Journal of Medical Imaging

Objectives To assess the effectiveness of our proposed urinary segmentation method based upon image processing, and to contrast its performance against various approaches. Subjects and Methods Non-contrast-enhanced computed tomography (NCCT) scans of instances with stone disease were collected. An experienced urologist and a senior radiology resident generated ground truths by manually drawing the regions of interest (ROIs) of stone samples. The proposed method took loosely defined ROIs as input and included three steps: (1) finding mean_HU_rim, the mean Hounsfield unit (HU) of the stone rim, on each NCCT slice by an edge detection operator; (2) considering pixels with HU higher than the minimum value of all the mean_HU_rim; and (3) removing noise. The entire data was divided into a developing set of 406 samples to evaluate the performance of the proposed method as well as to fine-tune a nnUNetv2 model and a comparison test set of 125 samples to compare the performance and inference time of various segmentation methods. In addition, the reproducibility of our method was tested on the 30 samples randomly selected from the entire data. Results A total of 531 stone samples from 287 instances were included. Our method showed a high level of agreement with the ground truth in the developing set, with the median (interquartile range) Dice coefficient of 0.86 (0.81 - 0.89). The reproducibility of our method was remarkably high with the median (interquartile range) Dice coefficient of 1.0 (0.991 - 1.0). On the comparison test set, our semi-automatic segmentation outperformed the fixed thresholding and deep-learning-based methods. Lastly, the inference time of the proposed method was less than a second. Conclusion The proposed semi-automatic approach provided a consistently reliable, simple, and robust method to segment urinary stones.

Purpose: To evaluate how segmentation architecture and dataset-adaptive configuration influence uterine MRI segmentation across heterogeneous benign and malignant tasks. Methods: U-Net, Swin-UNETR, and MedNeXt were compared with nnU-Net as a self-configuring reference across T2-weighted MRI datasets: public multiclass UMD anatomy/fibroid segmentation (n=300), institutional endometrial cancer tumor segmentation (n=206), and institutional uterine mass lesion segmentation (n=234). A relabeled external UMD-style cohort (n=12) assessed domain shift. Models used fixed partitions, fold ensembling, Dice, HD95, ASSD, volume error, and paired bootstrap comparisons with Holm correction. Results: MedNeXt was the strongest manually controlled architecture. nnU-Net achieved the highest performance on all internal datasets and external testing. Macro-Dice reached 0.761, 0.746, and 0.814 for nnU-Net on UMD, endometrial cancer, and uterine mass datasets, respectively, versus 0.722, 0.726, and 0.789 for MedNeXt. The nnU-Net-MedNeXt gap was largest for multiclass UMD segmentation and smaller in binary tasks. External testing degraded all models; nnU-Net remained highest (0.542), followed by MedNeXt (0.490), U-Net (0.396), and Swin-UNETR (0.287). Conclusions: Uterine MRI segmentation performance depended on task, architecture, and evaluation domain. MedNeXt supported modern convolutional design as a strong manual baseline, but nnU-Net remained the most robust overall, emphasizing the importance of dataset-adaptive configuration and external validation.

Convolutional neural networks (CNNs) have shown strong capabilities for image analysis. However, deploying these models in medical settings is complicated by their limited transparency. Over recent years, many approaches have been developed to overcome the so-called "black box" problem of deep neural networks. Here, we show how such explainable AI (XAI) approaches can be applied to not only improve transparency but also training data quality and model performance, with classification of lung damage in histopathology images as use case. First, we conducted a thorough exploratory data analysis and visually compared the compressed multi-dimensional representations of the histology images from the last CNN layers with labels given by pathologists to reveal flaws in the training data. Second, we used Gradient-based Class Activation Mapping (Grad-CAM) as well as SHapley Additive exPlanations (SHAP) values to identify image regions that significantly contributed to the model decisions. To overcome identified shortcomings, we then finetuned additional top layers of the CNNs and introduced model architectures with attention which improved model performance. In summary, we developed a practical workflow that uses interpretability analyses to examine model perception, assess label consistency, and guide model refinement on the example of lung histopathology scoring, demonstrating how XAI approaches can increase both transparency and model performance. The code for this paper is shared at https://github.com/Aitslab/Histology_XAI.git.

