The missing piece of the puzzle

Artificial Intelligence
Cognitive computing is presenting an exciting route through which the healthcare sector might transform itself and, with increasing pressures, it looks like we might be ready to embrace it. Experts argue cognitive technologies may address future challenges, starting from their impact on the global data explosion crisis, compounded by workforce shortages. But with complex systems and challenges, how do we bring cognitive out of research and into the real-world. Where does one start?
Experts have for some time advocated the application of cognitive computing in the healthcare sector, but many wonder how it can be used when today’s data is predominantly
unstructured, systems don’t talk to each other, and clinicians have an incomplete “picture” of the patient, coupled with the fact that they are spending the majority of their time on administrative tasks. IBM believes clinicians and radiologists are the key to solving the puzzle; cognition may be the missing piece.
At this year’s UK e-Health week, a panel of clinical experts weighed in on the exciting possibilities of cognitive computing for everything from making smart imaging request forms to “batch” sorting exams so radiologists can focus their attention on the abnormals; from comparing and tracking measurements across time and across large numbers of prior comparison exams to seamlessly “import and report” capabilities for streamlined reporting. “What would happen if a system gave clinicians the opportunity to confidently say that a given set of scans is normal, allowing radiologists to concentrate on assessing the ones that are abnormal?” That is the question Mike Fisher, CCIO at Royal Liverpool and Broadgreen University Hospitals NHS Trust, asked IT leaders and suppliers in the audience at the conference. Fisher went on to say, “It’s an information overload problem, if we’re going to address the data deluge we need innovative technology like cognitive.”

What does AI really mean?

Chairing the discussion, Mark O’Herlihy, IBM Watson Health Imaging Managing Director for International, explained IBM’s tendency to define AI as ‘augmented intelligence’ instead of ‘artificial intelligence.’ There is a lot of confusion amongst people regarding the terms AI, cognitive, machine learning and deep learning. “We believe the first step is understanding that our approach is about the clinician working in concert with the expanded ability of a computer to gather different types of information from multiple sources, organise that information and present it in the most useful way. This allows the clinician to work more reliably and accurately for the benefit of patients.”
During the panel, Dr Shankar Sridharan, CCIO at Great Ormond Street Hospital for Children NHS Foundation Trust, seemed to agree with the use of the term ‘augmented intelligence,’ explaining that cognitive computing, for example, should complement the work of clinicians, helping “make medicine smarter, make doctors smarter”. He
told the audience in London “I am the superhero and cognition is the cape.”

Clinician-informed innovation

IBM says clinicians are directly involved in the development of the cognitive software they will use to administer care. Watson Health Imaging aims to provide a comprehensive
patient imaging record that enables multi-specialty collaboration so clinicians can have a more complete view of the patient. 
The approach IBM takes aims to remove some of the main barriers and improve productivity, focusing on interoperability, adding another dimension to the “digital imaging maturity” approach. Cognitive is the piece that potentially enables providers to evolve beyond simply storing, sharing and viewing images from a single department (a traditional PACS system), to connecting a myriad data (images, text, records), from multiple departments and care systems, adding advanced cognitive analytics and reporting to
ultimately work more in a more streamlined, efficient way.
Jonathan Benham, Consultant Radiologist at Sandwell and West Birmingham Hospitals NHS Trust, explained the new type of reporting that Merge, an IBM company, is developing can improve processes for radiologists:“I have been interested in structured reporting for a long time. There hasn’t been an easy way of doing it. We have moved from using tape to digital dictation and then typing by a transcriptionist, and then moved onto voice recognition. Now, we need to move towards something that would enable use of a template on a screen which can be accessed interactively.”
Standardisation of reports is expected to allow radiologists to spend more time viewing and interpreting medical images, reducing the need to edit and correct transcription errors.“Real-world cognitive solutions like some of the works-in-progress from IBM Watson Health Imaging show great promise and it’s exciting to be a part of that discussion and development,” Benham added.

Technology striving to intelligently address challenges

Technological advancements must carefully address legacy and autonomy issues to handle the growth in medical data. Nevertheless, people are increasingly becoming more inclined to learn about augmented intelligence and open themselves to change, but will that be enough to solve the puzzle?
[Published ahead of press]

IBM Watson Health

Merge, an IBM company together with Watson Health Imaging are providing enterprise clinical imaging and cognitive computing solutions to advance healthcare globally. Visit