Stay Ahead of the Curve in Radiology AI.
RadAI Slice is your weekly intelligence briefing on the most critical developments at the intersection of radiology and artificial intelligence. Stop searching. Start leading.Subscribe to join 11k+ radiology professionals tracking the future of imaging AI.
Featured In
Lastest Issues
The latest developments in Radiology & AI.
Reports & Market Maps
Explore our deeper briefings on radiology AI trends, market structure, and event-level intelligence.
SIIM 2026 AI Landscape
Analysis of SIIM 2026 sessions, themes, speaker mix, and abstract activity through the lens of imaging AI.
ECR 2026 AI Product Landscape
A launch-level view of ECR 2026 product announcements: AI share, product mix, and modality focus.
ECR 2026 AI Industry Landscape
A data-driven map of ECR 2026 exhibitors: AI share, vendor mix, and regional patterns.
Your Weekly Slice of Innovation
Each issue is precisely structured to give you exactly what you need. No fluff, just facts and forward-looking insights.
Recent Industry News
The Latest Research
FDA Approvals Database
Latest News & Trends
Private Equity Backs AIRS Medical to Expand MRI AI Globally
TA Associates is investing in AIRS Medical to accelerate its global expansion of AI-powered MRI efficiency solutions.
Radiology Maintains Lead in FDA-Cleared AI Algorithms, Cardiology Follows
Radiology remains the top specialty for FDA-cleared AI, with cardiology as a strong second, particularly in cardiovascular imaging.
AI Clinical Reasoning in Diagnostics and Digital Fatigue in Healthcare
Recent JMIR features explore large language models in clinical diagnostics and digital fatigue among healthcare professionals.
From the Research Hub
An advanced ensemble deep learning framework for accurate multi-class lung cancer classification using IUNet++ and MResNext.
Early and precise diagnosis is crucial for increasing patient survival because lung cancer is still one of the top causes of cancer-related death globally. Even while deep learning-based methods have shown encouraging results in the research of lung cancer, many of the existing models have significant computational complexity, overfitting, and limited capacity for generalization. An ensemble deep learning framework for automated lung cancer diagnosis from CT scan images is proposed in this research. The CT images are first pre-processed utilizing Gabor filtering, CLAHE-based contrast enhancement, and data augmentation to increase image quality and address class imbalance. An Improved UNet++ (IUNet++) model combined with a Convolutional Block Attention Module (CBAM) is then used to segment the lung tumor region in order to improve feature representation and localization accuracy. The most informative features are then selected using the Enhanced Elephant Herding Optimization Algorithm (EEHOA) after histogram, texture, binary, and rotational-scale-translational (RST) features are extracted. Lastly, lung cancer is classified into benign, malignant, and normal categories using a Modified ResNeXt (MResNeXt) model that incorporates Leaky ReLU and an improved layer structure. The IQ-OTH/NCCD lung cancer CT dataset, which is accessible to the public, was used for the experiments. 99% accuracy, 99% precision, 99% recall, and 98.66% F1-score were attained with the proposed framework. The proposed approach beat a number of contemporary state-of-the-art techniques, including CNN, KNN, Modified YOLOv3, ShuffleNet, and EfficientNet-based models, according to a comparative analysis. Additionally, compared to traditional deep learning techniques, the incorporation of attention-guided segmentation, better feature selection, and an improved ResNeXt architecture decreased model complexity and increased training efficiency. For multi-class lung cancer diagnosis using CT images, the proposed ensemble framework offers a practical and computationally efficient solution. The outcomes of the experiments show that it has the potential to be a dependable computer-aided diagnostic tool for clinical decision support.
A clinically anchored radiomics dictionary for explainable TI-RADS-based thyroid nodule classification in ultrasound; dictionary version TU1.0.