Background: Point-of-care lung ultrasound (LUS) interpretation requires specialized training. Vision-language models (VLMs) may automate multi-sign assessment but lack systematic clinical benchmarking. We benchmarked three LUS tasks on 125 evaluable cases drawn from 150 point-of-care ultrasound (POCUS) Atlas clips to establish baseline capabilities and isolate clinically relevant failure modes for VLM-based LUS interpretation. Methods: We evaluated a normalized multi-model benchmark on three tasks: pleural sliding detection (T1), lung-rocket (B-line) and consolidation classification (T2), and Posterolateral Alveolar and/or Pleural Syndrome (PLAPS; comprising posterolateral consolidation and pleural effusion) assessment (T3). Inputs were 10 uniformly sampled frames for all tasks, plus synthetic M-mode images for sliding. Results: We report four findings. First, M-mode improved T1 sliding over static frames for some models, but performance remained modest with wide uncertainty. Second, T2 pathology, identifiable from single frames, was discriminated most clearly above chance on a balanced cohort, with strong lung-rocket and anterior-consolidation F1 achieved by open-weight rather than closed-weight Claude models, although a significant between-model difference held only for lung rockets. Third, T3 PLAPS F1 was high for most models but reflected high positive-class prevalence rather than strong discrimination. Fourth, inter-model agreement was near chance (task-averaged Cohen's kappa = 0.02-0.29) and well below each model's self-consistency, indicating distinct, non-redundant error patterns. Conclusions: Static-frame VLMs produced discriminative labels for B-lines and anterior consolidation but remained unreliable for motion-dependent signs (pleural sliding) and prevalence-inflated outcomes (PLAPS). Synthetic M-mode partially recovered temporal information but did not achieve clinically sufficient sliding accuracy. These results support morphology-assisted use for B-lines and consolidation; motion-dependent signs and balanced endpoints require larger validation cohorts before clinical deployment

The accurate segmentation of lizard claws is important as they are materially heterogeneous, comprising both bone and keratinous tissue. This study presents a deep learning framework for the automated segmentation of lizard claw tissues, specifically bone and keratin, from CT imaging data. A dataset comprising 14 lizard claws was used in this work, with annotations generated through a superpixel based labeling approach to provide ground truth reference segmentations. To evaluate the effect of spatial context on segmentation performance, both 2D and 2.5D CNN architectures using DeepLabV3 with ResNet-50, ResNet-101, and Inception-ResNet-v2 backbones were investigated, with predictions subsequently reconstructed into three-dimensional volumes for analysis. Performance was assessed using a leave one out cross validation (LOOCV) strategy and evaluated with 3D Dice Similarity Coefficient (DSC), Intersection over Union (IoU), Sensitivity (Recall), 95th Percentile Hausdorff Distance (HD95), and Relative Volume Error (RVE). Experimental results demonstrate that 2.5D CNN architectures consistently outperform their 2D counterparts across all evaluation metrics, highlighting the importance of incorporating inter-slice contextual information for volumetric tissue segmentation. From amongst the models, the 2.5D Inception-ResNet-v2 achieved the best overall performance, reaching a validation accuracy of 97.5% and producing segmentation results that closely align with ground-truth tissue structures. Our findings demonstrate the effectiveness of 2.5D deep learning approaches for the high accuracy segmentation of heterogeneous lizard claw tissues from CT data, whilst providing a robust framework for automated morphological analysis in comparative anatomical studies.

