Medical Image Analysis

Accurate prenatal screening from ultrasound imaging is severely constrained by the extreme spatial heterogeneity of fetal structures and the data scarcity inherent to rare congenital anomalies. To overcome these computational bottlenecks, we present RadarMAE, a data-efficient self-supervised representation learning framework tailored for fetal ultrasound analysis. By employing a novel, structure-aware masking strategy, RadarMAE enables robust pre-training on highly limited clinical datasets. Benchmarking the framework on first-trimester screening for trisomy 21 (Down syndrome), our model achieved exceptional diagnostic performance on an internal cohort, with an area under the curve (AUC) of 0.99, 97% accuracy, 94% sensitivity, and 98% specificity. Crucially, through domain-adversarial adaptation, the framework demonstrates robust cross-institutional generalization under extreme source-target data imbalance, bounding performance degradation to less than or equal to 0.04 across all key metrics in external testing. Furthermore, by integrating spatial registration with explainable AI, we constructed the first population-level facial risk atlas for trisomy 21. This data-driven interpretability framework validates standard clinical markers while revealing the comparable diagnostic weight of underutilized features, such as mandibular dysmorphology. Ultimately, this approach provides a scalable, transparent, and high-performance solution for equitable prenatal diagnostics and phenotypic biomarker discovery.

Timely detection of fetal distress during labor is a central pursuit of modern obstetric care. The primary measure used in this regard is fetal heart rate and contraction monitoring (echocardiotocography or CTG), which provides a continuous signal during labor. The vast majority of machine learning applications to CTG assume access to a 30 to 60 minute window just prior to delivery, which makes an untenable assumption that the timing of delivery is known. In this work, we reframe the problem as a time-to-event prediction task. We develop deep time-to-event models to predict the joint probability of imminent delivery and fetal acidosis, a biomarker of fetal hypoxia collected from umbilical cord gas after delivery. We develop a framework for this problem, drawing on deep discrete-time survival methods and marked point processes, adapting each to this particular setting (i.e., continuous physiological signals, fully observed events, partially observed labels). One approach dubbed Marked DeepHit accurately estimates delivery within one hour at multiple elapsed-time landmarks, with test-set AUROC of 0.837 [95% CI 0.692--0.970] and 0.935 [0.858--0.964] at 6 and 12 hours into labor, respectively. We find that these models also achieve competitive performance near delivery time compared to prior models trained specifically on the end of tracing, and outperform other clinical feature-based proposals, both on internal and external validation at a second site. These results give a proof-of-principle to support future prospective deployment of AI-EFM during labor.

Text-driven medical image segmentation demands unified models based on multi-modal learning, capable of processing diverse imaging modalities while providing spatially reliable prediction estimates for clinical practice. Nevertheless, existing vision-language model solutions suffer from poor domain generalization and exhibit limited capability for uncertainty estimation to model the cumulative unreliability of cross-modal alignment across network layers. In this paper, we propose UniMedSeg, a unified vision-language framework for multi-modal medical image segmentation built upon frozen pre-trained vision-language backbones. Our framework incorporates modality-aware adaptation, cross-layer uncertainty-guided encoding, and a frequency-spatial dual-prompt decoder, wherein the progressive uncertainty propagation mechanism of the encoder suppresses misleading cross-modal features, whereas the decoder disentangles anatomical structure and boundary details in the frequency domain under dual textual prompts. Auxiliary losses further sharpen boundary representations and calibrate pixel-wise uncertainty maps. Extensive experiments demonstrate that UniMedSeg preserves competitive in-domain segmentation performance, improves domain generalization, and yields clinically useful spatial risk estimation.

