Ultrasound imaging is pivotal in clinical diagnostics due to its affordability, portability, safety, real-time capability, and non-invasive nature. It is widely utilized for examining various organs, such as the breast, thyroid, ovary, cardiac, and more. However, the manual interpretation and annotation of ultrasound images are time-consuming and prone to variability among physicians. While single-task artificial intelligence (AI) solutions have been explored, they are not ideal for scaling AI applications in medical imaging. Foundation models, although a trending solution, often struggle with real-world medical datasets due to factors such as noise, variability, and the incapability of flexibly aligning prior knowledge with task adaptation. To address these limitations, we propose an orchestration learning framework named PerceptGuide for general-purpose ultrasound classification and segmentation. Our framework incorporates a novel orchestration mechanism based on prompted hyper-perception, which adapts to the diverse inductive biases required by different ultrasound datasets. Unlike self-supervised pre-trained models, which require extensive fine-tuning, our approach leverages supervised pre-training to directly capture task-relevant features, providing a stronger foundation for multi-task and multi-organ ultrasound imaging. To support this research, we compiled a large-scale Multi-task, Multi-organ public ultrasound dataset (M²-US), featuring images from 9 organs and 16 datasets, encompassing both classification and segmentation tasks. Our approach employs four specific prompts-Object, Task, Input, and Position-to guide the model, ensuring task-specific adaptability. Additionally, a downstream synchronization training stage is introduced to fine-tune the model for new data, significantly improving generalization capabilities and enabling real-world applications. Experimental results demonstrate the robustness and versatility of our framework in handling multi-task and multi-organ ultrasound image processing, outperforming both specialist models and existing general AI solutions. Compared to specialist models, our method improves segmentation from 82.26% to 86.45%, classification from 71.30% to 79.08%, while also significantly reducing model parameters.

Recent vision-language foundation models deliver state-of-the-art results on natural image classification but falter on medical images due to pronounced domain shifts. At the same time, training a medical foundation model requires substantial resources, including extensive annotated data and high computational capacity. To bridge this gap with minimal overhead, we introduce MedBridge, a lightweight multimodal adaptation framework that re-purposes pretrained VLMs for accurate medical image diagnosis. MedBridge comprises three key components. First, a Focal Sampling module that extracts high-resolution local regions to capture subtle pathological features and compensate for the limited input resolution of general-purpose VLMs. Second, a Query Encoder (QEncoder) injects a small set of learnable queries that attend to the frozen feature maps of VLM, aligning them with medical semantics without retraining the entire backbone. Third, a Mixture of Experts mechanism, driven by learnable queries, harnesses the complementary strength of diverse VLMs to maximize diagnostic performance. We evaluate MedBridge on five medical imaging benchmarks across three key adaptation tasks, demonstrating its superior performance in both cross-domain and in-domain adaptation settings, even under varying levels of training data availability. Notably, MedBridge achieved over 6-15% improvement in AUC compared to state-of-the-art VLM adaptation methods in multi-label thoracic disease diagnosis, underscoring its effectiveness in leveraging foundation models for accurate and data-efficient medical diagnosis. Our code is available at https://github.com/ai-med/MedBridge.

Predicting the risk of developing breast cancer is an important clinical tool to guide early intervention and tailoring personalized screening strategies. Early risk models have limited performance and recently machine learning-based analysis of mammogram images showed encouraging risk prediction effects. These models however are limited to the use of a single exam or tend to overlook nuanced breast tissue evolvement in spatial and temporal details of longitudinal imaging exams that are indicative of breast cancer risk. In this paper, we propose STA-Risk (Spatial and Temporal Asymmetry-based Risk Prediction), a novel Transformer-based model that captures fine-grained mammographic imaging evolution simultaneously from bilateral and longitudinal asymmetries for breast cancer risk prediction. STA-Risk is innovative by the side encoding and temporal encoding to learn spatial-temporal asymmetries, regulated by a customized asymmetry loss. We performed extensive experiments with two independent mammogram datasets and achieved superior performance than four representative SOTA models for 1- to 5-year future risk prediction. Source codes will be released upon publishing of the paper.