Automated lesion segmentation is increasingly central to PSMA PET/CT interpretation, supporting staging, treatment planning, and response assessment at a scale that outpaces available nuclear-medicine expertise. However, automated PSMA-PET/CT whole-body lesion segmentation models are trained on images alone, with no knowledge of where in the body prostate metastases actually tend to occur. Radiologists use clinical domain knowledge of metastatic spread, but its absence in machine learning models produces false positives in anatomically implausible locations and missed lesions in high-risk sites such as the liver. In this work, we explore whether population-level spatial knowledge of metastatic spread can be used to augment deep learning segmentation predictions, and how such a prior should be fused with a network's output, without additional training. We build a data-driven metastasis atlas from 375 expert-annotated whole-body PSMA PET/CT scans and investigate its fusion with a trained segmentation network under a Bayesian framework, in which prediction probabilities from an nnU-Net-based lesion segmentation model serve as the likelihood and the data-driven atlas as the prior. Because metastases occupy only a small fraction of whole-body voxels, the atlas's peak probability is too low, and standard power-scaled or naive Bayesian pooling references lack the tools to deal with this shortcoming. This causes these standard fusion strategies to fail and, in the naive Bayesian case, to sharply degrade performance. We instead derive a calibrated, background-referenced log-odds fusion, one of many possible approaches to combine a population atlas with a deep learning model's predictions, distinct from classical multi-atlas label fusion in that it fuses a single population prior with a trained network's softmax rather than combining several registered atlases. Furthermore, this approach is neutral outside atlas support by construction, reduces exactly to the baseline network when unweighted, and requires no retraining. This atlas fusion significantly improved mean Dice over the baseline nnU-Net on a disjoint internal test set ($+0.011$, Holm-adjusted $p=0.021$) and on an independent external cohort ($+0.0129$, Holm-adjusted $p=3.8\times10^{-16}$), with lesion sensitivity improving from 0.849 to 0.861 internally and Dice improving over baseline in every stratified anatomic region, including the rare, high-risk sites motivating this work, while naive Bayesian pooling degrades performance sharply and power-scaled pooling underperforms it throughout. Our findings suggest that population-level spatial priors can meaningfully augment deep learning predictions in whole-body oncologic segmentation, provided the fusion rule is calibrated to where the prior actually carries signal.

Multi-phase contrast-enhanced computed tomography (CT) is the gold standard for renal cell carcinoma (RCC) characterization, yet clinical interpretation relies on subjective visual comparison across phases. We present PRISM (Phase-Resolved Isotropic Subtraction Mapping), an open-source automated pipeline that transforms multi-phase CT acquisitions into registered digital subtraction angiography (DSA) volumes with color-coded enhancement maps. PRISM integrates six sequential processing stages: (1) DICOM loading with automated contrast-phase classification, (2) deep learning-based isotropic interpolation via RIFE, (3) automated kidney segmentation using TotalSegmentator v2, (4) enhancement-based tissue detection, (5) three-step deformable registration (rigid, affine, B-spline) using SimpleITK, and (6) dual-channel digital subtraction visualization. We present a systematic parameter optimization study comprising 200 registrations across five patients and five experiments. Key findings: We identify an efficient registration configuration combining a 40 mm B-spline grid (within 6% of the 30 mm quality optimum at 36% lower computational cost), 5% metric sampling (equivalent quality to 25% at 3.2x speedup), and a single-level multi-resolution pyramid (avoiding the 5.4x overhead of a 4,2 pyramid with no quality benefit); we show that registration quality is effectively independent of interpolation target spacing from 0.5-3.0 mm, enabling a coarse-register/fine-apply strategy that computes the full transform at 3.0 mm (approximately 4 minutes per phase) and applies it to 0.5 mm volumes for high-resolution visualization. We also determine that a 40 HU subtraction noise threshold optimally balances signal-to-noise ratio (2.00) against sensitivity (14.2% enhancing volume retained), with higher thresholds (60-80 HU) favoring specificity and lower thresholds (20 HU) favoring sensitivity.

