Purpose:To develop and validate a combined model for predicting gout risk by integrating ultrasound (US) features as novel risk factors with clinical data and predictions from deep learning (DL) models. Patients and Methods:This retrospective study included 609 cases who underwent first metatarsophalangeal (MTP1) joint US at two centers. Data from Center 1 were divided into a training group (70%, n = 355) and an internal testing cohort (ITC) (30%, n = 162). Data from Center 2 served as an external testing cohort (ETC) (n = 92). A DL diagnostic model based on MTP1 US images was developed to obtain diagnostic predictions. Clinical data, US features, and DL predictions were integrated, and logistic regression analysis was performed to identify independent risk factors. Various models were constructed (clinical, US, clinical-US, clinical-DL, and combined), and the best model was interpreted with a nomogram. Multicollinearity was assessed using the variance inflation factor. Model performance was evaluated using the receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). Results:The combined model, incorporating clinical data (gender, serum uric acid [SUA]), US features (tophus, double contour sign (DCs), bone erosion), and DL predictions, exhibited the best performance. For the ITC, the area under the curve (AUC) and Brier scores were 0.904 (95% CI: 0.843~0.965) and 0.100 (0.066~0.140), respectively. For the ETC, they were 0.881 (95% CI: 0.815~0.947) and 0.160 (0.107~0.221). DCA confirmed the clinical utility of the combined nomogram. Conclusion:A nomogram was constructed based on seven risk predictors (gender, SUA, estimated glomerular filtration rate (eGFR), tophus, bone erosion, DCs, and DL prediction) to predict and quantify gout risk.

Background:Although thyroid nodules are detected in up to 60% of adults on ultrasound, the vast majority are benign, creating a substantial decision-making burden compounded by heterogeneous practice guidelines. Large language models (LLMs) show promise in processing unstructured medical text and are emerging as tools for report interpretation among both clinicians and patients. However, their reliability across distinct clinical tasks in thyroid ultrasound interpretation remains poorly characterized. Objective:This study evaluates 2 LLMs, ChatGPT-4o and DeepSeek-R1, in interpreting thyroid nodule ultrasound text reports across three clinical tasks-benign-malignant differentiation, Chinese Thyroid Imaging Reporting and Data System (C-TIRADS) classification, and management recommendation-with concurrent assessment of output stability for each task. Methods:We retrospectively analyzed 1063 ultrasound text reports from 3 medical centers, including 306 with histopathological confirmation. Each nodule report was submitted to both LLMs via their consumer web interfaces using task-specific prompts, with 5 repetitions per model; final outputs were determined by mode voting. Diagnostic performance was assessed by receiver operating characteristic analysis with DeLong testing; agreement was quantified using squared weighted κ and Cohen κ; and stability was measured using Krippendorff α and Fleiss κ. Results:For benign-malignant differentiation, DeepSeek-R1 showed higher sensitivity (0.879 vs 0.692; P<.001) and accuracy (0.729 vs 0.644; P=.008) than ChatGPT-4o. With access to images and clinical context unavailable to the LLMs, senior radiologists showed higher performance (area under the curve=0.865; accuracy=0.804). For C-TIRADS classification, DeepSeek-R1 showed substantial agreement with radiologists, exceeding ChatGPT-4o (κ=0.770 vs 0.688; Δκ=0.082, 95% CI 0.048-0.122). Both models yielded moderate, comparable agreement with clinicians on management recommendations (κ=0.606 vs 0.608). Stability was near perfect for C-TIRADS classification (α=0.864 vs 0.866) and management recommendations (κ=0.853 vs. 0.849) in both models; however, DeepSeek-R1 showed markedly greater stability than ChatGPT-4o in benign-malignant differentiation (κ=0.869 vs 0.609; Δκ=0.260, 95% CI 0.191-0.321). Conclusions:Both LLMs demonstrate clinical potential for thyroid nodule ultrasound report interpretation, with DeepSeek-R1 showing advantages in diagnostic accuracy, classification consistency, and output stability. However, both LLMs remained inferior to senior radiologists, suggesting their role as decision-support tools rather than stand-alone diagnostic systems. These findings provide preliminary evidence to inform the responsible integration of LLMs into thyroid imaging workflows while highlighting the need for further evaluation before patient-facing deployment.

