Coronavirus: AI is fighting on the frontline
With our NHS making frantic preparations for a torrent of incoming Covid-19 patients, we explore how machine learning (ML) technologies can support frontline workers and act as a crucial weapon in the fight against corona.
Optimizing Coronavirus Diagnoses
Since its initial outbreak in Wuhan province, local Chinese hospitals have been deploying AI to help diagnose Covid-19 cases as early, and accurately, as possible.
For example, China-based tech companies, including Alibaba, offer computer vision technology that analyses CT scans of potential patients’ lungs and identifies visual signs of associated pneumonia symptoms. Neural networks were trained on CT scan images and accompanying data of over 5000 confirmed cases. The resulting network capably determines whether a given patient has the infection based on a simple CT scan of their chest. Alibaba claims accuracy rates of over 96% and state that the diagnosis process takes less than 20 minutes – a considerable improvement on the human-led methods that require physicians to analyse multiple images if they are to make fully-informed decisions.
Avoiding Hospital Acquired Infections
AI-powered robots may prove a valuable asset in curtailing the spread of Covid-19 within clinical settings.
As hospitals struggle to contain this exceedingly contagious virus, autonomous robots offer a safe option for disinfecting facilities. These self-driving devices navigate rooms, disinfecting surfaces by emitting concentrated ultra-violet light that eradicate bacteria, viruses and other harmful microbes. The UV light damages the microbes’ DNA, thus impeding their ability to grow and multiply. According to developers, the room is entirely cleaned in under 20 minutes. Thus, the ‘rob-cleaners’ can optimise patient flow by prepping rooms in lightning quick time, and eliminate the risk of further transmission via contaminated cleaning staff.
Helping Make the Toughest Choices
With limited medical supplies and escalating numbers of incoming patients, doctors in northern Italy are already having to make life-or-death decisions. They must assess an individual’s chances of overcoming the disease so as to prioritise resources and maximise survival rates.
Emerging Ai diagnostic tools offer a helping hand through this unenviable process. These systems analyse blood samples and predict survival probabilities on a case-by-case basis. This will inform doctor’s choices and help ensure that ICU beds are being used in the most efficient way possible.
There are, of course, countless nuanced ethical considerations here. For one, should doctors be prioritising those with the best chance of survival? Or should those with the highest risk of death get treatment first? Leaving such decisions up to AI would be a highly controversial and potentially harmful option. However, AI’s input eases the psychological strain that burdens physicians and empowers them to take the right course of action.
Developing Drugs To Tackle Corona
AI’s transformative impact on the pharma world has been increasingly evident in recent years and the current pandemic is further highlighting its potential value.
First, machine learning algorithms are combing through drugs currently on the market and identifying candidates for re-purposing. These systems analyse the make-up of drugs and assess whether they may be useful in tackling Covid-19.
Furthermore, tech giants including Facebook, Google and Alibaba are offering up their supercomputers to hasten vaccine development. By modelling solutions and clarifying the virus’ make-up, these machines are empowering researchers as they attempt to identify vaccines in record time.
Thereupon, as health services around the globe face their greatest challenge in decades, AI technologies are set to play a crucial role in minimising Covid-19’s adverse effects. Whether it is reducing hospital-acquired infections with state-of-the-art cleaning bots, or hastening vaccine development with ML algorithms, AI will prove an invaluable ally to hospital workers as they fight corona on the frontline.
https://www.techtimes.com/articles/248237/20200321/researchers-develops-an-artificial-intelligence-tool-to-help-doctors-make-choices-for-coronavirus-patients-should-it-help-make-life-or-death-decisions.htmAI, Coronavirus, Covid-19, Machine Learning, Pandemic