AI in Radiology Practices

Today, Radiology published our latest study on breast cancer. This work, led by Felipe Oviedo Perhavec from Microsoft’s AI for Good Lab and Savannah Partridge (UW/Fred Hutch) in collaboration with researchers from Fred Hutch , University of Washington, University of Kaiserslautern-Landau, and the Technical University of Berlin, explores how AI can improve the accuracy and trustworthiness of breast cancer screening. We focused on a key challenge: MRI is an incredibly sensitive screening tool, especially for high-risk women—but it generates far too many false positives, leading to anxiety, unnecessary procedures, and higher costs. Our model, FCDD, takes a different approach. Rather than trying to learn what cancer looks like, it learns what normal looks like and flags what doesn’t. In a dataset of over 9,700 breast MRI exams—including real-world screening scenarios—our model: Doubled the positive predictive value vs. traditional models Reduced false positives by 25% Matched radiologists’ annotations with 92% accuracy Generalized well across multiple institutions without retraining What’s more, the model produces visual heatmaps that help radiologists see and understand why something was flagged—supporting trust, transparency, and adoption. We’ve made the code and methodology open to the research community. You can read the full paper in Radiology https://lnkd.in/gc82kXPN AI won't replace radiologists—but it can sharpen their tools, reduce false alarms, and help save lives.

Case Tuesday: Cardiac CT A patient presents with chest pain. The question is urgent: is this a heart attack waiting to happen, or something else? A CT coronary angiogram is performed. For the radiologist, this means carefully assessing coronary arteries, looking for stenosis, calcifications, and subtle plaques. The challenge: Coronary CTs generate hundreds of slices, often complex to interpret. Subtle plaques can be easily overlooked. Quantifying calcium scores and stenosis consistently takes significant time. This is where #AI is showing real promise: Automated calcium scoring to assess cardiovascular risk Plaque detection and quantification to support precise diagnosis Tools that standardize reporting and improve communication with cardiologists The radiologist’s expertise is essential in interpretation and clinical context but AI ensures that the assessment is faster, more reproducible, and more actionable. The impact: Earlier detection of coronary artery disease. Better risk stratification for patients with chest pain. Closer collaboration between radiology and cardiology teams As Chief Medical Officer at GE HealthCare, I see cardiac CT as a shining example of how AI doesn’t just enhance workflows it helps us move toward preventive, precision medicine that saves lives before catastrophe strikes. Do you see AI as the tipping point that will make cardiac CT more widely adopted as a first-line test for chest pain? #CaseTuesday #CardiacCT #AIinHealthcare #Radiology #HeartHealth #GEHealthcare

AI helps spot breast cancer on MRI scans by overlaying clear visual indicators, like colored heatmaps, that make suspicious areas stand out from healthy tissue. 🔬 The Core Visualization: Heatmaps & "Glowing" Cancer When AI analyzes a breast MRI, it creates a visual overlay to highlight areas of concern for the radiologist. This is primarily done through: · Spatially Resolved Heatmaps: The AI generates color-coded maps over the MRI image. Areas the model believes are abnormal are highlighted in color, allowing radiologists to focus on specific regions for further investigation. · Enhanced Delineation: Using advanced techniques like synthetic correlated diffusion imaging (CDI), AI can be optimized to make cancerous tissue appear to "glow" or "light up" next to healthy tissue, significantly improving the clarity of tumor boundaries. 🧠 How AI Achieves This Detection This capability is powered by AI models, often based on deep learning, that are trained on vast datasets of breast MRI exams. A key advancement is the development of explainable AI anomaly detection models. Unlike older systems: · They learn a robust pattern of what normal, benign breast tissue looks like. · They then flag significant deviations from this pattern, which helps in identifying malignancies even when such cases are rare in the training data. · This approach is particularly promising for screening populations, where cancer prevalence is low. 📊 Documented Performance and Potential Impact Research shows these AI systems are becoming highly effective clinical tools. The key findings are summarized below. Key Performance Metrics: · Detection Accuracy: Matches the performance of board-certified breast radiologists in clinical studies. · False Positive Reduction: Can potentially help avoid up to 20% of unnecessary biopsies in certain patient groups. · Generalizability: Models trained on large datasets (e.g., over 21,000 MRI exams) have shown robust performance when tested on external patient data from other institutions and countries. Primary Clinical Goal: The technology is designed as a powerful assistant to radiologists, not a replacement. It aims to improve reading efficiency, reduce unnecessary procedures, and help detect cancers that might otherwise be missed.

