The adoption of Transformer-based architectures in the medical domain is growing rapidly. In medical imaging, the analysis of complex shapes - such as organs, tissues, or other anatomical structures - combined with the often anisotropic nature of high-dimensional images complicates these adaptations. In this study, we critically examine the role of Positional Encodings (PEs), arguing that commonly used approaches may be suboptimal for the specific challenges of medical imaging. Sinusoidal Positional Encodings (SPEs) have proven effective in vision tasks, but they struggle to preserve Euclidean distances in higher-dimensional spaces. Isotropic Fourier Feature Positional Encodings (IFPEs) have been proposed to better preserve Euclidean distances, but they lack the ability to account for anisotropy in images. To address these limitations, we propose Anisotropic Fourier Feature Positional Encoding (AFPE), a generalization of IFPE that incorporates anisotropic, class-specific, and domain-specific spatial dependencies. We systematically benchmark AFPE against commonly used PEs on multi-label classification in chest X-rays, organ classification in CT images, and ejection fraction regression in echocardiography. Our results demonstrate that choosing the correct PE can significantly improve model performance. We show that the optimal PE depends on the shape of the structure of interest and the anisotropy of the data. Finally, our proposed AFPE significantly outperforms state-of-the-art PEs in all tested anisotropic settings. We conclude that, in anisotropic medical images and videos, it is of paramount importance to choose an anisotropic PE that fits the data and the shape of interest.

Large language models (LLMs) have advanced the field of artificial intelligence (AI) in medicine. However LLMs often generate outdated or inaccurate information based on static training datasets. Retrieval augmented generation (RAG) mitigates this by integrating outside data sources. While previous RAG systems used pre-assembled, fixed databases with limited flexibility, we have developed Radiology RAG (RadioRAG) as an end-to-end framework that retrieves data from authoritative radiologic online sources in real-time. RadioRAG is evaluated using a dedicated radiologic question-and-answer dataset (RadioQA). We evaluate the diagnostic accuracy of various LLMs when answering radiology-specific questions with and without access to additional online information via RAG. Using 80 questions from RSNA Case Collection across radiologic subspecialties and 24 additional expert-curated questions, for which the correct gold-standard answers were available, LLMs (GPT-3.5-turbo, GPT-4, Mistral-7B, Mixtral-8x7B, and Llama3 [8B and 70B]) were prompted with and without RadioRAG. RadioRAG retrieved context-specific information from www.radiopaedia.org in real-time and incorporated them into its reply. RadioRAG consistently improved diagnostic accuracy across all LLMs, with relative improvements ranging from 2% to 54%. It matched or exceeded question answering without RAG across radiologic subspecialties, particularly in breast imaging and emergency radiology. However, degree of improvement varied among models; GPT-3.5-turbo and Mixtral-8x7B-instruct-v0.1 saw notable gains, while Mistral-7B-instruct-v0.2 showed no improvement, highlighting variability in its effectiveness. LLMs benefit when provided access to domain-specific data beyond their training data. For radiology, RadioRAG establishes a robust framework that substantially improves diagnostic accuracy and factuality in radiological question answering.

November 23, 2022 -- The Radiological Society of North America (RSNA) has announced the official results of the RSNA Cervical Spine Fracture AI Challenge. Conducted by RSNA in collaboration with the American Society of Neuroradiology (ASNR) and the American Society of Spine Radiology (ASSR), the aim of the challenge was to explore whether artificial intelligence (AI) could be used to aid in the detection and localization of cervical spine injuries. The top eight teams will be recognized in a presentation on Nov. 28, in the AI Showcase during RSNA's 108th Scientific Assembly and Annual Meeting in Chicago (RSNA 2022). The RSNA Cervical Spine Fracture AI Challenge was conducted on a platform provided by Kaggle, Inc. The top performing competitors will be awarded a total of $30,000.

