The pace of regulatory clearance of artificial intelligence (AI) algorithms for radiology continues to accelerate, and numerous algorithms are becoming available for use in clinical practice. End ...users of AI in radiology should be aware that AI algorithms may not work as expected when used beyond the institutions in which they were trained, and model performance may degrade over time. In this article, we discuss why regulatory clearance alone may not be enough to ensure AI will be safe and effective in all radiological practices and review strategies available resources for evaluating before clinical use and monitoring performance of AI models to ensure efficacy and patient safety.
To achieve consensus on the performance, interpretation and reporting of MS imaging according to up-to-date guidelines using the Peer Learning Methodology.
We utilized the Peer Learning Methodology ...to engage our clinical and radiology colleagues, review the current guidelines, acheive consensus on imaging techniques and reporting standards. After implementing changes, we collected radiologist feedback on the impact of the optimized images on their interpretation.
Survey responders indicated a strong preference for the new protocol in terms of overall image quality, individual lesions conspicuity and confidence in the ability to detect an MS lesion. The new protocol was preferred for both MS diagnosis and MS surveillance in 25 of 28 responses.
The Peer Learning Methodology is an effective tool to standardize and improve MR imaging quality, interpretation and reporting for Multiple Sclerosis in accordance with current guidelines.
The ACR's mission statement identifies five pillars of excellence. One of its pillars is research. ACR is recognized by many as supporting one of the premier research endeavors sponsored by a ...professional medical society of which the ACR Clinical Research Center is the largest component. The center is comprised of four entities: ACRIN(®), RTOG(®), QRRO(®), and ACR Image Metrix™. The Clinical Research Center encompasses personnel with extensive clinical trial expertise, a state-of-the-art IT infrastructure, and an imaging and radiation oncology core laboratory. This research enterprise supports a global network of researchers in the conduct of medical imaging and radiation oncology clinical trials. This paper's focus is on the Clinical Research Center's value to the radiology and radiation oncology professions, to the practices engaged in the clinical research, and to our patients.
Medical imaging accounts for 85% of digital health's venture capital funding. As funding grows, it is expected that artificial intelligence (AI) products will increase commensurately. The study's ...objective is to project the number of new AI products given the statistical association between historical funding and FDA-approved AI products.
The study used data from the ACR Data Science Institute and for the number of FDA-approved AI products (2008-2022) and data from Rock Health for AI funding (2013-2022). Employing a 6-year lag between funding and product approved, we used linear regression to estimate the association between new products approved in a certain year, based on the lagged funding (ie, product-year funding). Using this statistical relationship, we forecasted the number of new FDA-approved products.
The results show that there are 11.33 (95% confidence interval: 7.03-15.64) new AI products for every $1 billion in funding assuming a 6-year lag between funding and product approval. In 2022 there were 69 new FDA-approved products associated with $4.8 billion in funding. In 2035, product-year funding is projected to reach $30.8 billion, resulting in 350 new products that year.
FDA-approved AI products are expected to grow from 69 in 2022 to 350 in 2035 given the expected funding growth in the coming years. AI is likely to change the practice of diagnostic radiology as new products are developed and integrated into practice. As more AI products are integrated, it may incentivize increased investment for future AI products.
The purpose of this study was to evaluate the use of virtual monoenergetic images (VMI) in pre-operative CT angiography of potential donors for living donor adult liver transplantation (LDALT), and ...to determine the optimal energy level to maximize vascular signal-to-noise and contrast-to-noise ratios (SNR and CNR, respectively).
We retrospectively evaluated 29 CT angiography studies performed preoperatively in potential liver donors on a spectral detector CT scanner. All studies included arterial, early venous, and delayed venous phase imaging. Conventional polyenergetic images were generated for each patient, as well as virtual monoenergetic images in 10 keV increments from 40 –100 keV. Arteries (aorta and celiac, superior mesenteric, common hepatic, right and left hepatic arteries) were assessed on arterial phase images; portal venous system branches (splenic, superior mesenteric, main, right, and left portal veins) on early venous phase images; and hepatic veins on late venous phase images. Vascular attenuation, background parenchymal attenuation, and noise were measured on each set of virtual monoenergetic and conventional images.
Background hepatic and vascular noise decreased with increasing keV, with the lowest noise at 100 keV. Vascular SNR and CNR increased with decreasing keV and were highest at 40 keV, with statistical significance compared with conventional ( P < 0.05).
In preoperative CT angiography for potential liver donors, the optimal keV for assessing the vasculature to improve SNR and CNR is 40 keV. Use of low keV VMI in LDALT CT protocols may facilitate detection of vascular anatomical variants that can impact surgical planning.