Pro-AI doctor on why tech is needed in medicine

AI makes healthcare more economical, says Dr David Shusterman, who supports the use of the tech in day-to-day medicine.

Healthcare is being transformed by artificial intelligence (AI), whether we like it or not, bringing with it ways to automate tasks like never before.

Here at SiliconRepublic.com, we often cover advancements in the medtech sector, from new products to speedy vaccine development to AI-powered 3D medical-imaging technology among so much more.

And just as AI makes its way into the latest cutting-edge supports for healthcare, the revolutionary tech is also seeping into the day-to-day lives of healthcare professionals as they help patients.

For example, the BBC recently looked into AI usage in Dublin’s Mater hospital – one of the first in the country to use the tech in its radiology department. It reported that Mater uses AI to analyse head scans for bleeds, chest scans for blood clots and x-rays for fractures. And a new trial AI model at the Mater has been trained to create “synthetic” MRIs from CT scans.

The hospital finds that the tech is particularly helpful in assisting younger doctors when they don’t have support from experienced consultants.

To understand how AI is transforming everyday healthcare, SiliconRepublic.com spoke to Dr David Shusterman, a New York-based board certified urologist.

Several use cases

Dr Shusterman has always been pro-modernisation, he says, referring to himself as a “high-tech urologist”. He believes that the healthcare sector is always apprehensive of newer technology, it “always takes a lot longer” he says.

“I mean, healthcare is going to be revolutionised by it, whether doctors like it or not, because it becomes more economical to run an office.”

His urology practice uses AI in a number of areas, from helping with documentation to treatment analysis.

According to various surveys, documentation and administrative work costs healthcare professionals in UK’s NHS around 13 hours a week and some specialists in the US as much as 19 hours weekly. However, with AI, manual documenting time can be cut down significantly.

Dr Shusterman uses an AI service which makes transcription easy while producing meeting notes with his patients. While on the more complicated side, he uses AI to “continuously” monitor patient outcomes with treatment, in a way that would be complicated to do without.

His centre inputs treatment outcomes into a database, which is then analysed to see which outcomes are better. While this is possible without the use of AI, he explains that the tech allows for continuous analysis, which would be difficult to do manually.

“With AI, you could actually kind of look at it as it’s happening,” he says.

From the stats his patients had before to what they have after a treatment, the AI model can analyse the best possible treatment depending on the aggregated outcomes.

Image: David Shusterman

Shusterman’s medical specialty requires extended work with ultrasounds, and he says that new AI-infused ultrasound tech is on the way.

“I know that ultrasound companies right now are working on self-reading ultrasounds, meaning the ultrasound machine will give a full read and full chart note of what we find.

“Just basically helping the stenographer identify things that AI thinks is important.

“I think ultrasound is going to be one of the things that gets updated for AI purposes in the near future and the same could be said for CAT scanners and other machines,” he adds, especially as AI computing power continues to go down in proportion to its output.

Although, it is important to note that the International Energy Agency predicts a dramatic four-fold increase in energy demand arising from AI-led data centres.

Moreover, Shusterman acknowledges that “liability issues” are the reason advanced tech such as this is not on the market yet, although, with that addressed, “it really shouldn’t be that hard” to release the tech “very soon”, he hopes.

Innovation means change

“I’ve been trying to kind of get out there and preach modernisation of practices,” says Shusterman. He is a believer in fast adoption of cutting-edge technology and doesn’t think there’s much long-term harm using AI in his field.

“Everyone, especially in medicine – we’re always nervous about new technology and how it can have bad consequences when we ignore all the great consequences that come with [it].”

According to him, 95pc of healthcare is “very basic”, which means, “it’s not difficult to understand”. He says that a lot of day-to-day healthcare can be relegated for the most part to AI.

In fact, he expects that AI can learn and make suggestions that could be “better than what [healthcare professionals] do traditionally”.

While they would still need to analyse all the cases that come their way, doctors and AI operators would only need to use their “own intelligence in about 5pc of cases”, he explains.

“Analysing something that’s a finished product is much easier than coming up with a finished product.”

However, studies demonstrate that patients want doctors who establish an “empathic relationship”, while a bioethics expert from University of California, Berkeley says that AI cannot replace human empathy in the healthcare setting.

Moreover, AI in healthcare also means a loss of jobs that have been commonplace in clinics and hospitals for decades.

“Stenography is one of these industries that really with the advent of AI will be kind of left behind,” Shusterman acknowledges. “We had scribes that are in the office that take notes and instead of having a scribe – we actually replace them with AI technology.”

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.