If you follow the latest trends and emerging technologies in healthcare, you have surely seen endless headlines about artificial intelligence (AI) in radiology. The buzz is well within reason. Artificial intelligence encompasses an extensive range of computerized technologies to solve problems in ways that imitate human thinking.
Machine learning, a subset of AI and form of clinical data science, provides an opportunity to improve the quality of care for patients and is an exceptional opportunity for radiology departments to remain on the forefront in informatics. Machine learning can be viewed as a toolbox of mysterious, but important, mathematical techniques (or algorithms) that empower computers to improve task performance.
So, does this mean machines or robots will be running radiology departments soon? No, but this form of artificial intelligence will play a significant role in the future of imaging and value-based care.
Medical imaging data has grown exponentially, with the potential to transform how we care for patients. Radiology images are intriguing, and image interpretation allures informatics professionals with expertise in machine learning. Algorithms have improved, and technology is advancing. Most of us can remember the days of film and would agree that radiology departments have quickly transformed into digital environments. AI will soon generate compelling insight into the data we acquire, and radiology departments will work alongside these powerful industry techniques.
Machine learning can provide powerful insight when used in image interpretation, clinical decision support, measuring radiology patient outcomes, data integration and knowledge management. Machine learning algorithms can provide enhanced efficiency and accuracy in interpretation of images as well as facilitate follow-up management of reported findings.
Current uses of machine learning include computer-aided detection and diagnosis (CAD), content-based image retrieval, automated classification of radiology reports, prediction of overall survival of cancer patients using “radiomics” and prediction of transition from mild cognitive impairment to Alzheimer’s disease.
It is imperative that radiologic technologists understand the vocabulary of informatics and its emerging technologies, such as machine learning. In a recent article in the Journal of the American College of Radiology, Kruskal, et al. stated that “understanding imaging informatics is as central to radiology education as radiation biology and MRI physics.” Radiologists and technologists are going to be working alongside AI as we continue to pioneer informatics applications in our field, and comprehension is key to successful implementation. The real value in machine learning lies in the potential for deeper knowledge of the data acquired and future of value-based patient care. The possibilities are endless with informatics!
Want more insights from Kori? Watch "Biomedical Informatics and Its Impact on Radiologic Technology" today and earn 1.0 CE credit from the ASRT!