Purpose: To develop and evaluate a recurrent neural network that operates directly on undersampled golden-angle radial k-space, enabling low-latency cardiac cine reconstruction for interventional cardiac MRI (iCMR). Methods: KOSTAL (K-space Optimized Spatial Temporal Architecture for Low latency) combines a per-coil k-space convolutional gated recurrent unit (convGRU), which learns a mask-conditioned view-sharing rule over the accumulated k-space, with an image-space recurrent module. It was trained and tested on the open-access multi-coil OCMR dataset using 10 simulated golden-angle spokes per frame against a 1000-spoke reference. Ablations varied recurrent module architecture and placement; temporal behavior was assessed using a imaging-plane transition test. Generalization was evaluated without retraining in 23 pediatric patients prospectively scanned during free-breathing using a radial golden-angle bSSFP sequence with binned XD-GRASP and clinical breath-held cine serving as references. Results: At 10 spokes per frame, KOSTAL reached a median SSIM of 0.907 (gridding: 0.622) in under 34 ms of processing time per frame. Image quality (IQ) matched gridded results formed from 7-11 times more spokes. Recurrence in k-space rather than image space accounted for most of this gain (0.891 versus [≤]0.650). IQ recovered from an abrupt plane transition in 75 ms, versus 303 ms for 200-spoke gridding. KOSTAL had the best agreement with the binned reference and the highest long-axis CNR. KOSTAL accurately tracked left-ventricular blood-pool volume (end-diastolic volume bias -0.7 mL) with a shorter lag (71 ms) than iterative reconstruction (239-287 ms). Conclusion: Learned, mask-conditioned recurrence in k-space yields high image quality at a latency compatible with interventional guidance.

Coarse to fine segmentation is a natural strategy for three dimensional lesion delineation becausefull resolution processing of an entire scan is expensive while only a small fraction of the volume contains pathology. Its main failure mode, however, is structural: a lesion missed by the coarse stage is usually absent from the region of interest and therefore cannot be recovered by the fine stage. This problem is most severe for small, low contrast, and multifocal lesions. We introduce RECALLREFINE, a coverage constrained coarse to fine framework that treats candidate-region selection as a recall critical learning problem rather than a deterministic crop around the coarse mask.A full-volume network first predicts coarse lesion probabilities and multi-scale features. A candidate score network then combines coarse probability, predictive uncertainty, feature novelty, and spatial context to score overlapping 3D blocks. During training, a differentiable selector is optimized with a component-wise coverage objective that requires every annotated lesion instance to remain covered by at least one selected block under a fixed refinement budget. Selected blocks are processed at native resolution by a local refiner with a context-ring branch, and overlapping predictions are merged back into the full volume. Across four public 3D lesion benchmarks, RECALLREFINE improves mean tumor Dice from 76.5% for the strongest matched coarse-to-fine baseline to 78.4%, while increasing sensitivity for lesions below 10 mm from 60.9% to 70.8%.At a 20% candidate budget, the selector covers 94.4% of lesion instances, substantially reducing irreversible coarse-stage misses. The results suggest that the central problem in cascaded lesion segmentation is not only how to refine a detected lesion, but how to guarantee that difficult lesions are still represented in the refinement set.

Brain foundation models have shown promising advances in learning generalizable representations from functional magnetic resonance imaging (fMRI) data. However, existing models are typically tied to specific brain atlases and cannot be applied to fMRI data represented using different atlases without full retraining, thereby severely limiting their broader utility across neuroimaging studies. To address this bottleneck, we introduce AREION, atlas-agnostic representation learning of brain connectomes. AREION overcomes atlas dependency through two integrated innovations. First, a Brain Region Contextualizing technique maps brain regions and connectivity entries from diverse atlases into a unified representation space based on their spatial, functional, and morphological characteristics. Second, a Hierarchical Brain Transformer Encoder processes region-specific connectivity profiles as variable-length token sequences and jointly captures local functional structure and global brain functional organization, conditioned on brain regional context, without assuming a fixed number of regions. Together, these two techniques enable the learning of generalizable representations across brain atlases with different parcellation resolutions and connectivity dimensions. AREION is pretrained on a large-scale multi-atlas cohort of fMRI data via self-supervised modeling and evaluated across diverse downstream tasks. Extensive experiments show that it outperforms strong supervised baselines and fMRI foundation models while using substantially fewer parameters, demonstrating strong transferability to atlases unseen during pretraining, generalization to external datasets, and excellent interpretability at the brain regional level.