Background. Deep denoisers restore low-dose chest radiographs but emit a single point estimate with no indication of where the output is reliable, so a confident-looking reconstruction can be locally wrong. In paediatric radiography, where the radiation-dose imperative is sharpest, we ask whether a per-pixel uncertainty map can be attached at no meaningful cost to reconstruction quality. Methods. We develop PP-VAE-Hformer, a hybrid convolution-transformer denoiser with a variational-autoencoder bottleneck (epistemic uncertainty map) and a heteroscedastic dual head (aleatoric map), and run a 19-arm ablation isolating five composite-loss terms on Poisson-Gaussian-degraded paediatric chest radiographs (Kermany collection; 624-image held-out test set), against eight discriminative baselines retrained on identical data and noise. Calibration is assessed by reliability diagrams and post-hoc-scaling; comparisons use Bonferroni-corrected Welch tests with Cohens d as the primary discriminant. Results. A heteroscedastic negative-log-likelihood (NLL) objective buys a calibrated aleatoric map for 1.2 dB PSNR, which two-stage fine-tuning recovers to a statistically indistinguishable 0.010 dB. For the variational variant the Kullback-Leibler (KL) annealing schedule, not the bottleneck itself, is decisive: cyclic annealing closes a 1.10 dB gap over posterior-collapsing linear warmup. Predicted uncertainty tracks realised error (Pearson r > 0.8) with an optimal recalibration scale within 1% of unity. A matched-severity cross-degradation test shows this calibration is model-circular, however: at equal severity the aleatoric map becomes modestly (8-16%, a lower bound) over-confident once the noise model changes, and the Monte-Carlo epistemic map is a weaker error predictor that does not compensate. We additionally document a previously unreported destructive interaction between structural-similarity and Sobel-edge supervision at literature-default weights. Conclusions. A calibrated per-pixel uncertainty map can be attached to a paediatric low-dose CXR denoiser at effectively no reconstruction cost. Its near-perfect calibration is tied to the trained (inverse-crime) noise model, and the Monte-Carlo epistemic map does not reliably substitute where the aleatoric map fails, so an architecture-independent uncertainty estimate (e.g. deep ensembling), a blinded reader study, and real low-dose data are the immediate next steps. Keywords. image denoising; uncertainty quantification; calibration; variational autoen-coder; paediatric chest radiography; low-dose imaging

Prostate cancer, a leading cause of cancer-related deaths among men globally, necessitates the development of precise diagnostic and treatment strategies. Accurate segmentation of prostate cancer in medical imaging, particularly in MRI scans, is crucial for early diagnosis and clinical decisions. Traditional manual segmentation techniques, while efficient, are labor-intensive and necessitate significant expertise, resulting in an increasing demand for automated alternatives. The RSAUNet is a novel deep learning architecture designed to improve prostate cancer segmentation. It incorporates essential components, including Residual Blocks, Swin Transformer Blocks, and Attention Mechanisms within a U-Net architecture. This markedly enhances the model's capacity to discern complex anatomical features and accurately segment malignant tissues. Using sophisticated deep Learning methods, RSAUNet addresses the complexities of prostate imaging, delivering reliable, consistent segmentation results. The model was evaluated against various cutting-edge techniques on extensive multi-parametric MRI datasets, attaining an impressive Dice coefficient (DC) of 0.998 and a Jaccard Index (IoU) of 0.965. These findings highlight the innovative characteristics of RSAUNet and its capacity to transform prostate cancer diagnosis and treatment strategies. The proposed model surpasses current methods and shows potential for practical clinical applications, providing an efficient, precise, and scalable solution for automated prostate cancer segmentation and fostering optimism for the future of medical imaging and diagnosis.

