Coronavirus: AI algorithm to predict kidney injuries among COVID-19 patients developed

A newly-developed artificial-intelligence-based algorithm might help health professionals derive which COVID-19 patients face a higher risk of developing acute kidney injury (AKI) during dialysis. Presented virtually during the ASN Kidney Week 2020 Reimagines October 19-October 25, the preliminary reports of the research indicate that AKI is relatively common among the patients suffering from the novel coronavirus infection. 

As per reports, the experts from Icahn School of Medicine at Mount Sinai used the data from at least 3,000 hospitalised patients of COVID-19 for the study. The researchers then trained a model based on machine learning, a kind of artificial intelligence to predict the acute kidney injury that requires dialysis. Moreover, the data collected in the first 48 hours of the study was incorporated in a bid to make the predictions when patients were admitted to a medical facility.

Read - Trump's Claim Of Higher Covid-19 Testing Doesn't Explain Rise In Cases In The US: Study

Read - COVID-19 Escalated Into 'syndemic' For Chronic Disease Patients, Says Study

Model demonstrated high accuracy

On the basis of the research, the new AI-based model showcased high accuracy (AUC of 0.79) and these features were essential to predict the blood levels of creatine and potassium, age along with other vital signs of heart rate and oxygen saturation. The co-author of the research, Lili Chan, MD, MS, also reportedly said that the machine learning model demonstrated “good performance” and noted that such models can be “useful” during a future hike in COVID-19 cases.

Chan also informed that the experts are in the middle of deploying these AI-based models to different healthcare facilities to enhance the care of patients battling the novel coronavirus infection. 

"A machine learning model using admission features had good performance for prediction of dialysis needs. Models like this are potentially useful for resource allocation and planning during future Covid-19 surges," said co-author Lili Chan. "We are in the process of deploying this model in our healthcare systems to help clinicians better care for their patients."

Meanwhile,  a separate study on a small sample size concluded that a significant proportion of COVID-19 patients discharged from hospital experience breathlessness, fatigue, anxiety, and depression for 2-3 months after the infection. Led by the University of Oxford, the researchers observed the medium-term impact of COVID-19 on 58 patients hospitalised due to the novel coronavirus infection.

Read - Antibodies Against Coronavirus Detectable Up To Seven Months Post COVID-19 Onset, Says Study

Read - COVID-19: US Likely To Report More Than Half A Million Deaths By February, Says Study

Image: Representative/PTI