Magnetic resonance imaging (MRI) exhibits substantial scanner and protocol dependent appearance variations, making harmonization challenging without distorting patient specific anatomy, particularly in abdominal imaging. We propose AC-Diff, an anatomy--contrast disentangled diffusion framework that formulates MRI harmonization as a factor-specific generative intervention. Using aligned multi-contrast supervision and cross-patient factor swapping, AC-Diff learns to separate anatomical structure from acquisition-dependent contrast. Unlike conventional image or latent diffusion, AC-Diff restricts stochastic generation to the contrast subspace while the source anatomy bypasses diffusion and is directly reused during reconstruction. Experiments on an in-house multi-vendor cohort and the external Duke Liver MRI dataset demonstrate improved target-domain alignment with strong structural preservation. In downstream liver segmentation, AC-Diff improves Dice from 0.942 for the original inputs to 0.954 for the harmonized images. These results support contrast-specific latent generation as a promising approach to anatomy-preserving MRI harmonization.

Standard cardiac biomarkers such as ejection fraction, ventricular volumes, and strain compress complex three-dimensional heart dynamics into a small set of predefined measurements, leaving clinically important heterogeneity unrecognised. Here we propose a personalised normative cardiac framework in which each individual is compared to a digital twin representing expected healthy cardiac function given their age, sex, and body habitus. We show that structured patterns of deviation from these normative twins, discovered through systematic exploration of a generative cardiac dynamics space, constitute novel dynamic phenotypes that are not captured by conventional biomarkers. Applied independently to two UK Biobank cardiac cohorts with distinct shape representation, a 128-dimensional four-chamber statistical shape descriptor (HeartSSM, n = 52,365) and a 64-dimensional deep generative latent space from a conditional variational mesh autoencoder (MeshHeart cVAE, n = up to 39,376), we identify reproducible, continuous deviation-from-normative-twin scores (three discrete deviation phenotypes in the HeartSSM cohort) that improve cardiovascular risk stratification beyond ejection fraction, ventricular volumes, myocardial mass, and strain. The strongest effect was seen for incident atrial fibrillation among individuals with preserved ejection fraction (at least 50%): incremental C-index delta C = +0.068 (5-fold cross-validated delta C = +0.063; likelihood-ratio P = 3.3 x 10^-8), with a separate, smaller effect for incident stroke in the same preserved-EF subgroup in the independent MeshHeart cVAE representation (which showed no signal for atrial fibrillation) (delta C = +0.024, P = 2.6 x 10^-4). Both effects were checked against placebo controls in which the deviation scores were replaced by Gaussian noise. To test whether the normative twin generalises beyond UK Biobank's specific scanning and processing pipeline, we applied the MeshHeart cVAE normative twin, unmodified, to an independent external cohort of healthy volunteers imaged at a different hospital (GenScan/Hammersmith, n = 315): deviation scores in this cohort were of comparable magnitude to the UK Biobank healthy reference (ratio of means = 1.09; Mann-Whitney P = 1.4 x 10^-5), indicating that the normative twin captures cross-site-transferable structure rather than a scanner- or pipeline-specific artefact. Our findings suggest that cardiac phenotyping can move beyond predefined measurements toward data-driven dynamic phenotypes that better capture the heterogeneity underlying cardiovascular disease risk, even among individuals whose ejection fraction appears normal.

Background and purpose: Accurate identification of cardiac MRI volume orientations is essential for reliable image interpretation and for enabling downstream automated analysis pipelines. However, orientation labels and commonly assigned manually, making the process time-consuming and prone to variability. Recent advances in vision-language models offers new opportunities for automated label generation by jointly modeling visual and semantic information. Methods: In this study, we systematically compare a conventional end-to-end transformer-based captioning model with a CLIPCap architecture for automated cardiac MRI view labelling. Both models were evaluated on a dataset of 1184 de-identified cardiac MR images spanning across six clinically relevant orientations: short axis, two-, three-, four-chamber views, left and right ventricular outflow tract views. The caption transformer was trained from scratch, whereas CLIPCap leveraged pretrained CLIP image embeddings combined with lightweight mapping network and a frozen language model. Results: Experimental results demonstrate a substantial performance advantage from CLIPCap, which achieved an overall accuracy of 98% compared to 68% for the caption transformer. CLIPCap consistently delivered high precision and recall across all orientation classes, including those of high clinical importance, while the end to end transformer exhibited unstable performance and complete failure in certain views. Conclusion: These findings highlight the benefits of leveraging large scale multimodal pretrained model for medical image labeling tasks, particularly in data limited settings. The study suggests that CLIPCap provides a more robust and clinically reliable solution for automated cardiac MRI orientation labeling, supporting its integration into efficient and scalable clinical workflows.

