Hong Kong-based AI developer creates COVID-19 survival calculator

The online tool provides a COVID-19 patient's risk score, expected time to death and survival probability curve based on 15 health variables.
By Adam Ang
02:04 am
Share

Credit: Deep Longevity

AI systems developer Deep Longevity, a unit of Hong Kong-listed Endurance RP Limited, has launched an online tool that can potentially estimate the survival of a hospitalised COVID-19 patients.

WHAT IT DOES

The web-based COVID Risk Calculator provides a patient's COVID-19 risk score, expected time to death (TTD) and survival probability curve. It requires a physician to input 15 health variables, such as symptoms and comorbidities.

The tool is based on research published last week in the peer-reviewed journal Life. The study collected the blood test results of more than 5,000 COVID-19 positive patients admitted to 11 public hospitals in New York.

It then analysed those test reports through a deep learning neural network called BloodAge to "quantify the intensity of the ageing process" and check a person's biological age, which can be higher or lower than their chronological age.

BloodAge predictions can have "more impact on a patient's survival than chronological age," based on two survival models: Cox proportional hazards and logistic regression. It was said that each extra BloodAge year is equivalent to a one-day reduction in TTD.

WHY IT MATTERS

In a press statement, Deep Longevity said that risk stratification or assigning risks to admitted patients remains an "essential, albeit grim, necessity" as hospitals around the globe continue to be overwhelmed with new cases.

It noted that age is the main risk factor affecting a COVID-19 patient's survival. To provide an "accurate" survival analysis, it developed the COVID Risk Calculator to quantify a patient's ageing rate.

Deep Longevity stressed that the calculator has limitations and urged potential users to read its research paper.

THE LARGER TREND

Early in the year, consumer genomics company 23andMe released a similar tool called COVID-19 Severity Calculator which informs users how certain non-genetic factors may impact the risk for hospitalisation due to SARS-CoV-2, the virus causing COVID-19. However, it does not predict an individual person’s risk for severe symptoms.

Last year, there were risk assessment tools launched that can tell the probability of a person testing positive for COVID-19. One such tool was developed by Etihad Airways with Vienna-based Medicus AI. Another tool by US-based Cleveland Clinic uses an algorithm to provide a risk score based on a user's list of symptoms, details of potential exposure to the disease and demographic data from hospitals. ​​​​

ON THE RECORD

"Age was recognised as the main risk factor affecting patients' survival at the very onset of the pandemic. The elderly have been reported to have the highest mortality rate, as well as suffer from more complications in numerous studies. In the meantime, most such studies ignore that there is no universal pace of ageing. Some people age faster than others. This notion is obvious to medical professionals, who have gained the ability to tell over-agers and under-agers apart throughout the years of practice. However, the official records lack any information on the true, biological age of COVID-19 patients," Endurance Longevity CEO Jamie Gibson said.

"The research project by Deep Longevity in collaboration with Lincoln Medical Center highlights the importance of quantifying ageing rate for accurate survival analysis," he added.

Share