January 3, 2017 | Written by: Jen Clark
Categorized: Health and medicine
Share this post:
The healthcare industry is facing a problem. There is too much data to keep up with, and most of it – as much as 80% – is going unused. Information from patient records, research, clinical trials and medical journals is largely unstructured, and there are reams and reams of it – far too much to sift through manually.
Unfortunately, this means that much of the value of this information is currently getting lost, whereas if extracted, it could have useful or even life-saving implications for patients: from helping them to manage long-term medical conditions at home, to creating targeted treatment plans for those who don’t respond positively to mainstream drugs.
Now, however, it may be possible to extract useful insights from many disparate sources thanks to the cognitive power of IBM’s super computer: Watson.
The data explosion
Here’s where the numbers get scary. Recently, the volume of healthcare data reached 150 exabytes. Just to put it into context, one exabyte is equivalent to one quintillion bytes – enough to fill a stack of DVDs that would reach from Earth to Mars. That’s a crazy amount of information, and the manpower required to read and understand it all would be extraordinary.
This is where Watson can help. The Watson Health Cloud brings together vast amounts of medical data into a centralized, cloud-based hub. It can read 200 million pages of text in 3 seconds, and bring advanced analytics to help turn the raw data into useful information that can help patients.
Watson in action: fighting disease
I like to think of Watson as a super-powered personal assistant: someone who can review huge quantities of data, conduct a literature review and summarise it into plain English, ready for doctors to consider when treating a patient. As we’ll discover, some of Watson’s insights are already being put to good use in treating cancer and diabetes.
Targeted treatments for cancer patients
Watson is already making a valuable contribution to Oncology, suggesting tailored treatment plans for individual patients based on their unique medical records, an understanding of the latest advances in care, and a knowledge of what clinical trials might be available to them. The need for tailored treatment is palpable, since not everyone with cancer responds to the same treatment, and on average only 25% of patients benefit from the first drug they are given.
So how can you tell which treatments might benefit which patients, without having to go through the process of trying them out? The answer is in the data. By using information from cancer genome sequencing, a patient’s notes and the body of research available, Watson is able to build up a picture of each patient’s medical history, and predict which treatment methods are most likely to see a positive response.
Cancer genome sequencing
IBM Watson has teamed up with Quest Diagnostics to work on cancer genome sequencing. After Quest Diagnostics has sequenced a patient’s tumor, Watson analyses the genetic alterations found, and identifies mutations that are potentially treatable. By comparing DNA from tumor cells to DNA from healthy tissue, Watson can spot patterns and identify targeted therapies for each individual patient. It can also point out clinical trials that may benefit the patient, and predict which options to avoid – for instance, whether chemotherapy might be especially toxic for a particular individual.
Case study: IBM and New York Genome Center partnershipIBM is already putting this ability to use in a project with the New York Genome Center, creating a national cancer tumor repository as part of the White House Precision Medicine Initiative. The repository will help pinpoint which genetic characteristics impact how a patient may react to treatment. The initial project included 200 patients from hospitals around the New York area, many of whom had failed to respond to regular treatment options. The team hope to scale up the project so that it can benefit over 100,000 patients a year, making the repository more widely available.
Tackling diabetes: a case study
Big data analysis can be applied to the study and management of diseases besides cancer. IBM is collaborating with the American Diabetes Association to produce cognitive apps that help with prevention, identification and management of diabetes. Earlier this year, they unveiled a Medtronic & Watson app designed to help patients manage their own condition on a day-to-day basis. Based on retrospective analysis of a patient’s insulin levels, continuous glucose monitors and nutritional data, the app can help people understand how their behavior affects glucose fluctuations in real time. Patients will be able to spot links between lifestyle choices, and hopefully have the information they need to avert a crisis.
These are just some of the ways that Watson and the Internet of Things is having an impact on healthcare. IBM is undertaking many other projects to help improve patients’ quality of life, including a project with Thomas Jefferson University Hospital to enable patients to take control over their hospital stay: operating lights and window blinds through voice activation, and asking questions about the availability of hospital facilities.
Take a look at our website to find out more, and explore how Watson is ushering in the era of cognitive health.