Applied AI in Healthcare

"I don't think any physician today should be practicing without artificial intelligence assisting in their practice. It's just impossible (otherwise) to pick up on patterns, to pick up on trends, to really monitor care." - Bernard Tyson, CEO of Kaiser Permanente

Healthcare – One of the most promising sectors for AI

Not one body is the same. Or so is our perception of health. Health is a personal matter and we want doctors and nurses to treat us as unique individuals. But more and more of the human body can be described using data.

Healthcare is increasingly becomeing a discipline of collecting the right data, understanding that data, and transforming it into clinical actions.

“Machine learning and AI are already demonstrating its potential to drive efficiencies and improve the quality of care and will continue to be significant in the next 20 years."

- Charles Koontz, President and CEO of GE Healthcare

Healthcare is one of the areas with the most data about each of us. Just think about it; we can measure the blood pressure and glucose levels, we can measure the pulse and monitor our wellness data on smartwatches, X-Ray machines and MR scanners produce medical images, and EKG diagrams hold data about our heart and circulation. Even the smallest building blocks, our DNA, is mapped and discussed.

In Denmark, our healthcare data dates many years back and spans a multitude of clinical disciplines. All the disciplines are connected digitally through healthcare platforms using the unique patient id, the Danish CPR-number.

AI in healthcare represents a collection of multiple technologies enabling machines to sense, comprehend, act and learn, so they can perform administrative and clinical healthcare functions. With a large amount of data available, a key conditions for developing applied AI, numerous healthcare solutions are being developed across a large range of healthcare disciplines. And not just developed - many are already in use.

Looking at the AI investments in “Healthcare” since January 2014, more than $2 billion has been invested just in startup companies developing AI based solutions for Healthcare. Healthcare is the top industry segment for AI investments. The number of investments goes up and the application areas for the technology is spreading all over the healthcare discipline. From drug discovery, to diagnostics to mental health and clinical trials.Healthcare graph.png

The graph only shows money flowing into startup or early stage companies within AI in Healthcare. So to fully understand the ambition behind AI in Healthcare, the internal projects and spending in companies like IBM, Google and Apple, should be added. And they are investing big time.

One example is Google, and their Google Medical Brain unit. Google Medical Brain is part of Google’s AI unit, Google Brain, and the Medical Brain unit is solely focused on doing AI research and early product development within Healthcare. Alphabet (the holding company the owns Google) also owns the company Verily, so just within Google numerous projects and hundreds of people work with AI for Healthcare.  Healthcare AI is already an impressive tech industry.

Judging by the numbers, even more money will be used on developing solutions that will change Healthcare dramatically over the next five to 20 years.

What's hot in artificial intelligence in healthcare?

Based on the number of new companies entering the space, the most popular use case of artificial intelligence in healthcare is imaging and diagnostics.

The promise of AI in diagnostics lies in a computer's ability to detect diseases like cancer at an early stage. Machine learning algorithms can compare a medical image with those of millions of other patients picking up on nuances that the human eye might miss.

The second most popular investment category is insights and risk analytics. This includes companies crunching structured and unstructured clinical data to surface information and predict risk.

Closely related to this application is AI in clinical trials where start-ups raise funds to match patients with ongoing trials based on their medical history.

Having looked at the AI space in healthcare more closely during the Applied AI project a multitude of solution areas have surfaced. The most interesting areas identified during our research are:

Clinical insights & risk analytics — Companies focused on mining and analyzing health data from disparate sources using machine learning and natural language processing. Interesting companies include Roam, Flatiron, PrognosHealth Fidelity and  AiCure.
Genetic.png
Genetic research  Companies using artificial intelligence for personalized healthcare, genome sequencing, and CRISR research. Interesting companies include LabGenius and Desktop Genetics.
Ripcage.png
Medical imaging & diagnostics — Companies using deep learning and computer vision to identify anomalies in medical images and scans. AI promises early identification of hard-to-detect illnesses by comparing thousands of images at a speed and scale impossible for humans. Interesting companies include Enlitic, Lunit, Zebra Medical Vision and Google Research
psychological.png
Mental health — Here the focus is on early detection of mental illnesses, as well as consumer-focused products for lifestyle management and cognitive training. Interesting companies include Ginger.ioCognoa and Koko.
pills.png
Clinical trials & drug discovery — Companies that leverage AI to find the most effective combinations of drug compounds reducing the time and capital required in drug discovery. It also includes companies focused on clinical trial management. Interesting companies include Mendel.aiNotable Labs, Recursion Pharma, Deep 6 AI, IBM Watsontwoxar and Atomwise.
VA.png
Virtual assistants — Voice and text-based AI assistants that interact with patients and analyze their responses. Interesting companies include Babylon, Ada, Your.md and Push Doctor.
hosp.png
Emergency room & hospital management — Companies developing solutions for monitoring emergency rooms in real-time and reducing hospital readmission rates, among other applications. Interesting companies include Gauss Surgical, Qventus, Jvion and Medy Match.
heart.png
Lifestyle management & monitoring — Consumer-focused health monitoring and wellness applications, including tracking medication adherence and personalized diabetes management recommendations. Interesting companies include Wellframe, Ovia HealthDreaMed DiabetesFlo and SkinVision

Companies transforming healthcare using AI powered solutions

Through the Applied AI Project, we have selected three sample companies working with AI solutions for healthcare.  Will  they be successful in the long run is hard to say, but they illustrate AI solutions available today that support such different scenarios as diagnostics, matching patients to clinical trials and a smart assistant to the general practitioner.

The Capital Region working with applying AI at Danish hospitals

RegionH20logo.pngThe Capital Region (Region Hovedstaden) is one of five regions responsible for providing public healthcare services in Denmark. This responsibility covers the development and daily operation of 11 hospitals, psychiatric institutions and other healthcare practices in the Region.

The Capital Region covers the geographic area around Copenhagen, employs 38.000 people, has its administrative center in Hillerød and is the largest region in terms of citizens served (1,8 million as of January 2017). Other key figures from 2016-17 include: 

Sources:  [1] Region H - Budget 2018-2021, [2] Faktafolder om Region Hovedstaden

The Capital Region has a political vision of being internationally recognized as a green, innovative metropol that ensure their citizens and patients a high life quality and delivers a consistent top-quality medical treatment. The ambition is to become one of the five most preferred places in the world for the development of health and welfare solutions.

Together with IBM, the Capital Region has carried out a seven month AI pilot project during 2016-17 looking at specific use cases for AI. Other themes, including legal and ethical aspects of using AI in healthcare, was also explored in that period.

Overall, the results of the project have been positive. The project group expects that artificial intelligence will have a major impact on the future of providing health care services in relation to prevention, diagnosis, treatment and planning of healthcare.

The Capital Region has decided to proceed with AI in larger scale following the pilot project. The ambition is to develop strategic partnerships and solutions, benefitting the Danish healthcare system and creating growth in Greater Copenhagen.

Where does The Capital Region see the use of AI?

The Capital Region has identified a wide range of areas with great potential. These include oncology, diabetes treatment, and mammography screening. The initial findings reveal that further development and local adaptation is required before the technology can be fully implemented in the clinics.

Below are three scenarios where Region Hovedstaden imagines practical use of AI technology is possible. None of the described cases are implemented, but they are all project candidates for the Region’s continued work with AI in 2018.