Artificial Intelligence-based radiomics models for thyroid ultrasound (US) often lack interpretability, limiting clinical trust. This study develops and evaluates an interpretable radiomic feature (RF) framework for thyroid-nodule classification by linking quantitative-US features to the Thyroid-Imaging-Reporting-and-Data-System (TI-RADS) semantic lexicon through a clinically grounded radiomics dictionary. A radiomics dictionary was constructed to map TI-RADS categories, including composition, echogenicity, shape, margin, and echogenic foci, to Image-Biomarker-Standardization-Initiative-compliant RFs extracted from two-dimensional-US images. Relationships were defined through expert consensus (four physicians, three physicists, one radiology expert, one biologist) and further examined using Shapley-Additive-Explanations (SHAP) as a model-based interpretability analysis. Three multicenter datasets were combined, yielding 5,542 nodules, from which 107-RFs were extracted using PyRadiomics and normalized with min-max scaling. Twenty-seven feature-selection methods were paired with twenty-five classifiers and evaluated using stratified five-fold cross-validation on 70 % of the data, followed by evaluation on a held-out multicenter testing set comprising the remaining 30 % for benign-versus-malignant nodule classification. Robust model selection employed a stability-aware-composite-scoring framework combining mean performance and variability across accuracy, precision, recall, F1-score, and Receiver-Operating-Characteristic-Area-Under-the-Curve (ROC-AUC). The dictionary enabled direct interpretation of radiomic signatures in TI-RADS terms. The Select-From-Model (logistic regression) plus Extra-Trees classifier achieved strong testing performance (ROC-AUC:0.941 ± 0.004). SHAP identified texture heterogeneity as the dominant malignancy signal, with Gray Level Run Length Matrix non-uniformity, intensity dispersion, and kurtosis aligning predictions with high-risk TI-RADS descriptors. This study introduces an interpretable radiomics dictionary and stability-aware model selection framework, addressing interpretability limitations and enabling transparent thyroid nodule risk stratification from US.
An Explainable Multimodal AI Framework with Reinforcement Learning for Post-Surgical Clinical Decision Support
Post-surgical adverse outcomes, including mortality, intensive care readmission, and complications, remain major challenges for clinical decision-making. Existing machine learning approaches focus on outcome prediction while operating as opaque systems, limiting clinical trust and the translation of predictions into treatment decisions, and many clinical studies rely on synthetic data in which shared intermediate variables create circular dependencies between inputs and targets that compromise reported performance. We aimed to develop an explainable multimodal architecture and a rigorous evaluation methodology that address these gaps. We designed a two-stage architecture integrating supervised deep learning for risk prediction with conservative Q-learning for action recommendation. The first stage uses five modality-specific encoders for structured records, physiological time-series, chest radiographs, clinical notes, and surgical metadata, unified through cross-modal attention into a shared patient-state representation. The second stage applies offline reinforcement learning to recommend clinical actions while preventing value overestimation. We formally characterized a target-leakage flaw in synthetic pipelines and propose a real-data methodology using a verified clinical database, with event-censored temporal separation and uncertainty-weighted per-task training. Component-level behavior was validated on a controlled synthetic benchmark, demonstrating that the architecture functions as designed without claiming clinical validity. The cross-modal attention and risk-prediction components behaved as expected, whereas the offline reinforcement learning stage did not converge on the benchmark, indicating that value estimation requires further investigation on real clinical data. The architecture provides dual-level explainability through attention visualization and value decomposition, contributing a deployable design, a formal methodological critique of synthetic-data practices, and a complete framework for clinically valid evaluation.
From the RadAI FDA Approvals Database
Allengers Medical Systems Limited
Digital C-ARM with 3D
The Digital C-ARM with 3D is a mobile fluoroscopic X-ray system that provides real-time 2D and 3D imaging. It assists clinicians in visualizing patient anatomy during procedures, improving accuracy and patient outcomes.
Sonio
Sonio Suspect
Sonio Suspect is AI-powered diagnostic software that helps clinicians identify lesions in medical images that may be suspicious for cancer, enhancing early detection and aiding diagnostic decisions.
Topcon Healthcare Solutions EMEA Oy
Harmony Referral System with Harmony OCT Viewer (Octave)
The Harmony Referral System with Harmony OCT Viewer (Octave) is a system designed to manage and view ophthalmic images, specifically OCT (Optical Coherence Tomography) scans. It helps clinicians by organizing, displaying, and facilitating the analysis of retinal images, aiding in the diagnosis and management of eye conditions.
The Sharpest Insights, Effortlessly
Save Time
We scour dozens of sources so you don't have to. Get all the essential information in a 5-minute read.
Stay Informed
Never miss a critical update. Understand the trends shaping the future of your practice and research.
Gain an Edge
Be the first to know about the tools and technologies that matter, from clinical practice to academic research.
Ready to Sharpen Your Edge?
Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.
We respect your privacy. Unsubscribe at any time.