AI-Guided Prescriptions May Help Prevent Bacterial Co-Infection Deaths In COVID Patients | Weather.com – The Weather Channel

Representational image

In addition to the intrinsic severity of COVID-19, there is a hidden threat many patients face when inflicted with the virus: secondary bacterial infections. With the immune system already working overtime to liberate the body of the coronavirus, other harmful invaders such as bacteria have a much easier time entering us in our weakened states and inflicting illnesses.

It is important to note that secondary bacterial infections can be deadly, and must never be underestimated. As University of Queensland Professor Kirsty Short elaborates, most to all COVID-19 fatalities have a secondary co-infection as a culprit.

Now, a team of researchers led by Dr Short has harnessed the power of machine learning to predict the risk of secondary bacterial infections.

Ever since the discovery of Penicillin, we've made monumental strides in the field of antibiotics, and a few of them would certainly be effective against such secondary infections.

However, while treating all patients with antibiotics might seem logical, over-prescription can set a dangerous precedent for bacteria to evolve to evade our current antibiotics. Essentially, experts warn that unregulated antibiotic usage could end up starting an antibiotic-resistant superbug revolution, something we obviously hope to avoid at all costs.

One solution is to predict which patients are more likely to develop a secondary bacterial infection, and begin their treatment promptly and preemptively. Fortunately, with a little AI magic, Kirsty might have an answer.

To help those hospitalised and at risk from secondary bacterial infections, the researchers developed LASSO (least absolute shrinkage and selection operator). This technique analysed blood samples from COVID-19 patients across six countries, identifying seven genes whose expression levels could predict the risk of developing a secondary bacterial infection within 24 hours of hospital admission.

Dr Meagan Carney, a key researcher on the project, emphasises that LASSO has the potential to revolutionise antibiotic prescription, and put doctors more at ease during the process. Further, Carney explains that the LASSO technique is a relatively simple method compared to the complex AI algorithms often discussed in the media. Its straightforwardness could make it easier for scientists from other fields to build upon the procedure, and accelerate the adoption of data science in the medical industry.

We should strive towards making data science less of a black box and inspire scientists across the world to better understand how it can revolutionise the medical industry, she remarks.

The findings of this research have been published in Lancet Microbe and can be accessed here.

**

For weather, science, space, and COVID-19 updates on the go, download The Weather Channel App (on Android and iOS store). It's free!

Go here to see the original:

AI-Guided Prescriptions May Help Prevent Bacterial Co-Infection Deaths In COVID Patients | Weather.com - The Weather Channel