Aim: This study aimed to develop a deep learning (DL) model for automatic detection and diagnosis of gouty arthritis (GA) in the first metatarsophalangeal joint (MTPJ) using ultrasound (US) images.Materials and methods: A retrospective study included individuals who underwent first MTPJ ultrasonography between February and July 2023. A five-fold cross-validation method (training set = 4:1) was employed. A deep residual convolutional neural network (CNN) was trained, and Gradient-weighted Class Activation Mapping (Grad-CAM) was used for visualization. Different ResNet18 models with varying residual blocks (2, 3, 4, 6) were compared to select the optimal model for image classification. Diagnostic decisions were based on a threshold proportion of abnormal images, determined from the training set.Results: A total of 2401 US images from 260 patients (149 gout, 111 control) were analyzed. The model with 3 residual blocks performed best, achieving an AUC of 0.904 (95% CI: 0.887~0.927). Visualization results aligned with radiologist opinions in 2000 images. The diagnostic model attained an accuracy of 91.1% (95% CI: 90.4%~91.8%) on the testing set, with a diagnostic threshold of 0.328.Conclusion: The DL model demonstrated excellent performance in automatically detecting and diagnosing GA in the first MTPJ.

Although struma ovarii (SO) is mostly benign, it is frequently overtreated due to challenges in differentiating it from malignant tumors. Therefore, the purpose of this study was to investigate the clinical, laboratory, ultrasonographic, and pathological characteristics of SO to improve the accuracy of preoperative diagnosis. We retrospectively reviewed the clinical data and imaging characteristics of 103 patients with postoperative pathology for SO at the Affiliated Hospital of Qingdao University from May 2013 to November 2023. Among the 103 patients diagnosed with SO, 95 were benign SO (median age: 44.9 years (range, 15–83 years)) and 8 were malignant struma ovarii (MSO) (49.6 years (range, 15–83 years)), malignant cases accounted for 7.8

Objective: To develop a machine learning (ML) model using clinical data and ultrasound features for gout prediction, and apply SHapley Additive exPlanations (SHAP) for model interpretation. Methods: This study analyzed 609 patients' first metatarsophalangeal (MTP1) joint ultrasound data from two institutions. Institution 1 data (n = 571) were split into training cohort (TC) and internal testing cohort (ITC) (8:2 ratio), while Institution 2 data (n = 92) served as external testing cohort (ETC). Key predictors were selected using Random Forest (RF), Least Absolute Shrinkage and Selection Operator (LASSO), and Extreme Gradient Boosting (XGBoost) algorithms. Six ML models were evaluated using standard performance metrics, with SHAP analysis for model interpretation. Results: Five key predictors were identified: serum uric acid (SUA), deep learning (DL) model predictions, tophus, bone erosion, and double contour sign (DCs). The logistic regression (LR) model demonstrated optimal performance, achieving Area Under the Curve (AUC) values of 0.870 (95% CI: 0.820-0.920) in ITC and 0.854 (95% CI: 0.804-0.904) in ETC. The model showed good calibration with Brier scores of 0.138 and 0.159 in ITC and ETC, respectively. Conclusion: This study developed an interpretable ML model for gout prediction and utilized SHAP to elucidate feature contributions, establishing a foundation for future applications in clinical decision support for gout diagnosis.

OBJECTIVES:The rising global prevalence of gout necessitates advancements in diagnostic methodologies. Ultrasonographic imaging of the foot has become an important diagnostic modality for gout because of its non-invasiveness, cost-effectiveness, and real-time imaging capabilities. This study aims to develop and validate a deep learning-based artificial intelligence (AI) model for automated gout diagnosis using ultrasound images. MATERIALS AND METHODS:In this study, ultrasound images were primarily acquired at the first metatarsophalangeal joint (MTP1) from 598 cases in two institutions: 520 from Institution 1 and 78 from Institution 2. From Institution 1's dataset, 66% of cases were randomly allocated for model training, while the remaining 34% constitute the internal test set. The dataset from Institution 2 served as an independent external validation cohort. A novel deep learning model integrating a patch-wise attention mechanism and multi-scale feature extraction was developed to enhance the detection of subtle sonographic features and optimize diagnostic performance. RESULTS:The proposed model demonstrated robust diagnostic efficacy, achieving an accuracy of 87.88%, a sensitivity of 87.85%, a specificity of 87.93%, and an area under the curve (AUC) of 93.43%. Additionally, the model generates interpretable visual heatmaps to localize gout-related pathological features, thereby facilitating interpretation for clinical decision-making. CONCLUSION:In this paper, a deep learning-based artificial intelligence (AI) model was developed for the automated detection of gout using ultrasound images, which achieved better performance than other models. Furthermore, the features highlighted by the model align closely with expert assessments, demonstrating its potential to assist in the ultrasound-based diagnosis of gout.

