IBM was betting big when it invested billions on its cognitive learning platform, Watson, that could think, reason and learn. It invested $26 billion dollars in big data and analytics and now spends close to one third of its R&D budget in developing this technology. In 2013, Watson could analyze 600,000 pieces of medical evidence, 2 million pages of text (from 42 medical journals and clinical trials) and 1.5 million patient records before suggesting oncology related treatment, all in a matter of seconds. With so much power, Watson should be dominating the healthcare analytics market, but instead it's losing its lead to competitors that are lean, nimble and more focused. What is cognitive computing, and why has it suddenly captured the interest of government, financial institutions, investors, healthcare providers, and technology companies?
Cognitive computing is the simulation of human thought processes in a computer. Cognitive computers use machine learning algorithms to continually acquire knowledge from disparate data sources and then present the information as actionable material. Cognitive computing is not an entirely new science. It was first discussed in 1950 when Alan Turing published a paper 'Computing Machinery and Intelligence', in which he proposed an experiment, the Turing Test , which tests a machine's ability to exhibit intelligent human behavior. However, the science gained momentum only recently with technological developments in data mining, pattern recognition and natural language processing.
A cognitive computer sifts through structured and unstructured data and tries to find the hidden body of knowledge/patterns in the data. The information captured by cognitive computers is subsequently verified or discarded by humans. This process happens iteratively, and with each iteration, the cognitive computer becomes better at identifying patterns. Cognitive computing is immensely valuable to fields such as medicine in which there are no black-and-white answers, and the best answer is often based on evolving and ambiguous evidence that is colored by individual experiences or intuition.
Startups in Cognitive Computing
Although cognitive computing is closely associated with Watson, many other companies and organizations are developing products and services that are as good, if not better than Watson. IBM has acquired some of these rivals but it looks like it will have to compete with the others for a share of this market.
Flatiron Health, a New York based startup of oncology-focused analytics software has challenged Watson in its home turf. It recently raised Series-C funding of $175 million from Roche Holding and Google Ventures (now GV) to further develop its OncologyCloud platform. Like Watson, Flatiron Health gathers and analyzes oncology related data from all possible sources - doctor's notes, medical journals, publications, clinical research, laboratory reports, hospital EMRs - and makes the information available as actionable material. Flatiron Health is currently focused on cancer but plans to expand its data analytics platform to other diseases as well.
Digital Reasoning's Synthesys platform is similar to Watson. It uses Natural Language Processing (NLP) and Machine Learning techniques to understand digital communication by analyzing entities and relationships in context, within vast amounts of data; revealing what is most critical to its clients. This extracted knowledge is accessible through a robust Application Programming Interface (API). Synthesys uses cognitive computing technology that gathers, reads and understands structured and unstructured data and analyzes it as a whole. Knowledge gaps are filled, problems are detected sooner, and unexpected insights are brought to light.
Narrative Science's Quill is a perfect example of how cognitive intelligence can simplify data embedded in charts, tables and databases into meaningful information. Powered by artificial intelligence, its Quill platform can analyze data from disparate sources, understand what is important, and then automatically generate perfectly written narratives to convey meaning from the data. Quill is being used by leading organizations to convert mountains of data into actionable insight that supports strategic business goals and objectives. In 2014, the National Health Service (NHS) in the UK chose the Quill platform to transform its vast and comprehensive data, stored as numbers in spreadsheets and tables, into narrative descriptions that anyone can read, understand, and use to make decisions.
Future of Cognitive Computing
Cognitive computing has a very bright future in the connected world driven by Bigdata, IoT and Cloud Computing. IBM has extended Watson's reach through acquisitions and by developing new learning capabilities at its own R&D labs. Watson can now analyze medical images, treat cancer, offer recommendations for gene-based treatments, screen skin diseases, understand information in the EMR, and assist doctors in managing patients with chronic diseases. Watson's success has inspired other companies to develop similar products using open source tools. Startups like Lumiata and Enlitic have developed small and powerful analytic solutions that assist healthcare providers in diagnosis and prediction of disease conditions.
The healthcare industry is poised to derive significant benefits from investment in cognitive computing and related technologies. Leading technology companies like Intel and Qualcomm have designed solutions and platforms to capture health parameters in realtime and integrate it with decision support systems. The data captured through these products and platforms will be a gold mine of information. Cognitive computers will be able to analyze the information captured through wearables and intelligent point-of-care devices along with the information available through traditional data sources, such as EMR and doctor's notes. This will help in uncovering new biomarkers and lead to significant improvements in the prediction and management of disease conditions.
Cognitive computing has gained a foothold in the developed world. Developing countries like India can quickly take advantage of the technological developments in this science by deploying standardized EMR solutions in all private and government medical institutions of the country. This will greatly assist in developing tools to analyze the medical history of patients and reduce the number of medical errors / preventable adverse effects (PAE), estimated at 5.2 million medical injuries a year. A startup that partners with a leading hospital chain and a pathological laboratory will get the first mover advantage in commercializing this technology in India.
To connect with me, send me a Linkedin invite or email me at firstname.lastname@example.org. To learn more about this topic and to check other publications written by me, please visit my blog