Diffusion models achieve high reconstruction quality in low-dose computed tomography (LDCT), but their iterative sampling trajectories impose substantial computational costs. Unlike unconditional generation, paired LDCT reconstruction starts from an image that already contains the anatomy and spatial structure of the standard-dose CT (SDCT) target; reconstruction primarily requires correcting dose-related noise and artifacts. We therefore introduce Residual Endpoint Flow Matching (REFM), an LDCT reconstruction method that learns to transport an LDCT image directly toward its paired SDCT endpoint rather than defining a noise-to-image trajectory. REFM predicts the residual velocity along linear interpolations between both images and supports single-step and multi-step reconstruction using the same trained network. We evaluate five model capacities using 1 to 50 Euler steps against deterministic U-Net and diffusion-based baselines. Across all REFM variants, one-step inference consistently provides the highest reconstruction quality. On the TCIA validation set, REFM Base achieves 50.98 dB PSNR and 0.9865 SSIM at 94.54 fps, compared with 50.92 dB, 0.9847, and 9.26 fps for DDPM-10. REFM Small retains 50.71 dB while increasing throughput to 198.56 fps. Without fine-tuning, REFM Base also matches the 25-step DDPM baseline on the external Mayo Clinic dataset, although DDPM remains stronger on synthetically degraded CRLM images. Thus, our results show that exploiting paired anatomical correspondence enables diffusion-level LDCT reconstruction with a single step reconstruction.

Background and Objectives: Automatic segmentation of the nasal cavity and paranasal sinuses from CT aids diagnosis and surgical planning, but clinical datasets in this domain remain small, which can affect training stability and evaluation validity. This study investigates whether the frequency-domain mechanisms of the Adaptive Frequency-Spatial Dual-Stream Network (AFS-DSN) can be transferred, in a lightweight form, into the self-configuring nnU-Net framework, and whether the resulting performance improvement survives rigorous statistical validation. We integrated the AFS-DSN frequency branch and cross-domain attention mechanism into nnU-Net, and corrected a zero-initialization gating deadlock in the original design, using the corrected model (v2) as a common base. On this base, we evaluated three variants: X1, an input-conditioned learnable spectral gate applied to 24 wavelet sub-bands; X2, which adds a boundary-distance training objective; and X3, combining X1 and X2. Experiments used the 130-volume NasalSeg CT dataset, under a full 91/19/20 protocol and a rebuilt, genuine 3-fold cross-validation in which the training set for each fold was reduced by approximately 19%, from 91 to 73-74 cases. Under the full protocol, all variants showed a small but statistically detectable improvement over a matched nnU-Net baseline. X3 achieved a Dice score of 95.78% versus 95.59% for the baseline, an improvement of 0.19 percentage points (95% CI [0.09, 0.29]; Holm-adjusted p = 0.004), and X3 also reduced average surface distance (ASD) from 0.252 mm to 0.235 mm. X3 was statistically equivalent to either individual component within a margin of +/-0.10 percentage points, suggesting a shared performance ceiling rather than a complementary gain. Under reduced training data, baseline Dice dropped by 2.71 percentage points (95.59% to 92.88%), while X3 remained essentially unchanged (95.78% to 95.75%), a cross-fold advantage of +2.88 percentage points (fold-level 95% CI [0.80, 4.96]) that held across all 20 test cases (sign test, p = 1.9e-6) and was corroborated by average surface distance. The practical value of this combined approach lies primarily in within-distribution data efficiency, rather than in maximizing peak segmentation accuracy.

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.