Philips' industry-first Tube for Life guarantee minimizes lifetime operating costs and provides reliability to help ensure efficient operation Amsterdam, the Netherlands and Chicago, USA – Royal Philips (NYSE: PHG, AEX: PHIA), a global leader in health technology, today announced new additions to its CT imaging portfolio at the Radiological Society of North America (RSNA) annual meeting (November 28 – December 2, Chicago, USA). The new CT 5100 – Incisive – features CT Smart Workflow [1], a comprehensive suite of artificial intelligence* (AI) enabled capabilities designed to accelerate CT workflows, enhance diagnostic confidence, and maximize equipment up-time, helping imaging services to enhance patient outcomes, improve department efficiency, reduce operational costs, and meet ambitious financial objectives. CT 5100 – Incisive – with CT Smart Workflow [1] includes Philips' Tube for Life guarantee, which over the lifetime of the scanner can potentially lower operating expenses by an estimated USD 420,000 [2][3]. This newest CT innovation from Philips also provides access to Philips' Technology Maximizer program, which provides users with the latest software and hardware updates as they are released. "With the combination of CT 5100 – Incisive – and CT Smart Workflow, we have embedded AI into the tools that radiology departments use every day so they can apply their expertise to the patient, rather than unnecessary distractions associated with the CT imaging itself," said Frans Venker, General Manager of Computed Tomography at Philips.

An example of a fundus eye images taken from the UK Biobank. A simple eye exam combined with powerful artificial intelligence (AI) machine learning technology could provide early detection of Parkinson's disease, according to research being presented at the annual meeting of the Radiological Society of North America (RSNA). Parkinson's disease is a progressive disorder of the central nervous system that affects millions of people worldwide. Diagnosis is typically based on symptoms like tremors, muscle stiffness and impaired balance -- an approach that has significant limitations. "The issue with that method is that patients usually develop symptoms only after prolonged progression with significant injury to dopamine brain neurons," said study lead author Maximillian Diaz, a biomedical engineering Ph.D. student at the University of Florida in Gainesville, Florida. "This means that we are diagnosing patients late in the disease process."

The upper panels show the craniocaudal and the mediolateral oblique views. The lower panels show a close-up of the left breast... view more OAK BROOK, Ill. - Artificial intelligence (AI) can enhance the performance of radiologists in reading breast cancer screening mammograms, according to a study published in Radiology: Artificial Intelligence. Breast cancer screening with mammography has been shown to improve prognosis and reduce mortality by detecting disease at an earlier, more treatable stage. However, many cancers are missed on screening mammography, and suspicious findings often turn out to be benign. An earlier study from Radiology found that, on average, only 10% of women recalled from screening for additional diagnostic workup based on suspicious findings are ultimately found to have cancer.

Andy OramFans of data in health care often speculate about what clinicians and researchers could achieve by reducing friction in data sharing. What if we had easy access to group repositories, expert annotations and labels, robust and consistent metadata, and standards without inconsistencies? Since 2017, the Radiological Society of North America (RSNA) has been displaying a model for such data sharing. That year marked RSNA's first AI challenge. RSNA has worked since then to make the AI challenge an increasingly international collaboration.

IMAGE: Schematic diagram of the proposed multichannel deep neural network model analyzing multiscale functional brain connectome for a classification task. OAK BROOK, Ill. - Deep learning, a type of artificial intelligence, can boost the power of MRI in predicting attention deficit hyperactivity disorder (ADHD), according to a study published in Radiology: Artificial Intelligence. Researchers said the approach could also have applications for other neurological conditions. The human brain is a complex set of networks. Advances in functional MRI, a type of imaging that measures brain activity by detecting changes in blood flow, have helped with the mapping of connections within and between brain networks.

How do clinicians set rules that allow professionals "to make good use of technology to find patterns in complex data" but also "stop companies from extracting unethical value from those data?" Geis, from the American College of Radiology (ACR) Data Science Institute, is one of the authors of a joint statement that addresses the potential for the unethical use of data, the bias inherent in datasets, and the limits of algorithmic learning, and was the moderator of a session on the topic at the Radiological Society of North America (RSNA) 2019 Annual Meeting in Chicago. There's a very big grey area between an absolute ethical approach to data use and decisions that are profit-driven, he told Medscape Medical News. "Sitting on the sainthood side, I can stick to doing only what I see as good for my patients, maybe even taking vows of poverty," he said. "On the extreme other side, I'm doing things that put me in prison."

Artificial intelligence in medicine has made dramatic progress in recent years. However, much of this progress is seemingly scattered, lacking a cohesive structure for the discerning observer. In this article, we will provide an up-to-date review of artificial intelligence in medicine, with a specific focus on its application to radiology, pathology, ophthalmology, and dermatology. We will discuss a range of selected papers that illustrate the potential uses of artificial intelligence in a technologically advanced future.