AI can help patients interpret home tests for COVID-19

New machine learning research led by George Mason University Professor Farukh Alimi and Professor Janusz Wojtociak provides a way for patients and clinicians to better predict whether symptoms are due to COVID-19, influenza or RSV.

More accurate diagnosis leads to better decisions about the course of care to heal patients and prevent the spread of disease. With fellow researchers at George Mason University and Vibrent Health, Alemi and Wojtusiak recently published a series of articles in a special issue of Journal of Quality Management in Health Care Discuss how artificial intelligence (AI) can help diagnose COVID from a combination of symptoms And home tests.

With their research, Alemi and Wojtusiak are now working on a website to offer an AI-based resource to help individuals select recommended actions as a result of their clinical profile and at-home COVID test results.

“We see AI radically improving clinical triage and test-to-treat decisions,” said Wojtociak.

Al-Alimi added, “Artificial intelligence will allow individuals to feel more confident about their decisions to stay home, seek care, or socially isolate. A lot of people test at the end of their symptoms and, surprisingly, find that they are still positive. What does one do if you disagree? Symptoms and home test results? Our AI will help these individuals understand how to proceed.”

The study in Paper 1 (shown below) found that the timing of symptoms is important in diagnosing COVID. For example, a Runny nose As an early symptom increased the odds of testing positive for COVID, and a runny nose as a symptom that occurred later decreased the odds. Likewise, fever is always a late symptom, so lack of fever early on should not be used to rule out MERS.

The results in Paper 2 found that COVID could not be diagnosed from individual symptoms; However, a combination of three or more symptoms can help with the diagnosis. Results from Paper 4 found that the accuracy of diagnosing COVID symptoms was higher when presenting with different body symptoms. For example, a combination of common neurological and respiratory symptoms was more diagnostic than any combination of symptoms individually. In addition, COVID has different presentations depending on age, disease severity, and virus mutations.

Paper 3 discusses how AI symptom screening could improve — and for vaccinated individuals, at-home antigen tests replace it. Tests at home are not always accurate and require clinical review, but these tests are performed at home where no such review is available. AI symptom checking can help make these tests more accurate. The study suggests that an AI symptom check is more accurate than a second home test.

The four papers published in the Special Supplement are:

1. Order of occurrence of COVID-19 symptoms
2. The role of symptom clusters in the triage of COVID-19 patients
3. Shared symptom screening and home tests for COVID-19
4. Guidance for the triage of COVID-19 patients with multiple systemic symptoms

A fifth paper, Modeling the likelihood of infection with COVID-19 based on examination of symptoms and prevalence of influenza and influenza-like illness, from the same group of researchers was also published in Journal of Quality Management in Health Care In April/June 2022.

Al-Alimi was Mason’s main investigator. Mason was a subcontractor for Vibrent Health, with Praduman Jain as the principal investigator on the project. (Jain is on the advisory board for the Mason School of Public Health.) Other Mason-affiliated researchers on these projects include associate professor Amira Ross, associate faculty member Ji Fang, doctoral student Elena Guralnik, and former student and assistant faculty member Wejdan Bagis. Rachel Peterson and Josh Schilling of Vibrent Health and F. Gerard Moeller of Virginia Commonwealth University were also part of the research team.

The methods used in these five papers vary. In Paper 4, the researchers performed a meta-analysis of the literature, using data from the published papers. In other papers, the researchers surveyed patients who underwent a PCR test and examined the relationship between patients’ symptoms and PCR test results. Most of the research was conducted using data collected between October 2020 and January 2021, prior to current variants such as BA.5 or BQ.1.

Related previous publications by these investigators include a study looking at how computers distinguish between COVID-19 and influenza and Analysis of university students’ symptoms and social distancing.