Yales’ PHATE Data Analysis Algorithm Predicts Mortality Rates Among COVID-19 Patients

Yale researchers have announced their PHATE algorithm for data analysis can be used to detect COVID-19 among patients and indicate risks of mortality in infected patients.

The COVID-19 pandemic has inflicted permanent damage to the human race, killing millions and impairing the immunity of billions of humans. Kids as young as six months old are getting tested positive for COVID-19. In such desperate times where healthcare authorities are focusing on minimizing the rate of infection and bringing down the number of deaths, Artificial intelligence and deep learning tools could come to their rescue. In a recent update on the use of data analysis tools for healthcare and pharmacology, Yale researchers announced their recent invention.

Top Healthcare Technology Update: Health Hero Closes Oversubscribed Seed Round with Participation from Major VCs Animoca Brands and Blockchain Founders Fund

Yale revealed the scope of its multiscale PHATE algorithm last month in the journal of Nature Biotechnology. This Machine Learning tool for data analysis allows researchers to scan through millions of granular level data of millions of cells and zero onto a single infected cell within minutes. Developed in the lab of Dr. Smita Krishnaswamy, who is an Associate Professor of Genetics and Computer Science, PHATE machine learning algorithm for data analysis snips through the conventional challenges faced in data visualization of Big Data projects. PHATE’s machine learning algorithm is so precise that it can “zoom in and focus on specific subsets of data to perform detailed analysis,” said Manik Kuchroo, a doctoral candidate at Yale School of Medicine.

During the development of this ML-based data analysis tool for COVID-19, researchers at Yale found TH17 helper T-cell influenced the outcome of the infection. By measuring these TH17 cells clustered with the immune system cell IL-17 and IFNG, Yale researchers could find out if the infected person survived or died with an 83% precision. PHATE plot of immune cells clusters provides predictive modeling and analysis of COVID-19 mortality rates. Yale researchers would continue to fine-tune the assessment from the PHATE Machine Learning tool for other infections diseases as well.

Recommended: Conversational AI SaaS Marketing By 2025, Chatbot Messaging App Accesses to Reach 9.5 Billion Globally

[To share your insights with us, please write to sghosh@martechseries.com]