The biomechanical properties of the human aorta play a key role in cardiovascular function and disease, yet their population-wide trends and variations remain largely unknown. To fill this crucial gap, we quantified variations in aortic size, stiffness and stresses across 57,434 individuals from the UK Biobank using two-dimensional cine MR imaging and a combined deep-learning, probabilistic and biomechanical modelling framework. Aortic diameter and distensibility (inverse of stiffness) were calculated for the ascending and descending aorta, and their conditional distributions were characterised as a function of age, sex, body size and blood pressure. These observations were integrated with ex-vivo tissue mechanics to infer stress distributions across the aortic wall and the associated uncertainty. Our analysis revealed several new relationships. Larger aortas were found to be stiffer, and greater body size (height and/or weight) was associated with larger aortas with no change in stiffness. Aortic distensibility was negatively associated with systolic, but not diastolic, blood pressure. Apparent sex differences in aortic size were substantially reduced after accounting for body size. Importantly, the maximum stress was found to depend not only on aortic size, as previously thought, but also on distensibility, with lower distensibility associated with higher stresses. The results provide first-of-their-kind population-scale reference distributions for human aortic stiffness, which will improve the personalisation of computational models and digital twins. The correlation between stress and distensibility strongly supports progressing beyond the current diameter-based criteria (based on the law of Laplace) and for the inclusion of stiffness in the clinical risk assessment of aortic diseases. Overall, this study forms the foundation for personalised quantitative medicine and biomechanically informed risk stratification of aortic aneurysms and dissections.

Background. Automated segmentation of glioma subregions on multiparametric MRI can be improved by changing either the network architecture or the information presented to the network. These two design levers are usually studied together, which makes their individual contributions difficult to see. We present an integrated analysis of two completed, independent studies, each of which isolates one lever. Methods. In the architecture-level study (Slice), SIBA-UNet - a 2D U-Net combining depthwise separable convolution, inverted-bottleneck blocks and attention-guided skip fusion - was trained on 128 x 128 multimodal slices from BraTS 2020 with a patient-level split, and a sigmoid decision threshold was selected on validation data. In the representation-level study (Subtract), CMSR-3D augmented the four standard MRI sequences with four voxel-wise cross-modal subtraction maps (T1-T1CE, T2-FLAIR, FLAIR-T1CE, T2-T1CE) and evaluated 4-, 6- and 8-channel inputs with an identical 3D U-Net on 1,251 BraTS-GLI 2023 cases (1,001/125/125 split). Results. SIBA-UNet reached an overall validation Dice of 0.8601 at a validation-selected threshold of 0.51, with reported subregion Dice of 0.8795 (necrotic core), 0.8501 (edema) and 0.9125 (enhancing tumor); the source ablation was only partially reported. With the architecture held fixed, the 8-channel CMSR-3D configuration gave the highest mean Dice among the evaluated inputs (0.8699 vs. 0.8534 for the 4-channel baseline) and the lowest mean HD95 (5.9859 vs. 8.2308), whereas either two-map subset alone changed mean Dice by at most 0.0037. Conclusions. Architecture design and input-representation design are separable levers for brain tumor segmentation. The controlled representation ablation indicates that the complete set of subtraction maps was more informative than either subset under a fixed 3D U-Net. Both studies were evaluated internally on their respective datasets, and their numerical results are not directly compared because the datasets, dimensionality and evaluation protocols differ.

