Artificial Intelligence and health: A giant leap

Artificial Intelligence and health: A giant leap

Dr Watson is a new addition to the north American medical profession. But unlike other graduates who have undergone years of training, Dr Watson did things differently. Because Dr Watson is a computer. Watson, IBM’s supercomputer, is a cognitive computing system. That means instead of being programmed to complete a function, Watson learns from experience – data – in order to improve its decision-making processes.

Watson for Oncology will be “employed” to crunch large volumes of data, for assessing cancer tumours and recommending the appropriate drug. With human doctors, this process would take several weeks for each patient. With Watson, it’ll take “a few minutes”.

That’s not all. With every diagnosis made, Watson becomes faster and more “intelligent”. As genetic sequencing becomes more widespread, it will be possible to identify which genome is causing the cancer. And Watson, using its artificial intelligence, will then be able to identify the most suitable solution. Which could mean an end to treatments such as chemotherapy or radiation. By the end of 2015, 14 hospitals across the US and Canada will be using Watson.

New levels of insight

Imagine if your doctor could diagnose backed up by knowledge gained from 14 million patient visits and 3,700 healthcare providers. That’s the reality for doctors using Modernizing Medicine, a live database of medical information which can be “mined” to help doctors locate the right treatments for unknown ailments.

Listed #70 on the Forbes list of America’s most promising companies, the company has raised $50 million to develop telemedicine – where patients can share photos and have “remote exams with their physicians”.

Finding the right medication is one thing, but how to make sure the patient completes the treatment? Even if the patient is motivated, they may forget to take the right tablet at the right time. AiCure is an AI solution that checks if patients have ingested their medicine, gives regular reminders, and sends data securely back to a central database.

Using a webcam, the patient holds up medication and ingests “live”. AiCure registers what’s happening on the screen and sends updates and alerts to healthcare coordinators.

24-hour care

According to the UN, the world’s ageing population is unprecedented, pervasive, enduring and has “profound implications”. Which poses the question: What happens if a country’s population is getting older, and there aren’t enough young people around to look after them? A robotics and automation expert may have an answer.

Antonio Espingardeiro has developed an AI robot which can monitor, provide basic care, and offer companionship to patients. “We have to find more resources and have to get new ways of delivering those resources and delivering the quality of care,” he explains.

A new frontier

For Google DeepMind, a British Artificial Intelligence company bought by Google for £400 million in 2004, its mission is simple: “Solve Intelligence”. Co-founder and CEO Demis Hassabis sees data as the key to making breakthroughs in climate, economics, and disease. “They’re just tremendously complicated interacting systems. It’s just hard for humans to analyse all that data and make sense of it. And we might have to confront the possibility that there’s a limit to what human experts might understand. AI-assisted science will help the discovery process.”

How? DeepMind will learn from experience, and store it as data. And then start making new connections, with constantly evolving algorithms. Just like how the human brain develops by growing new neural networks.

But DeepMind will do this on a much bigger scale. And that’s where things start getting really “deep”. Because it means AI could start solving age-old problems, using methods humans have never been able to conceive of. When this will happen? Over to Hassabis:

“We’re decades away from anything that’s nearing human-level general intelligence. But we’re going to have systems doing useful things in five or ten years. We’re on the first rung of the ladder. As to how many rungs there are — there could be ten or 20 more breakthroughs before we’ve solved what intelligence is.”