What if radiology AI didn’t need labels at all? 👉 https://lnkd.in/exC2zuTa As MRI demand surges, radiologists' time is becoming a true bottleneck, both in the clinical and in algorithm development. A new Radiology: Artificial Intelligence paper points to a more scalable path forward. The ALIGN framework shows how AI can learn neuroradiology directly from images and the free-text reports that already exist, without explicit manual labeling by radiologists. Key details from the study: 💢 Zero-label training: Trained on 63,178 MRI exams using paired images and radiology reports, no human annotations required (beyond writing the initial report, of course!) 🤝 Text-vision alignment: Uses NeuroBERT to align 3D MRI scans with the semantic meaning of reports ‼️ Strong triage performance: 0.95 AUC for normal vs abnormal detection 🏥 Robust generalization: Maintained high performance across four external hospitals (AUC 0.85–0.90) 🎯 Zero-shot detection: Accurately identified conditions it wasn’t explicitly trained on, including stroke, MS, and hemorrhage (mean AUC 0.89) 🔎 Visual-semantic search: Enables radiologists to retrieve example images by typing a text query ALIGN reframes anomaly detection as a self-supervised learning problem, eliminating the need for curated labels. Approaches like this could unlock more scalable ways to build radiology AI without adding to already overburden radiologists’ workload. 👏 Great work David Wood, Sina Kafiabadi, Kishan Dissanayake, Matt Townend, Yiran Wei, Dr Asif Mazumder, Peter Sasieni, Sebastien Ourselin FREng FMedSci, Chike Onyekwuluje, Permesh Singh Dhillon, Carolyn Costigan McConachie PhD, Kavi Fatania, Mark Igra, Hilmar Spohr, Thomas Booth et al.

Ever wondered how AI is actually making a difference in the real world, or in healthcare in particular? The FDA has now cleared over 750 AI-powered technologies in radiology alone. And when you look across all specialties, including cardiology, neurology, ophthalmology, and even wearable seizure detection devices - the total climbs to nearly 1,000 AI/ML-enabled medical tools cleared as of mid-2024. It’s a staggering figure that underscores how AI is reshaping the future of diagnostics far beyond just imaging. Let’s consider radiology more deeply as an example: The specialty sits at the intersection of data richness and diagnostic urgency. Imaging data - high-volume, high-resolution, and already digitized - is a natural fit for AI. The work radiologists do, while deeply specialized, is rooted in pattern recognition across structured image formats. That makes it fertile ground for machine learning - especially deep learning models that can spot anomalies faster, more reliably, and with expanding scope. And we’re already seeing real-world traction: ✅ AI triage tools are flagging critical cases like head CT hemorrhages, enabling faster intervention. ✅ AI-assisted mammogram reads are now matching the accuracy of double human reads in large-scale studies. ✅ Early pilots show AI can cut reporting times by nearly half without compromising diagnostic precision. ✅ Two-thirds of U.S. radiology departments already use AI in some form and that number is rising quickly. This is happening across healthcare, though radiology is a particularly illuminating proving ground for how AI can embed meaningfully into clinical practice - not as a novelty, but as core infrastructure. Regulatory clarity, measurable outcomes, and seamless workflow integration are already unfolding here - and other specialties are not far behind. Companies like Aidoc and Quibim are pushing boundaries with FDA-cleared tools clinicians actually rely on. Industry heavyweights like GE, Siemens, and Philips are no longer experimenting - they’re scaling. If you’re building AI to improve healthcare, please tell us a bit about your solution in the comments below!

New Publication! I’m excited to share that my latest research paper, “Deep learning-based detection and segmentation of osseous metastatic prostate cancer lesions on computed tomography,” has been published in the European Journal of Radiology Artificial Intelligence. In this study, we developed and evaluated deep learning models to automate lesion detection and segmentation in CT scans, aiming to enhance precision and efficiency in oncology imaging. Key Highlights: - Evaluated EfficientNet, ResNet34, and DenseNet for lesion detection - Tested nnUNetv2, UNet, ResUNet, and ResAttUNet for segmentation - Compared deep learning-based segmentation with manual radiomic analyses - Assessed models using F1 score, precision, recall, AUC, and Dice Similarity Coefficient (DSC) Our findings demonstrate the potential of AI-driven automation in improving metastatic lesion detection, paving the way for enhanced clinical decision-making and patient care. A huge thank you to my co-authors, mentors, and collaborators for their support in this work! Especially S J Pawan, PhD, and Shreyas Malewar 🔗 Read the full paper here: https://lnkd.in/gEiYrEN7 #deeplearning #AI #medicalimaging #cancerResearch #CTScans #precisionMedicine #radiology #healthcareInnovation

Unlocking the Future of Cancer Care with AI: A Comprehensive Review This review explores how artificial intelligence (AI)—powered by deep learning, vision transformers, and large language models—is revolutionizing oncology across every stage of care. From virtual biopsies in radiology to real-time diagnostics in pathology and dynamic treatment planning using multimodal data, AI is transforming how we detect, interpret, and treat cancer. Key highlights include: —Advances in radiomics and image-based tumor profiling —Integration of AI into digital pathology workflows —The role of LLMs in summarizing medical records and supporting clinical decisions —A spotlight on lung cancer and the need for personalized, AI-driven screening and treatment models —Challenges like model bias, validation, and data sharing—and how to overcome them Lung cancer shows the potential—and urgency. Despite new therapies, risk modeling and real-world data remain underused. AI can close that gap. The review also emphasizes the emergence of generalist medical AI systems that can process diverse data types and provide holistic, continuous decision support—from prevention to survivorship. This considers a must-read for anyone interested in the intersection of AI, medicine, and innovation. https://lnkd.in/eWdNSSBD