Purpose: To evaluate whether visually assessed high-quality prostate MRI, as defined by the Prostate Imaging Quality (PI-QUAL) scoring system, is sufficient to ensure consistent AI model performance, or whether hidden acquisition-related technical effects introduce variability that cannot be detected through visual inspection. Methods: This retrospective study included 568 bpMRI scans from seven institutions, classified using PI-QUAL v2 into high-quality (scores 2-3) and low-quality (score 1) groups. Quantitative image quality features were extracted using MRQy from T2w and ADC sequences. Three experiments were conducted on high-quality scans: (1) unsupervised clustering to identify hidden technical batch effects; (2) V-Net models (M) for prostate capsule segmentation on T2w using single-cluster and balanced multi-cluster training strategies; and (3) VoxelMorph-based deformable registration (M2) to align ADC maps to T2w images. Performance metrics included DSC, precision, recall, volume ratio, HD95, ASSD, and TRE. Kruskal-Wallis tests assessed cross-cluster variability, Two One-Sided Tests (TOST) evaluated performance equivalence, and bootstrap resampling quantified training efficiency. Results: Clustering identified four clusters among 277 high-quality bpMRI scans. Single-cluster models showed variable performance (DSC=0.41-0.85) with significant batch effects (p<0.001), even with 140 training samples. The balanced multi-cluster model (n=60) eliminated batch effects (p=0.075), achieved equivalent performance (p<0.001), and demonstrated 2.23-fold greater training efficiency despite using 2.3-fold fewer samples. Registration performance also varied significantly across clusters (TRE=0.86-5.47 mm, p<0.001). Conclusion: Acquisition-related batch effects significantly impact AI generalizability in prostate MRI, even among high-quality scans. Quantitative clustering enables detection of technical variations and supports diversity-aware training strategies that achieve equivalent performance with 2.3-fold fewer samples while eliminating batch effects.

Reliable clinical deployment of automated liver segmentation requires mechanisms for detecting failures in rare and previously unseen scenarios. Achieving this goal requires an appropriately calibrated threshold that converts an out-of-distribution (OOD) score into a failure prediction. However, threshold calibration typically relies on expert-labeled failures, creating a substantial annotation burden when failures are rare. Building upon our prior work, which uses Pairwise Surface DSC scores as indicators of segmentation quality, we propose a label-free framework for calibrating OOD score thresholds. First, we fitted a log-t distribution to Pairwise Surface DSC scores from a validation set of 400 internal scans to approximate an in-distribution score distribution. New segmentations were assigned significance scores based on their extremity under this fitted distribution and categorized into Low, Medium, and High Risk review groups using statistically principled cutoffs of 0.25 and 0.05. The fitted log-t distribution provided a strong fit to the observed scores and remained robust to moderate contamination by OOD cases. On an independent test set of 500 internal and external scans, the combined Medium and High Risk categories achieved 100% sensitivity and 79% specificity, whereas the High Risk category alone achieved 78% sensitivity and 96% specificity. These results indicate that clinically meaningful failure detection can be derived from unlabeled data. Our code is available at https://github.com/marshalln7/Label_Free_OOD_Threshold_Selection.

Faster ultrahigh-resolution (UHR) photon-counting CT acquisitions facilitate breath-hold imaging, but identical reconstruction settings may conceal differences in spatial resolution. To assess spatial resolution and depiction of small airways and venous stents across helical UHR acquisition configurations. Stationary wire, airway, and venous-stent phantoms underwent one acquisition per configuration using five configurations on a dual-source scanner at 120 kVp. Reconstructions used 0.2- and 0.4-mm thicknesses, the same sharp kernel, and iterative reconstruction strength 1. We measured the full-width at half-maximum of the slice sensitivity profile (SSP) and spatial frequency at 20% of the modulation transfer function (MTF20). Rank-based F tests with Holm adjustment assessed configuration differences in resolution and variability across slices. Airway detection used a profile contrast-to-noise criterion; stent assessment was qualitative. Calculated acquisition time for 400-mm coverage decreased from 9.80 to 1.30 seconds between the slowest and fastest configurations. At 0.2-mm reconstruction thickness, median SSP increased from 0.316 to 0.496 mm, and median MTF20 decreased from 2.870 to 2.059 lp/mm (both P<.001). Interquartile ranges increased from 0.047 to 0.197 mm for SSP and from 0.256 to 0.654 lp/mm for MTF20 (both variability P<.001). Increasing reconstruction thickness to 0.4 mm increased SSP by a median 0.221 mm across 2459 matched slice-wire measurements (P<.001). At 0.2-mm reconstruction thickness, 0.4- and 0.6-mm lumens with 0.2-mm walls were detected only in the slowest configuration. Faster configurations showed less distinct stent struts and greater adjacent-wall distortion. Faster helical UHR photon-counting CT configurations had poorer spatial resolution despite identical nominal slice thickness. Consistent acquisition and reconstruction settings are important for longitudinal comparisons.

