Imagine if humans could train themselves to give more precise diagnoses. Over time, medical misdiagnosis would abate, save countless lives, and not to mention save hospitals, patients, and their families millions of dollars every year. Pharmaceutical industries, too, would be able to keep their stock orders in far more accurate conditions, keeping their reserves continuously maintained for predictive demand.
Now consider that in the distant future, your surgery could be performed by an enhanced robot. In your post-op appointment you talked, in detail, to a machine about your recovery.
There are many complex interconnections between artificial intelligence (AI), machine learning, and the way they help humans do their jobs more efficiently. We’re adapting to the idea that we can use machine checkouts for our own groceries at the supermarket, but machines commandeering the very state of our health? Many wouldn’t be too keen.
Despite the inherent reluctance to trust robots with human welfare, major advances underway in technology mean that the very framework of our healthcare system is becoming interlaced with machine learning-enhanced support. It’s already being implemented right in the middle of emergency room cases – the most nail-biting kinds of hospital encounters – where the waiting times seem long and an endless number of frantic questions need answering.
In this post, I’ll explore three major AI and machine learning-driven innovations that are underway in the healthcare space.
1. Individualized, personal genetics
The Human Genome project started out as an ambitious, costly endeavour. Its overall project cost eventually topped $3.8 billion dollars and resulted in “diddly squat” according to science writer David Dobbs.
The advent of machine learning algorithms—specifically, deep learning—has given rise to several companies who seek to leverage AI to understand genomic data. Deep Genomics, founded by Brendan Frey, seeks to find meaning in massive amounts of genomic data.
Kevin Loria of Business Insider describes what Deep Genomics seeks to do:
“Their learning software is developing the ability to try and predict the effects of a particular mutation based on its analyses of hundreds of thousands of examples of other mutations — even if there’s not already a record of what those mutations do. They’re trying to build not just a Rosetta Stone that explains a language, but a way to predict how a tiny change in the letter will create something new.”
Deep Genomics theorizes that if their machine learning algorithms can predict the effect of certain mutations, medical professionals will be able to perform pathologies and understand diseases in a completely new way.
2. Fast and efficient drug discovery
Developing a new drug is an incredibly expensive, labour intensive, and seemingly endless process that can drag on for an average of 12-14 years at a cost of around $2.6 billion.
According to a joint study conducted by Carnegie Mellon University and the Albert Ludwig University of Freiburg in Germany, the average cost of drug discovery could fall by up to 70% with the use of automation and machine learning.
University of Toronto researchers have recently developed new machine learning-driven algorithms that can significantly speed up the drug discovery process by using 3D protein structures.
Utkarshan Sinha explains how the algorithm works:
“In order to be effective, drugs must bind to specific proteins in a cell in the right orientation. They do this by changing the conformation of a protein, which results in a change in that protein’s function. Knowing the 3D structure of a protein can significantly enhance understanding of how they work in the body and consequently, aid in the development of drugs targeting the potential harmful effects of these proteins at unprecedented speeds and efficacies.”
The algorithm is able to cut down the time it takes to identify complex protein structures from weeks to several days, even when utilizing the processing speed of a basic desktop computer.
3. Better diagnoses
Machine learning-driven genetic diagnoses solutions are overturning the 2000-year old paradigm of symptomatic detection—where diagnoses is predicated on symptoms—and is moving towards a more preventative, proactivity-based paradigm. Sadly, 90,000 people in the United Kingdom alone are “unlikely to live longer than six months” due to the inability of doctors to identify the disease before it’s too late.
Instead of waiting for a patient to exhibit severe symptoms, companies like Freenome have developed tools to leverage genome data to aid in the proactive prevention of diseases. Freenome’s Adaptive Genomics Engine is driven by a philosophy that emphasizes early detection—it “detect[s] disease signatures in your blood…by using your freenome…the dynamic collection of genetic material floating in your blood that is constantly changing over time and provides a genomic thermometer of who you are as you grow, live and age.”
Innovative solutions such as these will be key in saving the lives of millions of people and unburdening healthcare systems across the globe. By detecting diseases early and facilitating preventive care, survival probability rates can be raised dramatically.