AI in healthcare: how artificial intelligence can help us fight pandemics



The potential for AI in healthcare is enormous yet largely untapped. The novel coronavirus, however, is forcing a rapid (and widespread) surge in medical AI adoption.

As nations around the globe struggle to contain and combat COVID-19, the demand for AI-powered healthcare solutions is soaring. And AI, it seems, is rising to the challenge.

Here, we explore the role of AI in healthcare, and how, amidst the COVID-19 pandemic, AI stands to help in the fight.


AI in healthcare

Artificial intelligence encompasses a vast range of technology tools and functionality. So, which AI functions are the most prevalent in healthcare?

•       Machine learning (ML)

Machine learning is one of the most common forms of AI. It’s also one of the most useful functions for AI in healthcare, allowing machines to take in, analyse and learn from masses of data.

In healthcare, machine learning (including deep learning), can help healthcare professionals spot patterns. From this, it allows computers to predict the effects of different symptoms or treatments. This means that ML can predict whether someone is likely to develop a given symptom or if a treatment is likely to work.

•       Natural language processing (NLP)

Another helpful function of AI in healthcare is natural language processing. NLP focuses on teaching machines to understand natural language. That is, language as we use it in daily life.

In healthcare, natural language processing relates to speech recognition, text analysis, and translation. Use cases include unlocking unstructured data in databases and documents, clinical trial matching, chatbot assistance, decision support, and more.

•       Deepfakes

Deepfakes are convincing content created by artificial intelligence. For the most part, the technology has cultivated a bad rap. But it can prove beneficial — and acts as another form of useful AI in healthcare.

By using deepfake technology, hospitals can create deepfake patients and true to life data that can then be used in testing. (For instance, when testing different treatments and cures.) All without putting real patients or their data at risk.

So, what exactly does AI in healthcare do, and how does it support our fight against pandemics?


Diagnosis

One use for AI in healthcare is in diagnosis. For example, AI can allow computers to recognise tiny details in medical imagery that would otherwise be missed. These can aid with diagnosis — making outcomes more accurate.

Computers can process masses of data, and far faster than a human can. This means that artificial intelligence can review and store every symptom, case study and journal made available to it. From there, it can suggest the most likely diagnoses for a healthcare professional to investigate.

Also, thanks to pattern recognition, AI can predict the likelihood of someone developing a condition or symptom. This aids with early detection and gives healthcare workers more time to help patients make a recovery.


Diagnosis in a pandemic

In the pandemic, AI can help healthcare professionals diagnose patients with the novel coronavirus.

•       Speed up the development of testing kits

South Korea has successfully ‘flattened the curve’ of new infections. They’ve conducted over 300,000 coronavirus tests in the process. To do this, they used artificial intelligence to help them develop a test for the virus in a matter of weeks, rather than months.

The computer was able to assist in the development of the test without having a sample of the virus. It used the genetic details about COVID-19 that had been released.

•       Predicting severe cases

With deep learning, artificial intelligence is also able to identify COVID-19 patients that are most at risk. That is, it can predict with greater accuracy which patients will progress to develop ARDS. (Acute Respiratory Disease Syndrome.) This is a severe complication of the novel coronavirus that requires intensive care.

This use of AI has helped doctors learn more about how the novel coronavirus interacts with the body. It also highlights the patients who are most at risk, which can help hospitals manage overwhelming patient numbers.


Research

AI in healthcare also extends to medical research. It can collect and organise research materials from a wide variety of sources. And it can translate foreign studies for greater accessibility.

Moreover, AI can help researchers make sense of the vast amount of material that’s out there. It can even notice patterns that might have otherwise been missed. In turn, this can spark more directed research into conditions, diseases and cures.

Beyond that, AI is an extremely versatile tool for healthcare research. It can help recreate genome sequences, and provide extra information to facilitate research.


Research in the pandemic

With regards to the pandemic, AI in healthcare is proving a helpful tool for researchers.

•       Genome sequencing

Already, scientists have managed to recreate the genome sequence of COVID-19 in under a month, with the help of AI. And whole-genome sequencing can provide a host of insight.

Knowing the genome of a virus can help researchers determine all sorts of useful things. For instance, susceptibility to diseases in individual patients. It can help us understand more about the virus, and track outbreaks. Researchers could even glean insight that helps with the creation of a vaccine.

•       Research mining and accessibility

Fighting COVID-19 is a global effort — and so a host of research is already pouring in. More, that is, that any one person can hope to keep up with. AI can (and is set to) help with this.

Artificial intelligence algorithms will take on the task of mining the mass of research — with the help of NLP. It can cross-reference papers, and search for patterns. So, any answers hidden in the sea of research will come to light.

Artificial intelligence can also help keep research accessible. Many countries are working tirelessly to fight the pandemic. As such, language barriers might prove problematic. But, with machine translation, AI helps to keep insights accessible.


A bright future for AI in healthcare

AI-based healthcare systems are being developed and deployed at an impressive rate. Indeed, by 2025, AI is expected to be implemented in 60% of all global hospitals.

And this rate only seems likely to accelerate as we turn to AI with our current COVID-19 strains.

AI is a tool — one that has plenty to offer the healthcare industry. And AI in healthcare is there to assist and support healthcare professionals in any setting. Be it a regular Tuesday, or during a pandemic.


Useful links

ELI5: what is deep learning?

Yes, positive deepfake examples exist

The EU AI guidelines explained


Download