Purpose: To build a 3D U-Net model, BioDeformUNet, to predict the deformation vector field (DVF) of the liver in near real-time, for efficient intra-procedural evaluation of the minimal ablative margin (MAM). Materials and Methods: This retrospective study included 170 contrast-enhanced computed tomography (CECT) image pairs from 157 patients who underwent liver ablation treatment between 2020-2024. Each data instance included one pre-ablation CECT (pre-CECT) and one post-ablation CECT (post-CECT). BioDeformUNet was trained under the guidance of DVFs generated by a biomechanical model-based deformable image registration (DIR) algorithm using a loss function that focused on large liver deformations. Data were split patient-wise into training (92-93 patients), validation (23-24 patients), and testing sets (42 patients). We compared our performance with two deep learning-based DIR methods: VoxelMorph and VFA. Evaluation metrics included: target registration error (TRE), Dice similarity coefficient (DSC), Minimum Ablation Margin (MAM), and inference time. For BioDeformUNet, we additionally evaluated the accuracy of the deformed tumor center-of-mass mapping by comparing the predicted tumor center location with that generated by Morfeus. A mapping error less than 3.0 mm (corresponding to the voxel size) was considered accurate. We used the Wilcoxon signed-rank test to assess the significancy of each test result. Our code is available at https://github.com/XinyueZhang831/BioDeformUNET. Results: The TRE of BioDeformUNet was not significantly different from Morfeus (3.31 BioDeformUNet; 3.23 Morfeus; p-value=0.41). The BioDeformUNet DVF magnitude was within 3.0 mm of Morfeus DVF for an average of 91.9% of the voxels. Tumor mapping errors greater than 3.0 mm occurred in only 8 cases. The inference time of BioDeformUNet was 0.6s per image pair, 0.2s for VoxelMorph, 0.3s for VFA, and 20.2s for Morfeus. Conclusion: BioDeformUNet achieved a similar performance to the biomechanical model-based algorithm but required fewer computational operations, resulting in a 34 times speedup in DVF computation.

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.

Objectives: To characterize image noise, lesion contrast, spatial resolution, and attenuation distribution for deep learning reconstruction (DLR) compared with hybrid iterative reconstruction (IR) in ultra-high-resolution (UHR) photon-counting CT (PCCT) of the lung, using a patient-derived 3D-printed phantom, and to apply the same analytical framework to patient examinations acquired on the same system. Materials and Methods: A patient CT scan-derived lung phantom containing three digitally inserted lesions of differing morphology was fabricated using PixelPrint 3D printing and imaged on a CZT-based PCCT prototype at four dose levels (CTDIvol 6, 4, 2, and 1 mGy). Images were reconstructed with hybrid iterative reconstruction at normal resolution (512 by 512) and UHR (1024 by 1024), and with DLR at UHR. Image noise, lesion contrast, and contrast-to-noise ratio (CNR) were measured within prespecified regions of interest replicated across reconstructions by coordinate transfer. Spatial resolution was assessed by the modulation transfer function (MTF) from a high-contrast edge at eight locations. Attenuation distribution and local texture were characterized using histograms and gray-level co-occurrence matrices (GLCMs) within a structured parenchymal region. Two patient examinations with part solid lung nodules acquired on the same prototype were analyzed as illustrative examples, with solid and ground-glass components segmented separately and their attenuation distributions compared. Results: In the phantom, DLR reduced image noise relative to UHR hybrid iterative reconstruction (7.8% at 6 mGy and 15.2% at 1 mGy). Lesion contrast was constant across reconstructions and dose levels, so CNR differences reflected differences in noise alone. MTF curves for DLR lay above those for hybrid iterative reconstruction across the evaluated spatial frequency range at both 6 and 1 mGy, averaged +0.4 lp/mm in MTF50. In a structured parenchymal region, the attenuation histogram mode became progressively narrower from hybrid iterative reconstruction at 1 mGy to DLR at 6 mGy (averaged -23%), while mode position and mean regional attenuation were preserved, consistent with reduced partial-volume mixing rather than altered attenuation. In the patient example, the attenuation distributions of solid and ground-glass components were more widely separated with UHR DLR (2.0) than with either hybrid iterative reconstruction (1.6). Conclusions: In UHR photon-counting CT of the lung, DLR reduced image noise while preserving lesion contrast and regional attenuation, increased measured MTF at a high-contrast edge, and produced attenuation distributions consistent with reduced partial-volume mixing at material boundaries. The same effect was observed in the separation of solid and ground-glass components in two patient examples.

