IBM has announced its intention to purchase Truven Analytics, a healthcare data management firm, providing analytics and services to hospitals, health systems, health insurance companies, pharmaceutical companies, and government agencies. IBM has offered to pay $2.6 billion for Truven, which was sold by Reuters to private equity firm Veritas Capital in 2012 for $1.25 billion.
IBM already holds 100 million patient records from past acquisitions of Phytel and Explorys, and with Truven’s acquisition the company will add data on another 200 million patients. In IBM’s own words, it will have “one of the most complete data sets on patients and healthcare in the world” as a result of the Truven acquisition. The deal is still subject to regulatory approval in the United States, and IBM owning such a large repository of health records could be an issue for regulators, as IBM will have an unparalleled competitive advantage, but also for privacy advocates who are likely to be concerned about how the data will be used, and whether HIPAA compliance is possible for such a large data set.
IBM’s main goal here is to expand the dataset that feeds into its artificial intelligence (AI) engine, Watson. The larger the dataset, the better Watson will be at training its models to make predictions, identifying correlations in the data, and providing unique insights. IBM calls this Cognitive Computing, another word for AI or machine learning based solutions that help companies make better decisions. For example, as a part of its Watson Health offering, IBM has targeted oncology treatment at the Memorial Sloan Kettering Cancer Center, where the data from the hospital has been used to train Watson, making it a “learned colleague” to support physicians.
Watson Health is also being used by partners like Apple, CVS, and Under Armour in the wellness area to introduce new offerings, and by hospitals to improve productivity. IBM has more than 30 partners already using its healthcare platform, and this number is growing at a considerable pace. With Truven, Watson Health’s capabilities grow by orders of magnitude, making it one of the most powerful AI-based health platforms in the market. IBM estimates the healthcare and life sciences market to be worth $200 billion, with over $2 trillion in waste within healthcare spending globally, which represents 1 in every 4 dollars spent.
Alphabet’s AI arm, DeepMind, has also recently launched a health platform called DeepMind Health. The company is initially helping the UK’s National Health Service (NHS) by empowering healthcare professionals with important data like identifying at-risk patients using data analytics, using a smartphone app called Streams. It also has another project where it helps clinicians and nurses to organize care better using an app called Hark.
In both cases, DeepMind is not known to be using any AI tools like deep learning with neural networks, which is at the heart of its value proposition. Instead, DeepMind has been busy learning to beat humans at games like Go, advancing the state of the art in terms of how machines can outperform humans at basic tasks like pattern recognition. On the other hand, Alphabet’s life sciences unit, Verily, is specifically working on healthcare wearables like a contact lens that tracks glucose or a health-tracking wristband. Its unclear how Verily might work in partnership with DeepMind, but it’s not hard to see synergies between Verily’s specialization on medical-grade hardware and software and the data analytics that can be performed on the back end with DeepMind.
Both IBM and Alphabet are at the cutting edge of artificial intelligence, with IBM pushing the limits of Watson but focusing on immediate commercial gains, and Alphabet having a much longer game plan with DeepMind, where the commercial proposition is quite unclear for now. Healthcare is an area where AI can be extremely useful, for starters improving diagnosis and treatment of chronic diseases, and helping the healthcare systems become more cost-efficient.
Tractica’s upcoming report on healthcare wearables covers some of these trends related to the digitization of healthcare, and the importance of AI-based data analytics. Wearables that target a certain chronic condition like diabetes, cancer, or heart disease are under development, and in the next few years will achieve regulatory compliance, as well as insurance compliance. Healthcare wearables will help automate data collection from patients, enabling remote care and improved population health management, which will ultimately power these AI based healthcare platforms, helping them make better decisions.