In recent years, scientists have been trying to convert magnetic resonance imaging (MRI) images into computed tomography (CT) images for dose calculations while taking advantage of the benefits of MRI images. The main approaches for image conversion are bulk density, Atlas registration, and machine learning. These methods have limitations in accuracy and time consumption and require large datasets to convert images. In this study, the novel 'voxels spawn voxels' technique combined with the 'orthonormalize' feature in Carimas software was developed to build a conversion dataset from MRI intensity to Hounsfield unit value for some structural regions including gluteus maximus, liver, kidneys, spleen, pancreas, and colon. The original CT images and the converted MRI images were imported into the Geant4/Gamos software for dose calculation. It gives good results (<5%) in most organs except the intestine (18%).

We present a new algorithm to automatically convert 3-dimensional left atrium surface meshes into a standard 2-dimensional space: a Left Atrial Positioning System (LAPS). Forty-five contrast-enhanced 4- dimensional computed tomography datasets were collected from 30 subjects. The left atrium volume was segmented using a trained neural network and converted into a surface mesh. LAPS coordinates were calculated on each mesh by computing lines of longitude and latitude on the surface of the mesh with reference to the center of the posterior wall and the mitral valve. LAPS accuracy was evaluated with one-way transfer of coordinates from a template mesh to a synthetic ground truth, which was created by registering the template mesh and pre-calculated LAPS coordinates to a target mesh. The Euclidian distance error was measured between each test node and its ground truth location. The median point transfer error was 2.13 mm between follow-up scans of the same subject (n = 15) and 3.99 mm between different subjects (n = 30). The left atrium was divided into 24 anatomic regions and represented on a 2D square diagram. The Left Atrial Positioning System is fully automatic, accurate, robust to anatomic variation, and has flexible visualization for mapping data in the left atrium. This provides a framework for comparing regional LA surface data values in both follow-up and cohort studies.

Machine learning, particularly through advanced imaging techniques such as three-dimensional Magnetic Resonance Imaging (MRI), has significantly improved medical diagnostics. This is especially critical for diagnosing complex conditions like Alzheimer's disease. Our study introduces Triamese-ViT, an innovative Tri-structure of Vision Transformers (ViTs) that incorporates a built-in interpretability function, it has structure-aware explainability that allows for the identification and visualization of key features or regions contributing to the prediction, integrates information from three perspectives to enhance brain age estimation. This method not only increases accuracy but also improves interoperability with existing techniques. When evaluated, Triamese-ViT demonstrated superior performance and produced insightful attention maps. We applied these attention maps to the analysis of natural aging and the diagnosis of Autism Spectrum Disorder (ASD). The results aligned with those from occlusion analysis, identifying the Cingulum, Rolandic Operculum, Thalamus, and Vermis as important regions in normal aging, and highlighting the Thalamus and Caudate Nucleus as key regions for ASD diagnosis.

Artificial Intelligence (AI) research needs to be clinician led; however, expertise typically lies outside their skill set. Collaborations exist but are often commercially driven. Free and open-source computational algorithms and software expertise are required for meaningful clinically driven AI medical research. Deep learning algorithms automate segmenting regions of interest for analysis and clinical translation. Numerous studies have automatically segmented paranasal sinus computed tomography (CT) scans; however, openly accessible algorithms capturing the sinonasal cavity remain scarce. The purpose of this study was to validate and provide an open-source segmentation algorithm for paranasal sinus CTs for the otolaryngology research community. A cross-sectional comparative study was conducted with a deep learning algorithm, UNet++, modified for automatic segmentation of paranasal sinuses CTs and "ground-truth" manual segmentations. A dataset of 100 paranasal sinuses scans was manually segmented, with an 80/20 training/testing split. The algorithm is available at https://github.com/rheadkaul/SinusSegment. Primary outcomes included the Dice similarity coefficient (DSC) score, Intersection over Union (IoU), Hausdorff distance (HD), sensitivity, specificity, and visual similarity grading. Twenty scans representing 7300 slices were assessed. The mean DSC was 0.87 and IoU 0.80, with HD 33.61 mm. The mean sensitivity was 83.98% and specificity 99.81%. The median visual similarity grading score was 3 (good). There were no statistically significant differences in outcomes with normal or diseased paranasal sinus CTs. Automatic segmentation of CT paranasal sinuses yields good results when compared with manual segmentation. This study provides an open-source segmentation algorithm as a foundation and gateway for more complex AI-based analysis of large datasets.

