Artificial intelligence (AI) is already transforming many aspects of our lives and clinicians stand at the brink of a revolution in the way healthcare is delivered. We don't yet know how this will unfold in clinical imaging, but it could eventually see machines autonomously interpreting images and integrating vast genetic and biochemical datasets for precision management of each patient. The radiology community has always been at the forefront of new technology -but the AI revolution could fundamentally change how clinical imaging is delivered. The Media have driven inflated expectations for the speed and success of AI in medicine and the future is probably more nuanced than we expect. In this talk the challenges and opportunities of using "Big Data" and AI in clinical radiology will be discussed as well as a vision for how machines and radiologists may interact in the future.
R EhmanMayo Clinic, Rochester, Minnesota, USA Introduction: Hepatic fibrosis is an excessive accumulation of proteins, particularly collagen, in the extracellular matrix of the liver, resulting from chronic inflammation and cellular injury. Hepatic fibrosis alone does not affect liver morphology unless it causes extensive hepatocellular necrosis or scarring and cirrhosis. Conventional MR imaging may show surface nodularity, regenerative nodules, expanded hilar and gallbladder fossa spaces, liver volume changes, and evidence of portal hypertension in advanced liver fibrosis. All of these are signs of advanced disease and morphologic imaging is not sensitive for detecting less than stage 4 fibrosis. MRI-based tissue characterization biomarkers that have proposed for assessing liver fibrosis include measurement of the apparent diffusion coefficient, (ADC), measurement of T1 or T1rho, and quantitative texture analysis without or with contrast material. Statistically significant changes from normal values have been demonstrated with all of these biomarkers advanced fibrosis is present. Like morphologic imaging, these biomarkers have not proved to be sensitive or specific for assessing fibrosis that is less than stage 4. It has long been known that fibrotic liver tissue is mechanically stiffer than normal liver and the degree of increased stiffness correlates with the severity of fibrosis. More recently, it was demonstrated that in progressive liver disease, the stiffness of hepatic tissue actually increases prior to the onset of histological fibrosis. This is thought to represent the effects of early changes in the macromolecular composition of the extracellular matrix (ECM). It has also been shown that increased ECM stiffness actually promotes the development of hepatic fibrosis through a process known as mechanotransduction. Magnetic resonance elastography (MRE) is an MRI-based technique for quantitatively assessing the mechanical properties of tissue [1,2,3]. The main clinical application at the present time is for noninvasive assessment of hepatic fibrosis. FDA-cleared upgrades to equip conventional MRI scanners for MRE are available from ...