Brown tumours, reactive osteolytic lesions caused by hyperparathyroidism, are usually identified by X-ray, computed tomography, and magnetic resonance imaging, with ultrasonography of brown tumours being only rarely reported in the literature. In this paper, we present a case of brown tumours initially detected during ultrasound examination and subsequently confirmed. We describe the ultrasonographic characteristics in detail and discuss the clinical value of ultrasound as a screening and diagnostic tool. Furthermore, we summarise the clinical characteristics, imaging features, and treatment options of brown tumours based on a thorough analysis of the available literature.

Objective Accurate segmentation of lesions in medical videos is crucial for clinical diagnosis and treatment.Unlike static medical images,videos provide continuous temporal information,enabling tracking of lesion evolution and morphological changes.However,existing segmentation methods primarily focus on processing individual frames,failing to effectively capture temporal correlations across frames.While self-attention mechanisms have been used to model long-range dependencies,their quadratic computational complexity renders them inefficient for high-resolution video segmentation.Additionally,medical videos are often affected by motion blur,noise,and illumination variations,which further hinder segmentation accuracy.To address these challenges,this paper proposes a novel medical video segmentation algorithm that integrates neighborhood attention and a State Space Model(SSM).The approach aims to efficiently capture both local and global spatiotemporal features,improving segmentation accuracy while maintaining computational efficiency. Methods The proposed approach comprises two key stages:local feature extraction and global temporal modeling,designed to efficiently capture both spatial and temporal dependencies in medical video segmentation.In the first stage,a deep convolutional network is used to extract spatial features from each video frame,providing a detailed representation of anatomical structures.However,relying solely on spatial features is insufficient for medical video segmentation,as lesions often undergo subtle morphological changes over time.To address this,a neighborhood attention mechanism is introduced to capture short-term dependencies between adjacent frames.Unlike conventional self-attention mechanisms,which compute relationships across the entire frame,neighborhood attention selectively attends to local regions around each pixel,reducing computational complexity while preserving essential temporal coherence.This localized attention mechanism enables the model to focus on small but critical changes in lesion appearance,making it more robust to motion and deformation variations.In the second stage,an SSM module is integrated to capture long-range dependencies across the video sequence.Unlike Transformer-based approaches,which suffer from quadratic complexity due to the self-attention mechanism,the SSM operates with linear complexity,significantly improving computational efficiency while maintaining strong temporal modeling capabilities.To further enhance the processing of video-based medical data,a 2D selective scanning mechanism is introduced to extend the SSM from 1D to 2D.This mechanism enables the model to extract spatiotemporal relationships more effectively by scanning input data along multiple directions and merging the results,ensuring that both local and global temporal structures are well represented.The combination of neighborhood attention for local refinement and SSM-based modeling for long-range dependencies enables the proposed method to achieve a balance between segmentation accuracy and computational efficiency.The model is trained and evaluated on multiple medical video datasets to verify its effectiveness across different segmentation scenarios,demonstrating its capability to handle complex lesion appearances,background noise,and variations in imaging conditions. Results and Discussions The proposed method is evaluated on three widely used medical video datasets:thyroid ultrasound,CVC-ClinicDB,and CVC-ColonDB.The model achieves Intersection Over Union(IOU)scores of 72.7%,82.3%,and 72.5%,respectively,outperforming existing state-of-the-art methods.Compared to the Vivim model,the proposed method improves IOU by 5.7%,1.7%,and 5.5%,highlighting the advantage of leveraging temporal information.In terms of computational efficiency,the model achieves 23.97 frames per second(fps)on the thyroid ultrasound dataset,making it suitable for real-time clinical applications.A comparative analysis against several state-of-the-art methods,including UNet,TransUNet,PraNet,U-Mamba,LKM-UNET,RMFG,SALI,and Vivim,demonstrates that the proposed method consistently outperforms these approaches,particularly in complex scenarios with significant background noise,occlusions,and motion artifacts.Specifically,on the CVC-ClinicDB dataset,the proposed model achieves an IOU of 82.3%,exceeding the previous best approach(80.9%).On the CVC-ColonDB dataset,which presents additional challenges due to lighting variations and occlusions,the model attains an IOU of 72.5%,outperforming the previous best method(70.8%).These results highlight the importance of incorporating both local and global temporal information to enhance segmentation accuracy and robustness in medical video analysis. Conclusions This study proposes a medical video segmentation algorithm that integrates neighborhood attention and an SSM to capture both local and global spatiotemporal features.This integration enables an effective balance between segmentation accuracy and computational efficiency.Experimental results demonstrate the superiority of the proposed method over existing approaches across multiple medical video datasets.The main contributions include:the combined use of neighborhood attention and SSM for efficient spatiotemporal feature extraction;a 2D selective scanning mechanism that extends SSMs for video-based medical segmentation;improved segmentation performance exceeding that of state-of-the-art models while maintaining real-time processing capability;and enhanced robustness to background noise and lighting variations,improving reliability in clinical applications.Future work will focus on incorporating prior knowledge and anatomical constraints to refine segmentation accuracy in cases with ambiguous lesion boundaries;developing advanced boundary refinement strategies for challenging scenarios;extending the framework to multi-modal imaging data such as CT and MRI videos;and optimizing the model for deployment on edge devices to support real-time processing in point-of-care and mobile healthcare settings.