Three-dimensional medical image segmentation is unusually heterogeneous in its computational needs: large homogeneous organ interiors are often resolved at coarse scale, whereas small organs, lesions, thin vessels, and ambiguous boundaries benefit from native-resolution processing. Yet modern 3D segmentation networks allocate nearly uniform computation across the entire volume. Existing dynamic-resolution or coarse-to fine methods reduce cost, but typically route regions using hand-crafted image complexity, prediction confidence, or a fixed crop, none of which directly estimates whether spending additional computation on a region will improve the segmentation. We introduce ERRORROUTE, a budgeted inference framework that learns the marginal refinement utility of candidate 3D blocks. A coarse branch first predicts global low-resolution semantics. During training, we estimate for each candidate block the loss reduction achievable by native resolution refinement and supervise a utility estimator to predict this gain before refinement. A differentiable budgeted selector then allocates a fixed compute budget to blocks with the highest expected utility per unit cost, while unselected regions remain on the efficient global path. Selected blocks are refined locally and merged through residual cross-scale fusion. Across six 3D CT/MRI benchmarks, ERRORROUTE achieves a mean Dice of 90.5% at 128 GFLOPs, improving the accuracy-compute frontier over dense 3D baselines and fixed coarse-to-fine systems. At a 30% routing budget, the learned utility estimator recovers 83% of the blocks selected by an oracle router, with the largest allocation concentrated on lesions, small structures, and complex boundaries. These results suggest that efficient medical segmentation benefits from learning where additional computation is useful, rather than merely where the model is uncertain.

Estimating population-level dependence between brain regions is challenging when the number of regions greatly exceeds the number of individuals. We introduce an information-borrowing generalized Bayesian method for high-dimensional covariance estimation from region-level brain measurements. A divide-and-conquer parameterization separates marginal variances from correlation calibrations: an empirical \emph{pilot} correlation matrix--by default, obtained by standardizing a sample covariance matrix (SCM)--provides a reference structure that is adaptively calibrated through region-specific parameters shared across pairwise correlations, reducing the number of free parameters from $\mathcal{O}(d^2)$ to $\mathcal{O}(d)$ ($d$ is the number of regions) while guaranteeing a positive-definite covariance through the inclusion of a ridge term. Horseshoe priors on the calibration parameters enable selective borrowing and shrinkage of pilot information. Under suitable conditions on pilot accuracy, we establish theoretical guarantees for covariance estimation while retaining scalability in ultra-high-dimensional settings via a low-rank likelihood representation and a Woodbury-accelerated Hamiltonian Monte Carlo algorithm. Simulations showed lower estimation error than the SCM, bagging and iteratively refined principal orthogonal complement thresholding (POETRY). Applications to a semi-synthetic cortical-thickness and ADHD resting-state fMRI data revealed significant group differences in the default-mode--visual network and somatomotor network--limbic network blocks after Benjamini--Hochberg correction.

Submillimeter functional magnetic resonance imaging (fMRI) based on blood-oxygenation-level-dependent (BOLD) signal enables the study of brain function at the submillimeter level, uncovering insights into fine-scale organisations like cortical layers and columns. However, its inherently low contrast-to-noise ratio (CNR) and signal-to-noise ratio (SNR) often limit its reliability and applicability. Noise Reduction with Distribution Corrected Principal Components Analysis (NORDIC PCA) is a locally low-rank denoising algorithm that reduces thermal noise levels in BOLD fMRI in a local patch manner. However, local patches often contain a mixture of signals from multiple tissues that negatively affects the low-rank structure of the patches, which limits the denoising capabilities of the algorithm. We propose an alternative approach for patch formation by gathering similar non-local voxels, dubbed voxel-matching (VM) NORDIC. The results on submillimeter-resolution BOLD fMRI data indicate that VM-NORDIC effectively promotes the low rank of the patches by boosting signal redundancy, allowing for more efficient noise attenuation. Moreover, the method barely affects spatial smoothness due to the non-local voxel selection based on time-series similarity. In particular, VM-NORDIC outperforms standard NORDIC with default local patching (Standard-NORDIC) in terms of temporal SNR (tSNR) (~9-90% larger than Standard-NORDIC; ~23-250% larger than the original) and spatial smoothness estimates (~20% of the smoothness induced by Standard-NORDIC). These improvements are fundamental to improving the validity and precision of fMRI studies at submillimeter resolutions.

