The amount of patient information collected and electronically transmitted during patient visits has grown significantly in recent years, generating interest among the radiology community to leverage Big Data in order to provide better, faster, and more affordable services.
A recent expert panel discussion, moderated by Lawrence Tanenbaum, MD, Vice President and Chief Technology Officer, Medical Director Eastern Division, Director of CT, MRI and Advanced Imaging, RadNet Inc., explored how emerging advanced technologies are being implemented across medical imaging.
Workflow efficiency remains a key issue in many radiology practices. High patient volumes, multiple imaging modalities and techniques, supporting rapid patient throughput demands, have created an environment where every second of a radiologist’s time counts. And while data flow in radiology has become largely seamless, a lack of enterprise integration and process-related inefficiencies continue to hinder productivity.
Fortunately, smart injectors, contrast and radiation dose monitoring systems, artificial intelligence (AI), and other new technologies are emerging to help imaging providers achieve improved workflow efficiency and consistency in complex imaging studies, providing better diagnostic accuracy, and increasingly personalized patient care.
Take power injectors, for example. Power injectors in various configurations have been available for some years. But new, syringe-less injector systems are dramatically improving CT imaging workflows; some increase throughput by as many as 2.6 patients per day, on average, compared with dual syringe injectors.1
Some vendors have developed integrated software capable of monitoring and recording contrast and radiation dose. These server-based systems standardize documentation and reporting with radiology information systems, and dictation and electronic medical record software. Boasting seamless connectivity and data collection, these tools can now automate tasks that were previously performed manually, ensuring patient safety while achieving good image quality. freeing radiologists to focus on patient diagnosis and care. .
In this video, Dr. Daniele Marin, Medical Director of Multi-Dimensional Image Processing Laboratory at the Duke University School of Medicine, and Dr. Mahadevappa Mahesh, Chief Physicist at The Johns Hopkins Hospital discuss the balancing act between radiation dose and importance of image quality.
Ryan Lee, MD, Section Chief of Neuroradiology and Vice Chair of Safety and Quality at the Einstein Healthcare Network, expanded upon the ability to move toward personalized care using machine learning in contrast monitoring.
“Where we want to go is to use machine learning to tailor care specific to each patient, and we want to identify any variables,” Dr. Lee said. “For example, by using AI assisted protocols and existing software, we can configure CT protocols to optimize radiation dose for individual patients, and in a similar fashion, we may be able to optimize CT contrast dose, for each individual patient, accounting for multiple different factors.” During his presentation at the panel discussion, Dr. Lee explained how a contrast monitoring program helped his practice achieve lower extravasation rates and overall reduce costs.
Panel member Matthew J. Kuhn, MD, FACR , a clinical professor of Radiology at the University of Illinois College of Medicine who also serves as chief medical officer at A.I. Analysis, Seattle, Washington, has also seen the benefits of machine learning to reduce contrast dose. The concept of personalizing contrast dose as a result of algorithmic learning has translated into significant improvement in patient safety, he said.
Emilio Vega, manager of CT Quality and Safety at, NYU Langone Health pointed out that AI algorithms are also being used at his practice to detect anatomical landmarks and remove operator inconsistency during image post-processing.
Emilio Vega describes the use of AI in his department. View video.
Dr. Marin agreed that consistency can be achieved through the use of these tools, “If you have the same patient, coming multiple times and you have inconsistent enhancement or image quality because you are changing your injection protocol, that's a big deal for some applications where you're trying to identify the enhancement of a tumor. I think its critically important for certain applications to be consistent.”
This inevitable shift towards harnessing informatics and AI is expected to improve best practices in imaging. By incorporating these tools and learning capabilities too aggregate quantitative diagnostic information, radiologists are poised to play a central role in personalized medicine.
“This artificial intelligence augmentation of human capabilities is going to help us be better radiologists,” Dr. Tanenbaum said. “It's going to make us more efficient and more productive.”
In this video, Dr. Tanenbaum explains the importance of becoming value-driven and patient-focused in radiology.
“I think we're moving in from a thirty-thousand-foot point of view towards personalized medicine, precision medicine. Doing the right thing for the right patient and giving the right dose of contrast media is part of that.”
- Matthew J. Kuhn, MD, FACR
Radiology Informatics | The Impact of Data on Personalized Medicine. Appl Radiol.