By Nancy Koenig and Murray A. Reicher, MD FACR
Cognitive computing involves self-learning systems that use data mining, pattern recognition, and natural language processing to mimic the way the human brain works. Instead of a software engineer devising an algorithm that remains static, cognitive systems can perform evolving tasks previously reserved for humans. The resulting opportunities are incredibly exciting.
At Merge Healthcare, we have no doubt that cognitive computing will play a major role in our industry. At the same time, we are cognizant of the numerous challenges facing our national healthcare system, including an aging population, declining reimbursements, and increased reporting responsibilities. As the industry changes, it is no longer sufficient for healthcare IT vendors to merely improve workflow and efficiency. We must endeavor to improve efficacy. And the thoughtful development and integration of new cognitive technologies will be critical to
While we feel fortunate to be a part of IBM, the world leader in cognitive computing technology, we also understand that no company will monopolize the advances to come. Rather, we understand that our success, and that of our customers, depends upon choices we make together. How do we harness this new technology to bring benefits to everyone—physicians, healthcare business leaders, and most of all, patients, while avoiding the pitfalls of other technological revolutions? We believe a key to the success of cognitive computing will be the development of collaborative workstations. Instead of attempting to replace what doctors and other caregivers do best, we plan to target the tasks that people find tedious and that are error-prone. Thus, cognitive computing will not only help improve the quality of healthcare, it will improve the quality of the lives of healthcare workers.
“Cognitive computing will not only help improve the quality of healthcare, it will improve the quality of the lives of healthcare workers.”
To reach this goal, we need to understand how people work in medical imaging and how imaging specialists diagnose disease. We need to comprehend human cognitive processes and weaknesses, such as our biases and physical limitations. With more than 25 years of experience in medical imaging, Merge brings this knowledge to bear, and is committed to growing our understanding of how imaging specialists interact with images, technology, patients, and referring doctors. We are also committed to collaboration with domain experts in order to invent and validate our offerings based on real data and evidence.
What will cognitive computing bring to medical imaging and when? The journey began with IBM Research more than a decade ago and will soon begin to yield real Merge and Watson Health Imaging products. Examples include technology for aggregating and filtering comprehensive patient records and well as technology for automated analysis of cardiac ultrasounds, particularly for improved diagnosis of aortic stenosis.
As the journey unfolds, we anticipate bringing two other innovations to market soon after. One is called Marktation, a new process for marking images with annotations and creating a report at the same time. Marktation strives to use image analytics to help radiologists keep their eyes and cognitive attention on the images. It will be introduced initially for mammography and then expanded for use in all anatomical locations. The result will be faster, more accurate reporting with fewer head and eye movements. Marktation represents a perfect example of how cognitive computing can help a physician – the doctor will still find the lesion, but the computer will describe the anatomical location and add the proper description to the correct location of the clinical report.
Another product being developed will be a new lesion tracking module deeply integrated with our PACS offerings. This module will leverage analytical technologies such as image series matching, classification of images by body part, tissue segmentation, and registration to help reading physicians more easily track lesions over time and semi-automatically create tracking tables in reports.
In the future, we will begin to see further advancements in image analytics and other innovations, all aimed at helping to enhance the performance and job satisfaction of our clinical users. To see demonstrations of cognitive works in progress from Merge and Watson Health, please contact us
to schedule a meeting at RSNA or visit our innovation theaters at booth #2538.