AI, ML and Big Data in Healthcare18th February 2019
Artificial intelligence (AI) isn’t a new idea, but it’s only in recent years that our technology has caught up to the point at which it has practical uses. We’re a long way away from true artificial intelligence like the human-like robots and computers that we see in science fiction movies, but we are at least in a place where AI can outperform human beings at certain tasks.
And AI could be particularly powerful in the health care industry. One piece of research from Accenture found that key clinical health AI applications can potentially create $150 billion in annual savings for the US health care economy by 2026. Another report from Tractica found that the AI health care market will be worth $34 billion by 2025.
AI is great at performing repetitive tasks, and there’s no shortage of them in the health care industry. In the United States, physicians spend more time filling out electronic health records (EHRs) than they do interacting with patients. If artificial intelligence could take on the bulk of that record keeping, it would free up their time to spend it with patients. In fact, it would have the same impact as hiring hundreds of thousands of new doctors.
AI has multiple impacts across the entire healthcare industry, but they can typically be categorized as aiding with one or more of the following.
- Keeping Well: New innovations like the smart belt, which warns people when they overeat, are helping to usher in a new era of preventative healthcare. The goal is to keep people well and to stop people from having to seek treatment in the first place.
- Early Detection: When problems do occur, AI can help to spot them earlier. For example, Microsoft is developing computers that work on a molecular level to fight cancer cells as soon as they’re spotted. AI is also being used to analyze online search engine behavior to spot mental health issues.
- Diagnosis: AI can help physicians to diagnose patients more quickly and arrive at insights based upon the 80% of health data that’s invisible to current systems because it’s unstructured.
- Decision-making: Clinical decision support systems and other AI-based tools can help doctors and patients to prioritize tasks. Examples include Quest Diagnostics’ Quanum and VitreosHealth.
- Treatment: AI-based tools are already being used across the board. Google DeepMind is reducing the time it takes to plan radiotherapy treatment while IBM’s Watson is making treatment recommendations based on patients’ medical records around the world, including in China, Thailand and India.
- End of Life Care: We’re living longer than ever, and our aging society is requiring more and more care at the end of their lives. AI-powered virtual assistants and even robots are being touted as the future, and indeed robots are already being used to care for the elderly in Japan.
- Research: AI can help to uncover new drugs and treatments, but it can also be used to research the diseases themselves, potentially allowing us to inoculate against them or to eliminate them. For example, Canadian start-up Meta uses AI to quickly analyze scientific papers and to provide easy insights.
- Training: AI-powered simulations can help surgeons and other healthcare professionals to hone their craft without putting real patients at risk. AI models are generally more realistic and reliable, and one of the advantages of using AI for training is that it can tailor the training to each different individual.
Will doctors become obsolete?
The short answer is no, or not all. Historically, technology has created jobs instead of destroying them, and this holds true all the way back to the industrial revolution. According to Gartner, AI will have eliminated 1.8 million jobs by 2020. At the same time, it will create 2.3 million new jobs, leading to an overall increase of 500,000. Other predictions are similarly optimistic.
Part of this is because of the way that AI and physicians would likely interact. AI-powered clinical decision support tools could provide physicians with suggestions based on hard data, but it would be down to physicians and their patients to take this data and to decide together on the best way to proceed.
But on top of that, there’s just something about health care that calls for a human touch. Just imagine that you’ve just been diagnosed with cancer. Would you prefer to be told the news by an emotionless robot that’s nothing more than an algorithm, or would you prefer to be told by a friendly family doctor?
Artificial intelligence goes hand-in-hand with machine learning, natural language processing and other technologies, all of which can be combined to process the huge amounts of big data that we create on a daily basis. In the health care industry, being able to process this data and to draw new conclusions isn’t just a matter of making money — it’s a matter of life and death.
It won’t be long until artificial intelligence is being used as standard practice throughout the health care industry, and that’s good news for all of us. After all, we’ll all become patients at some point in our lives, and AI has the potential to usher in a new era of health care in which we’re all treated with personalized health care plans based on data and not just the results of clinical trials.
And the good news is that we won’t even lose our doctors. AI won’t replace them — it’ll just help to make them more efficient. It’s a true case of man and machine working better together than either could in isolation, and it spells a bright future for all of us.
AI, ML and Big Data in Healthcare was originally published in Hacker Noon on Medium, where people are continuing the conversation by highlighting and responding to this story.