Deep learning techniques have significantly enhanced the convenience and precision of ultrasound image diagnosis, particularly in the crucial step of lesion segmentation. However, recent studies reveal that both train-from-scratch models and pre-trained models often exhibit performance disparities across sex and age attributes, leading to biased diagnoses for different subgroups. In this paper, we propose APPLE, a novel approach designed to mitigate unfairness without altering the parameters of the base model. APPLE achieves this by learning fair perturbations in the latent space through a generative adversarial network. Extensive experiments on both a publicly available dataset and an in-house ultrasound image dataset demonstrate that our method improves segmentation and diagnostic fairness across all sensitive attributes and various backbone architectures compared to the base models. Through this study, we aim to highlight the critical importance of fairness in medical segmentation and contribute to the development of a more equitable healthcare system.

Introduction:Triple-negative breast cancer (TNBC) is known for its high malignancy, limited clinical treatment options, and poor chemotherapy outcomes. Although some advancements have been made using nanotechnology-based chemotherapy for TNBC treatment, the controlled and on-demand release of chemotherapeutic drugs at the tumor site remains a challenge. Methods:We manufactured DOX/BaTiO3@cRGD-Lip (DBRL) nanoparticles as an ultrasound (US)-controlled release platform targeting the delivery of Doxorubicin (DOX) for TNBC treatment. The nanoparticles incorporate DSPE-Se-Se-PEG-NH2 as the liposomal membrane for ROS responsiveness, cRGD peptide for TNBC cell selectivity, and polyethylene glycol for minimized phagocytic cell absorption. Results:The DBRL+US group achieved significant tumor inhibition (70.27% compared to control group, p < 0.001), while maintaining excellent biocompatibility with over 90% cell viability in normal cells. The selective cytotoxicity was evidenced by a 55.70% cell death rate in 4T1 cancer cells under US activation. DBRL showed enhanced tumor accumulation with peak fluorescence intensity of (1.01 ± 0.33)×109 at 12 hours post-injection. Conclusion:This targeted nanocomposite material paves a new prospect for future precise piezoelectric catalytic therapy for the treatment of TNBC.

A lesion detection method in ultrasound images based on feature feedback mechanism is proposed to realize real-time accurate localization and detection of ultrasound lesions. The proposed method consists of two parts: feature extraction network based on feature feedback mechanism and adaptive detection head based on divide-and-conquer strategy. The feature feedback network fully learns the global context information and local low-level semantic details of ultrasound images through feedback feature selection and weighted fusion calculation to improve the recognition ability of local lesion features. The adaptive detection head performs divide-and-conquer preprocessing on the multi-level features extracted by the feature feedback network. By combining physiological prior knowledge and feature convolution, adaptive modeling of lesion shape and scale features is performed on features at all levels to enhance the detection effect of the detection head on lesions of different sizes under multi-level features. The proposed method is tested on the thyroid ultrasound image dataset, and 70.3％ AP, 99.0％ AP50 and 88.4％ AP75 are obtained. Experimental results show that the proposed algorithm can achieve more accurate real-time detection and positioning of ultrasound image lesions in comparison with mainstream detection algorithm.