Patients with implantable cardioverter-defibrillators (ICDs) often are excluded from receiving CMR scans at 3.0T due to ICD-induced image artifacts. This work aims to transfer sequence protocols established for CMR imaging in the presence of ICDs at 1.5T to a first human application at 3.0T. All measurements were performed on a clinical 3T MRI scanner. Phantom measurements were used to establish and optimize wideband PSIR LGE imaging at 3.0T. Volunteer measurements with externally placed ICDs were used for optimization of CINE imaging and wideband PSIR imaging. Thirdly, patients with implanted ICD were examined with the optimized protocols. Human measurements showed that SA stacks obtained with spoiled GRE after contrast agent showed a markedly lower artifact burden than the conventional approach. For scar imaging, a wideband PSIR with a bandwidth of 3kHz and a frequency offset of +750Hz allowed a sufficient suppression of relevant artifacts in the majority of scans. No strong correlation between the image quality and the distance between the cardiac silhouette and the device was detected. The presented work shows the feasibility of clinical cardiac MRI at 3.0T including GRE-CINE and wideband PSIR-LGE for patients with ICDs.

The choroid plexus (ChP) is a highly vascularized structure located in the brain ventricles that plays important roles in multiple central nervous system functions. ChP enlargement has been implicated in many neuropsychiatric conditions such as multiple sclerosis, depression, schizophrenia, and Alzheimer's disease. However, limitation in the accuracy of the current segmentation tools highlight the need for more reliable methods to quantify ChP volume in healthy and clinical human populations. The shape and size of the ChP vary drastically from person to person, and currently there are no large ChP datasets which allow comprehensive training of segmentation models. Here we introduce DataPlex, a collection of 700 structural MRI brain images where the lateral ventricle ChP has been manually labeled and checked by expert human raters. DataPlex is balanced and representative across a wide range of scanning parameters, demographics, and clinical profiles, and subject-level information will be available for each labeled image. DataPlex will be open source to support the creation of reliable segmentation tools for the human lateral ventricle ChP.

Foundation models (FMs) have rapidly become dominant in artificial intelligence and are increasingly being adopted in computational pathology. Numerous pathology-specific FMs have been developed and evaluated for a variety of downstream tasks, most commonly using frozen image embeddings with linear probes. While a pathology FM, at least implicitly suggests broad reusability across tasks, the transferability of these models to neurodegenerative disease-related tasks remains largely unexplored. In this study, we evaluated fourteen frozen feature extractors, including general-vision and pathology FMs, across four pathology image classification datasets spanning two specialized neurodegenerative disease immuno-histochemistry (IHC) tasks (Tau neurofibrillary tangle (NFT) and amyloid-{beta} plaque classification) and two cancer-related hematoxylin and eosin (H&E) tasks (the breast tissue BACH dataset and the multi-tissue TIL dataset). The cancer-related H&E datasets represent tasks more closely aligned with the predominant pretraining domain of many current pathology FMs, whereas the neurodegenerative IHC datasets represent a more specialized domain with limited representation in existing FM pretraining cohorts. Frozen embedding linear probes were compared against a conventional supervised ResNet-50 convolutional neural network (CNN) trained directly on image tiles using identical train, validation, and test splits. Across both neurodegenerative IHC datasets, the supervised CNN substantially outperformed all frozen FM linear probes. In contrast, pathology FMs achieved performance comparable to, and in some cases exceeding, the supervised CNN across the two cancer-related H&E datasets, with UNI2-h achieving the highest performance on BACH and several pathology FMs performing on par with the CNN on the larger TIL dataset. Furthermore, a supervised CNN trained using only 1% of the Tau NFT training data (1,938 tiles) still exceeded the performance of the best frozen FM linear probe trained on the complete dataset. Together, these results suggest the transferability of frozen pathology FMs may depend strongly on how well the downstream task is represented by their pretraining domain. These findings demonstrate frozen pathology FMs transfer effectively to the cancer-related H&E tasks evaluated here but may be less effective for specialized neurodegenerative IHC tasks less well represented in current FM pretraining cohorts. Expanding pathology FM pretraining datasets to include a broader range of disease domains and staining modalities may therefore improve transferability to specialized pathology applications. Our findings also highlight the continued importance of conventional supervised learning and motivate future work investigating nonlinear probes and end-to-end FM fine-tuning.