Foundation models for chest X-ray interpretation make it possible to adapt specialised visual representations with relatively small trainable modules. We report a retrospective study of Low-Rank Adaptation (LoRA) of Rad-DINO Vision Transformer Base with 14x14 patches (ViT-B/14) for 14-class multi-label classification on the National Institutes of Health (NIH) ChestX-ray14 dataset. The official test labels were accessed during earlier model development and configuration comparisons; consequently, every official-test result in this manuscript is explicitly descriptive and non-confirmatory. We used a patient-disjoint 90/10 split of the official trainval pool (77,988 training and 8,536 validation images) and retained the released 25,596-image test partition. The historically selected all-linear LoRA configuration with safe augmentation and g=37 produced a descriptive test macro AUROC of 0.8462 versus the frozen baseline of 0.8295. Comparisons of target modules, patch-token grids, and a Rad-DINO-specific local query head are reported as retrospective comparisons rather than unbiased model-selection evidence. A confident-learning diagnostic flagged 17,653 of 86,524 trainval images (20.4%); this is a model-based flag rate, not a ground-truth label-error rate. A separate counterfactual relabeling sensitivity analysis, which uses the same model to identify and rescore disagreements, changed the descriptive AUROC to approximately 0.9445 after 6,509 policy-defined flips. This value is not achieved model performance and is not a radiologist-audited label-quality ceiling. We provide a validation-only threshold and artifact protocol for future locked evaluation, but a genuinely untouched holdout and new locked selection are required for a confirmatory headline. The existing Zenodo record contains the 25 publication figures only.

Objective: To identify common photon-counting CT (PCCT) virtual monochromatic imaging (VMI) settings for accurate hepatic fat fraction (FF) quantification across different body sizes, including large body habitus. Methods: Six non-iodinated fat lesions (FF 5%-40%) were embedded in anthropomorphic liver phantoms representing medium-sized (25x32.5 cm^2) and large (31x39 cm^2) abdomens. Phantoms were scanned on a PCCT system (NAEOTOM Alpha) at 120 and 140 kV. CT numbers were measured in VMIs at 40-190 keV in 1-keV increments. Linear regression between the ground-truth FF and measured Hounsfield units (HU) was used to estimate FF. Common optimal VMI settings yielding the minimum relative root-mean-square error (rRMSE) in both phantoms were identified. Results: A single VMI setting of 70 keV at 140 kV demonstrated the best overall performance across body sizes, with FF (%) = -0.689HU + 36.51 (R^2 > 0.996), achieving rRMSE [&le;]3.4% and absolute RMSE [&le;]0.7% in both phantoms. Robust performance (rRMSE [&le;] 5%) was consistently maintained across 69-71 keV using identical calibration parameters for both phantoms. These results represented a substantial improvement over previously reported dual-energy CT (DECT) performance, while enabling accurate quantification on PCCT at radiation doses approximately 40% lower than those used in prior DECT protocols. Conclusion: PCCT enables accurate and robust hepatic fat fraction quantification independent of body size. A single protocol at 140 kV with VMIs of 69-71 keV consistently achieved low quantification errors, demonstrating strong potential for opportunistic liver fat assessment using PCCT, especially in obese patients.