Aim: To explore whether ultrasound (US) can be employed to identify the underlying characteristics associated with pain in patients with podagra by evaluating the relationship between ultrasound findings and clinical pain. Material and methods: Patients with podagra were recruited and grouped into a pain group (G1, 82 patients) and a non pain group (G2, 123 patients). US features were collected and compared. US data were analyzed by binary logistic regression analysis and ROC analysis. Interobserver reliability was assessed, too. Results: A total of 205 patients (196 male and 9 female) were enrolled in this study. In multivariate analysis, the thickness of the synovium (OR=1.928, CI=1.074-3.463), CD (color Doppler) signal of the synovium(OR=1.458, CI=1.011-2.103), and CD signal of the tophi (OR=1.576, CI=1.142-2.177) were identified as risk factors for clinical pain. Areas under the ROC curves (AUC) were 0.713, 0.686 and 0.641 for the three indicators, respectively. The best cutoff points were 1 mm for the thickness of the synovium, grade 1 for the CD signal of the synovium and grade 2 for the CD signal of the tophi. Conclusions: Ultrasound can provide valuable information for determining underlying features associated with pain in patients with podagra.

Objectives: To evaluate whether ultrasound findings of monosodium urate (MSU) crystal deposition predict frequent gout flares in index joints over 12 months. Methods: This single-center study enrolled people with at least one gout flare involving the MTP1, ankle or knee joint. The most painful or most frequently joint was identified as index joint for analysis. All participants were started on urate-lowering therapy and had an ultrasound scan of the index joints at the baseline visit. OMERACT scores (for tophus, double contour sign and aggregates) were used to analyze whether ultrasound scores predicted frequent (>= 2) gout flares in the index joint over 12 months. Results: Frequent flares were significantly higher in those with ultrasound findings in all index joints (MTP1: tophus: 85.0% vs 46.0%, P < 0.001, aggregates: 78.8% vs 59.0%, P < 0.01; ankle: tophus: 54.6% vs 20.8%, P < 0.001; aggregates: 60.0% vs 35.9%, P < 0.05; knee: tophus: 68.4% vs 28.6%, P < 0.05). For the MTP1, for each 1-point increase in tophus score, the odds of frequent gout flares increased by 5.19 [(95%CI: 1.26-21.41), 7.91 [(95%CI: 2.23-28.14), and 13.79 [(95%CI: 3.79-50.20)] fold respectively. For the ankle, a tophus score of 3 markedly improved the prediction of the frequent flares [OR= 9.24 (95%CI=2.85-29.91)]. Semi-quantitative sum scores were associated with frequent flares with an OR (95%CI) of 13.66 (3.44-54.18), P < 0.001 at the MTP1, 7.05 (1.98-25.12), P < 0.001 at the ankle. Conclusion: Ultrasound features of MSU crystal deposition at the MTP1 and knee predict subsequent risk of frequent gout flares in the same joints following initiation of urate-lowering therapy, with the highest risk in those with high tophus scores.

Objective: To explore the clinical and ultrasonographic predictors for aggressive behaviors preoperatively in sporadic medullary thyroid carcinomas (MTCs). Materials and Methods: The preoperative clinical and ultrasonographic characteristics of patients diagnosed with MTCs between January 2009 and May 2022 were retrospectively reviewed. MTCs were described and categorized according to the American College of Radiology (ACR) thyroid imaging reporting and data system classification by 2 radiologists. Interobserver agreement was evaluated by kappa test. Univariate and multivariate analyses were performed to identify predictors of aggressive behaviors in MTCs. The log-rank test was utilized to compare differences in Kaplan-Meier (K-M) curves for postoperative disease-free survival (PDFS). Results: A total of 120 patients were enrolled in the final study. Male sex was significant risk factor for metastasis, perithyroidal invasion, and lateral cervical lymph node (LCLN) metastasis [odds ratio (OR): 3.109, P = .019; OR: 5.316, P = .018; OR: 5.154 P = .012, respectively]. The kappa values for all ultrasonic characteristics were high (ranged from 0.811 to 0.941). Size, focality, and margin of the nodule were independent risk factors for metastasis, as well as for LCLN metastasis. Whereas margin (P < .001) and a subcapsular location (P = .021) were risk factors for perithyroidal invasion. According to K-M analysis, PDFS of patients differed significantly between groups with/without metastasis (P < .001), groups with/without perithyroidal extension (P < .001) and groups with/without LCLN metastasis (P < .001). Conclusions: Male sex is an independent risk factor for metastasis, perithyroidal invasion, and LCLN metastasis. The large size (≥2.55 cm for metastasis, ≥2.15 cm for LCLN metastasis, respectively), multifocality, and irregular margin of nodules were independent risk factors for both metastasis and LCLN metastasis. Extrathyroidal extension and a subcapsular location were risk factors for perithyroidal invasion. Moreover, patients with metastasis/perithyroidal extension/LCLN metastasis exhibited worse PDFS.

