Search

Comparative Evaluation of BI-RADS Classification, Clinical Management, and Diagnostic Performance in Breast Ultrasound Reports Using ChatGPT-4o and DeepSeek-V3 | Radiology Key

Two radiology-focused LLM evaluations compared BI-RADS classifications and management for breast ultrasound findings, using a fixed prompt, blinded readings, and standard performance metrics against pathology or follow-up as the reference.

AI will help radiologists. Pancreatitis or bowel obstruction will be detected faster

Translated title. Original: AI pomoże radiologom. Zapalenie trzustki czy niedrożność jelita będą szybciej wykryte

Researchers built a vision-language AI tool to speed up RUQ ultrasound interpretation, aiming to assist radiologists by producing initial, review-ready reports that combine image data with clinical information; initial validations show high diagnostic alignment with radiologist assessments, with real-world testing planned before deployment.

The effects of teleradiology - Radiology

Translated title. Original: A teleradiológia hatásai - Radiológia

The document reviews the clinical and organizational aspects of teleradiology, outlining its workflow, historical developments, benefits, adoption during the COVID-19 era, global penetration, and barriers, supported by international studies and case examples.

Lusíadas Health: Lusíadas Group opens Imaging Center... | Teamlyzer

Translated title. Original: Lusíadas Saúde: Grupo Lusíadas abre Centro de Imagem... | Teamlyzer

Grupo Lusíadas opens an imaging center at Hospital Lusíadas Lisboa, using AI-assisted diagnostic imaging and aligning assessments with international clinical guidelines; the unit includes Mammo Via Verde with a priority circuit and combines mammography, tomosynthesis, MRI, and ultrasound.

Artificial Intelligence has arrived in radiology. But where does it fit into the workflow? - Mobilemed

Translated title. Original: A Inteligência Artificial chegou à radiologia. Mas onde ela entra no workflow? - Mobilemed

AI is now part of radiology workflows when integrated into the DICOM viewer, supporting image analysis without supplanting clinician judgment. The focus is on seamless integration within existing PACS/workflow to minimize interruptions and keep radiologists at the center of decision-making.

AI was supposed to take over the job—but radiologists ended up with more work – Nerikes Allehanda

Translated title. Original: AI skulle ta jobben – men radiologerna fick mer att göra – Nerikes Allehanda

The article uses Jevons paradox to explain why AI has not reduced radiologists’ workload: cheaper image analysis can lead to more use of medical imaging. It says AI may improve access to care and prevention, but also raises concerns about budget cuts and access to health data.

Answers to NMO tests: Artificial Intelligence in Radiology: Modern Approaches to Mammogram Analysis - on RosMedInfo

Translated title. Original: Ответы на тесты НМО: Искусственный интеллект в рентгенологии: современные подходы к анализу маммограмм - на РосМедИнфо

The document describes AI concepts and their application to radiology, focusing on mammogram analysis, imaging modalities, model training and clinical integration, while stressing AI as a decision-support tool that requires physician oversight.

What’s Next for AI in Radiology?: Facilitating Data Orchestration Over Fragmentation | Diagnostic Imaging

AI should be embedded in health-care workflows to connect data and people, improving care delivery by reducing fragmentation. Success requires governance, interoperability, and cross-disciplinary collaboration, moving from isolated applications to enterprise-wide orchestration.

AI can assist in performing ultrasound | Learning in Poland

Translated title. Original: AI może wspomóc wykonywanie USG | Nauka w Polsce

Researchers built an AI tool that analyzes RUQ ultrasound images by combining visual data with clinical information to generate preliminary radiology reports, aiming to save radiologists' time and speed patient care. Validation on ~11,000 scans shows high accuracy but stresses the need for real-world testing and human oversight.

AI Radiology Solutions For CT Scans | Avicenna.AI

Avicenna.AI says its tools analyze CT scans as they are acquired, flagging urgent and incidental findings and standardizing scoring. The page cites studies reporting faster alerts with CINA-PE and higher scoring AUC with CINA-ASPECTS.