Reflections on RSNA 2022
- November 29, 2022
Welcome back to our Physician Education Hub, and greetings from Chicago. I’ve been at RSNA many times, and in recent years I have witnessed the growth in imaging technologies directed at the important challenge of diagnosing and treating liver disease. This year has been no exception, with notable developments in the applications of MRI, PET/CT, and ultrasound for the detection and diagnosis of both chronic liver disease and cancer (e.g., LI-RADS), highlighting the vital role radiology plays in this field. There have also been some promising developments in the utility of MRI in treatment planning and monitoring the effectiveness of interventions (e.g., Yttrium-90 therapy) in patients with liver cancer.
Delving deeper into these exciting times for liver imaging, an area of particular interest at RSNA, is the growing role of artificial intelligence (AI) in gastrointestinal radiology, which is covered in several sessions. On this topic, I’d like to introduce a paper recently published in Applied Radiology, where Dr. Matt Kelly walks us through how AI and convolutional neural network (CNN)-based approaches can save radiology time while also reducing subjectivity and inter-reader variability associated with conventional techniques. For example, AI- and CNN-based approaches have been used successfully in liver volumetry and quantitative three-dimensional reconstruction of the biliary tree from MRI (see Figure 1). The article also recognizes the need for widespread clinical adoption of AI in regulatory-cleared products to enable seamless integration into the radiology workflow. This should bridge the gap to the clinic and address challenges in liver disease management, imaging chronic liver diseases, and liver cancer, while ensuring that the role of radiology in managing these conditions continues to evolve.
-Dr. Carlos Duncker
- Dr. Carlos Duncker and Dr. Matt Kelly are Perspectum employees.
- LiverMultiScan, Hepatica, cT1 and MRCP+ are manufactured by Perspectum.