INTRODUCTION:The aim of the study was to explore the causes and clinical significance of hyperechoic renal medulla observed by ultrasonography in patients with primary gout. METHODS:This study included 2,107 patients with primary gout treated in the Gout Clinic of our hospital from 2016 to 2022. The clinical data and biochemical data of these patients were collected and analyzed. According to the presence or absence of punctate hyperechogenicity in the renal medulla on ultrasound examination, the patients were divided into the hyperechoic medulla (HM) and the normal hypoechoic medulla (NM) groups, and the HM group was further divided into the partial HM (P-HM) and fulfilled HM (F-HM) subgroups according to the distribution range of hyperechogenicity. RESULTS:Among the 2,107 patients with primary gout, 380 had hyperechoic renal medulla on renal ultrasound, including 106 patients with F-HM and 274 with P-HM. There were significant differences in the gout duration, urate arthropathy number, serum urate (SU) level, clinical tophi number, blood urea nitrogen, serum creatinine (sCr), and estimated glomerular filtration rate between the HM and NM groups or between the F-HM and P-HM subgroups (p < 0.05). Multivariate regression analysis showed that the presence of HM was positively correlated with gout duration, urate arthropathy number, gout attack frequency, SU, and sCr. The number of clinical tophi and sCr were closely related to F-HM. CONCLUSION:Ultrasound examination showed that a high medulla echo in patients with gout was often related to renal function damage. P-HM may be a transitory condition between NM and F-HM in patients with gout.

Rationale and Objectives To assess the consistency between ultrasound and dual-energy computed tomography (DECT) for the diagnosis of gout in the knee joint. Materials and Methods The ultrasound and DECT images of 176 knee joints from 167 patients diagnosed with gout at the Gout Specialty Clinic of Qingdao University Affiliated Hospital from February 2022 to December 2023 were retrospectively analyzed. The knee joint was segmented into five anatomical regions: intra-articular, anterior, posterior, medial, and lateral. The location of monosodium urate (MSU) crystal deposition was recorded. Tophi were classified as hypoechogenic, isoechogenic, hyperechogenic, or strongly echogenic. The Kappa test was used to assess the consistency between the two examination methods in different regions of the knee joint. The McNemar chi-square test was utilized to conduct a differential analysis between the DECT and ultrasound results. The chi-square test was used to assess differences in the rate of tophi detection with different echogenicities by DECT. Pearson’s correlation coefficient was used to assess the correlation between MSU crystal deposition volume and clinically relevant indicators. Results Double contour (61.4%) was the most common intra-articular ultrasound sign. In the extra-articular region, MSU crystals were commonly deposited in and around the popliteal groove region (ultrasound: 52.3%; DECT: 60.0%). Corresponding MSU deposits on DECT were found in 7 of 54 joints with aggregates detected on ultrasound, and in 15 of 108 joints with DC. Tophi with hyperechogenicity or strong echogenicity were more likely to be detected on DECT than those with hypoechoic or isoechoic features (84.3% and 90.9% vs. 55.1% and 27.8%, respectively). For the assessment of MSU deposits, ultrasound showed an overall higher positive rate than DECT (81.1% vs. 72.2%), with poor consistency between the two examinations (κ = 0.177). In distinct anatomical regions, ultrasound and DECT showed high consistency in the medial (κ = 0.651) and lateral (κ = 0.705) views, with no significant difference. The intra-articular (κ = 0.316) and anterior (κ = 0.346) regions exhibited only fair consistency, with statistically significant diagnostic differences. When exclusively assessing cases with tophi, ultrasound and DECT demonstrated similar consistency in the medial, lateral and anterior views (κ = 0.633, 0.712, and 0.400, respectively), with statistically significant differences. In the intra-articular region, the consistency was reduced (κ = 0.237), and the differences were statistically significant. Conclusion Ultrasound and DECT are effective methods to detect MSU deposition in gout of the knee. However, the consistency between the two techniques varies in different anatomical locations. Clinical assessment should be tailored based on the specific anatomical position. DECT is advantageous for the evaluation of intra-articular MSU deposits, while ultrasound is more sensitive for the early detection of scattered MSU deposits.