Blind source separation (BSS) recovers latent sources from observed mixtures. In brain imaging, sources occur in two domains, spatial and temporal, so a criterion applied in one leaves the other unconstrained. Independent component analysis, the most used BSS tool, exploits diversity such as non-Gaussianity and sample dependence, but in one domain at a time. Methods that use both domains apply the same criterion to each and weight them against each other. We propose spatio-dynamical decomposition (SDD), scoring maps by non-Gaussianity and timecourses by dynamical independence: how much each source's dynamics depend on the others. We normalize each score by how much it can improve on the data, eliminating the need for a weight. Across simulations and 130 naturalistic fMRI runs, SDD recovers sources distinguished by either maps or dynamics and reproduces both across disjoint episodes, while baselines trade off the two; map and timescale together also better predict stimulus tracking, with the largest gain for SDD. Together, these results show that separation is strongest when spatial and temporal domains each carry their own criterion.

Accurate breast lesion segmentation in ultrasound is difficult because speckle noise, heterogeneous lesion appearance, and poorly defined margins can obscure clinically relevant boundaries. This work presents UGB-SegNet, an uncertainty-guided breast ultrasound segmentation network that combines four components: multi-scale convolutional block attention (CBAM) at three EfficientNet-B0 encoder stages, a learnable feature pyramid network for adaptive multi-scale fusion, an entropy-guided decoder that routes fine-grained encoder information toward uncertain regions, and a boundary-aware composite loss with deep supervision. The model was evaluated on the public BUSI dataset (647 annotated lesion images) using a stratified 454/64/129 train/validation/test split. On the held-out test set, UGB-SegNet achieved 81.30% Dice, 68.49% IoU, 81.71% sensitivity, 98.14% specificity, and 15.63-pixel HD95. Under identical training conditions it outperformed five directly trained baselines: U-Net, Attention U-Net, SegNet, DeepLabV3+, and TransUNet. A five-fold resampling experiment over the development set with the held-out test set fixed yielded a mean Dice of 76.36% +/- 1.66% (95% CI 74.90-77.81). Reported ablations attribute gains of +2.06 Dice points to multi-scale CBAM, +1.52 to learnable FPN fusion, +1.22 to the uncertainty-guided decoder, and +1.24 to boundary-aware training. UGB-SegNet contains 14.00M parameters and runs at 62.4 FPS at 224 x 224 resolution on the tested NVIDIA GPU, providing a compact accuracy-efficiency trade-off for future prospective clinical validation.

Predicting implantation outcomes in vitro fertilization (IVF) is challenging because clinically relevant information becomes available at different points during treatment. In this study, we developed a stage-aware, leakage-conscious, and explainable machine-learning framework to predict implantation using a public IVF dataset. {beta}-HCG was excluded to avoid potential downstream target leakage. The data were divided into training, validation, and independent test sets. Several machine-learning models, including Logistic Regression, Random Forest (RF), Extra Trees, XGBoost, and Hist Gradient Boosting, were optimized using stratified cross-validation. A fixed RF model was then evaluated at three stages of the IVF process, using baseline couple characteristics, gamete and fertilization variables, and embryological variables, respectively. Model performance improved as additional treatment-related information became available, with the largest improvement observed after the inclusion of embryological variables. The final RF model showed good predictive performance on the independent test set. The number of embryos was consistently identified as the most important predictor across RF, permutation, and SHAP analyses. Overall, the findings suggest that IVF implantation prediction becomes more accurate as later-stage treatment information is incorporated, particularly information related to embryological characteristics. The proposed approach also provides interpretable predictions while reducing the risk of information leakage. Further evaluation using independent clinical datasets is needed before the framework can be considered for clinical use.