players to design a protein that would block the ability of the novel coronavirus SARS-CoV-2 that causes COVID-19 to infect human cells and replicate. The goal is to design a protein that could bind with the virus spike protein, thus interfering with its ability to bind with human cell receptors. After three rounds of competition, scientists at the institute selected 99 of the most promising submissions to test in the lab as part of the search for an effective antiviral treatment. Read coverage in GeekWire.
Improving risk modeling with data from wearable devices
Allen School researchers led by professor Tim Althoff are working with the Bill & Melinda Gates Foundation on a project to develop new spatio-temporal modeling techniques for predicting flu and COVID-19 infection risk. The team will use signals collected by wearable devices, such as heart rate, sleep and physical activity, along with GPS information to improve researchers’ ability to gain actionable insights from behavioral health studies — including studies that collect data on a smaller scale.
Combating the spread of misinformation online
During a pandemic, social media can be a useful tool for rapidly disseminating information in the public interest. It also tends to be a hotbed of conspiracy theories, unproven medical advice, and fake news. A team led by professors Franziska Roesner in the Allen School’s Security and Privacy Research Laboratory and Jevin West of the iSchool and the Center for an Informed Public are exploring how people react to misinformation they encounter online related to the COVID-19 outbreak. The team is also examining different social media platforms’ approaches to mitigating the spread of such misinformation among their users. The researchers recently posted preliminary findings based on a survey conducted in late March to understand people’s experiences with COVID-19 related misinformation on social media. See coverage in Adweek here (registration required).
Understanding the moral dimensions of the COVID-19 outbreak
A team of researchers led by Allen School professors Katharina Reinecke and Rajesh Rao, in collaboration with professor Andrew Meltzoff at the UW Institute for Learning and Brain Sciences, are embarking on a project to explore the moral dimensions of the COVID-19 outbreak. The group will study the impact of the coronavirus pandemic on people’s moral values across countries, cultures and age groups, using the LabInTheWild platform to gather data on a global scale. The researchers are particularly interested in exploring what people identify as the most socially accepted “solution” to a number of COVID-related moral dilemmas. Individuals may participate in the online study here.
Analyzing the behavioral impact of the COVID-19 response
Allen School professor Tim Althoff and his team are collaborating with researchers at Microsoft Research to better understand how COVID-19 has impacted population behavior. Among the impacts they plan to investigate are the widespread shift to working from home and the economic impacts associated with communities’ response to the pandemic.
If you would like to financially support the research efforts described on this page, you may make a tax-deductible gift online to the Allen School Annual Fund and specify to which project(s) you would like your gift directed. The "Your Information" link will send you to a form that includes a "Comments" box where you can indicate the efforts you wish to support. Simply write “COVID” if you have no preference; these gifts will be directed to the area of greatest need. Make a gift.
The Allen School is part of a larger University of Washington community that has pulled together to respond to the pandemic. If you are interested in supporting immediate needs related to virus testing, medical research and patient care, you may do so through the UW Medicine Emergency Response Fund. To provide support to students who have been impacted by COVID-19, you may direct a gift to the University’s COVID-19 Emergency Student Fund.
From pursuing leading-edge research and cross-campus collaborations, to providing life-saving medical care to patients, to applying creativity and innovation to help our students achieve their educational goals, together we will overcome the challenges of COVID-19. Learn more about UW’s response.