The assessment of resting energy expenditure (REE) is a challenging task with the current existing methods. The reference method, indirect calorimetry (IC), is not widely available, and other surrogates, such as equations and bioimpedance (BIA) show poor agreement with IC. Body composition (BC), in particular muscle mass, plays an important role in REE. In recent years, computed tomography (CT) has emerged as a reliable tool for BC assessment, but its usefulness for the REE evaluation has not been examined. In the present study we have explored the usefulness of CT-scan imaging to assess the REE using AI machine-learning models. Single-centre observational cross-sectional pilot study from January to June 2022, including 90 fasting, clinically stable adults (≥18 years) with no contraindications for indirect calorimetry (IC), bioimpedance (BIA), or abdominal CT-scan. REE was measured using classical predictive equations, IC, BIA and skeletal CT-scan. The proposed model was based on a second-order linear regression with different input parameters, and the output corresponds to the estimated REE. The model was trained and tested using a cross-validation one-vs-all strategy including subjects with different characteristics. Data from 90 subjects were included in the final analysis. Bland-Altman plots showed that the CT-based estimation model had a mean bias of 0 kcal/day (LoA: -508.4 to 508.4) compared with IC, indicating better agreement than most predictive equations and similar agreement to BIA (bias 53.4 kcal/day, LoA: -475.7 to 582.4). Surprisingly, gender and BMI, ones of the mains variables included in all the BIA algorithms and mathematical equations were not relevant variables for REE calculated by means of AI coupled to skeletal CT scan. These findings were consistent with the results of other performance metrics, including mean absolute error (MAE), root mean square error (RMSE), and Lin's concordance correlation coefficient (CCC), which also favored the CT-based method over conventional equations. Our results suggest that the analysis of a CT-scan image by means of machine learning model is a reliable tool for the REE estimation. These findings have the potential to significantly change the paradigm and guidelines for nutritional assessment.

The fetal standard plane acquisition task aims to detect an Ultrasound (US) image characterized by specified anatomical landmarks and appearance for assessing fetal growth. However, in practice, due to variability in human operator skill and possible fetal motion, it can be challenging for a human operator to acquire a satisfactory standard plane. To support a human operator with this task, this paper first describes an approach to automatically predict the fetal head standard plane from a video segment approaching the standard plane. A transformer-based image predictor is proposed to produce a high-quality standard plane by understanding diverse scales of head anatomy within the US video frame. Because of the visual gap between the video frames and standard plane image, the predictor is equipped with an offset adaptor that performs domain adaption to translate the off-plane structures to the anatomies that would usually appear in a standard plane view. To enhance the anatomical details of the predicted US image, the approach is extended by utilizing a second modality, US probe movement, that provides 3D location information. Quantitative and qualitative studies conducted on two different head biometry planes demonstrate that the proposed US image predictor produces clinically plausible standard planes with superior performance to comparative published methods. The results of dual-modality solution show an improved visualization with enhanced anatomical details of the predicted US image. Clinical evaluations are also conducted to demonstrate the consistency between the predicted echo textures and the expected echo patterns seen in a typical real standard plane, which indicates its clinical feasibility for improving the standard plane acquisition process.

The automatic screening of thyroid nodules using computer-aided diagnosis holds great promise in reducing missed and misdiagnosed cases in clinical practice. However, most current research focuses on single-modal images and does not fully leverage the comprehensive information from multimodal medical images, limiting model performance. To enhance screening accuracy, this study uses a deep learning framework that integrates high-dimensional convolutions of B-mode ultrasound (BMUS) and strain elastography (SE) images to predict the malignancy of TI-RADS 4 thyroid nodules with high-risk features. First, we extract nodule regions from the images and expand the boundary areas. Then, adaptive particle swarm optimization (APSO) and contrast limited adaptive histogram equalization (CLAHE) algorithms are applied to enhance ultrasound image contrast. Finally, deep learning techniques are used to extract and fuse high-dimensional features from both ultrasound modalities to classify benign and malignant thyroid nodules. The proposed model achieved an AUC of 0.937 (95 % CI 0.917-0.949) and 0.927 (95 % CI 0.907-0.948) in the test and external validation sets, respectively, demonstrating strong generalization ability. When compared with the diagnostic performance of three groups of radiologists, the model outperformed them significantly. Meanwhile, with the model's assistance, all three radiologist groups showed improved diagnostic performance. Furthermore, heatmaps generated by the model show a high alignment with radiologists' expertise, further confirming its credibility. The results indicate that our model can assist in clinical thyroid nodule diagnosis, reducing the risk of missed and misdiagnosed diagnoses, particularly for high-risk populations, and holds significant clinical value.