Primary melanoma of the parotid gland is an extremely rare and challenging tumor with a poor prognosis, and its ultrasonic characteristics have yet to be reported. This article presents a case of a 77-year-old man with a left parotid mass that was confirmed as a melanoma following surgery. The ultrasonic features of melanoma were examined in detail, with a particular focus on their diagnostic value. Furthermore, we summarized the clinical characteristics, treatment options, and outcomes associated with primary melanoma of the parotid gland based on a thorough analysis of the available literature.

The U-shaped architecture has emerged as a crucial paradigm in the design of medical image segmentation networks. However, due to the inherent local limitations of convolution, a fully convolutional segmentation network with U-shaped architecture struggles to effectively extract global context information, which is vital for the precise localization of lesions. While hybrid architectures combining CNNs and Transformers can address these issues, their application in real medical scenarios is limited due to the computational resource constraints imposed by the environment and edge devices. In addition, the convolutional inductive bias in lightweight networks adeptly fits the scarce medical data, which is lacking in the Transformer based network. In order to extract global context information while taking advantage of the inductive bias, we propose CMUNeXt, an efficient fully convolutional lightweight medical image segmentation network, which enables fast and accurate auxiliary diagnosis in real scene scenarios. CMUNeXt leverages large kernel and inverted bottleneck design to thoroughly mix distant spatial and location information, efficiently extracting global context information. We also introduce the Skip-Fusion block, designed to enable smooth skip-connections and ensure ample feature fusion. Experimental results on multiple medical image datasets demonstrate that CMUNeXt outperforms existing heavyweight and lightweight medical image segmentation networks in terms of segmentation performance, while offering a faster inference speed, lighter weights, and a reduced computational cost. The code is available at https://github.com/FengheTan9/CMUNeXt.

Accurate detection of thyroid lesions is a critical aspect of computer-aided diagnosis. However, most existing detection methods perform only one feature extraction process and then fuse multi-scale features, which can be affected by noise and blurred features in ultrasound images. In this study, we propose a novel detection network based on a feature feedback mechanism inspired by clinical diagnosis. The mechanism involves first roughly observing the overall picture and then focusing on the details of interest. It comprises two parts: a feedback feature selection module and a feature feedback pyramid. The feedback feature selection module efficiently selects the features extracted in the first phase in both space and channel dimensions to generate high semantic prior knowledge, which is similar to coarse observation. The feature feedback pyramid then uses this high semantic prior knowledge to enhance feature extraction in the second phase and adaptively fuses the two features, similar to fine observation. Additionally, since radiologists often focus on the shape and size of lesions for diagnosis, we propose an adaptive detection head strategy to aggregate multi-scale features. Our proposed method achieves an AP of 70.3% and AP50 of 99.0% on the thyroid ultrasound dataset and meets the real-time requirement. The code is available at https://github.com/HIT-wanglingtao/Thinking-Twice.

Medical ultrasound imaging technology is currently the preferred method for early diagnosis of thyroid nodules. Radiologists’ analysis of ultrasound images is highly dependent on their clinical experience and is susceptible to intra- and inter-observer variability. Although end-to-end deep learning technique can address these limitations, the difficulty of acquiring annotated medical image makes it very challenging. Transfer learning can alleviate the problems, but the large gap between source and target domain will lead to negative transfer. In this paper, a novel transfer learning method with distant domain high-level feature fusion (DHFF) model is proposed. It reduces the distribution distance between the source domain and the target domain while maintaining the characteristics of respective domains, which can avoid excessive feature fusion while enabling the model to learn more valuable transfer knowledge. The DHFF is validated by multiple public source and private target datasets in experiments. The results show that the classification accuracy of DHFF is up to 88.92% with thyroid ultrasound auxiliary source domains, which is up to 8% higher than existing transfer and distant transfer algorithms.