BMC Medical Imaging最新文献

Purpose: This study aimed to evaluate left ventricular (LV) deformation and tissue characteristics using cardiac magnetic resonance (CMR) in patients with hypertrophic cardiomyopathy (HCM) and heart failure with preserved ejection fraction (HFpEF), to examine their associations with heart failure status, and to explore the correlations between CMR parameters and the H₂FPEF score.

Methods: This retrospective study included 105 patients with HCM who underwent 3.0-T CMR. Participants were classified into HFpEF (n = 46) and non-HF (n = 59) groups according to the 2019 ESC HFA-PEFF algorithm. Global radial strain (GRS), global circumferential strain (GCS), global longitudinal strain (GLS), and corresponding systolic and early-diastolic strain rates were derived using CMR feature tracking. Myocardial tissue characterization included native T1 and T2 mapping, extracellular volume fraction (ECV), and late gadolinium enhancement (LGE). Group differences were assessed with t-tests or chi-square tests. Associations between strain, tissue parameters, and the H₂FPEF score were evaluated using Spearman correlations. Multivariable logistic regression was performed to identify independent CMR predictors of HFpEF.

Results: Compared with non-HF patients, those with HCM-HFpEF showed significantly reduced LV systolic and early-diastolic strain rates, including sGRSr (P = 0.010), sGCSr (P = 0.044), sGLSr (P = 0.018), and eGLSr (P = 0.006). They also demonstrated a higher prevalence and greater extent of LGE, as well as elevated native T1 and ECV values (all P < 0.05). Strain parameters correlated significantly with tissue characteristics (native T1 and mean ECV), except for GCS and ECV. In multivariable analysis, drinking, atrial fibrillation, lower LV-eGLSr, and higher ECV in segments with maximal wall thickness were independently associated with HCM-HFpEF. The H₂FPEF score showed weak but significantly correlations with native T1, ECV, and T2 values in both global and hypertrophied myocardial segments (r = 0.199-0.252, all P < 0.05).

Conclusions: HCM patients with HFpEF exhibit both systolic and diastolic dysfunction, accompanied by increased diffuse and focal fibrosis. Independent predictors of HFpEF include lower LV-eGLSr, higher segmental ECV, atrial fibrillation, and drinking. The H₂FPEF score shows significant associations with tissue-level abnormalities, highlighting the complementary role of CMR-derived strain and tissue characterization in the early detection and risk stratification of HFpEF in HCM.

Purpose: To develop and validate a multimodal model that integrates radiomics features (RFs) and deep learning features (DFs) derived from preoperative multisequence magnetic resonance imaging (MRI) for the prediction of lymphovascular space invasion (LVSI) in patients with endometrial cancer (EC).

Methods: This multicenter, retrospective study enrolled 892 patients with postoperative pathologically confirmed EC. Preoperative MRI comprised T2-weighted imaging, contrast-enhanced T1-weighted imaging, and apparent diffusion coefficient maps, were analyzed. Regions of interest (ROIs) were manually delineated for 2D and 3D analyses. RFs were extracted using PyRadiomics, and DFs were obtained using pretrained VGG 11, ResNet 101, and DenseNet 121 architectures. Five single-modality models (2D-RF, 3D-RF, VGG11-DF, ResNet101-DF, and DenseNet121-DF) were developed. In addition, the integration of RFs and DFs were explored to construct combined models. Models were trained in a training cohort (n = 378) and evaluated in both internal (n = 160) and external (n = 354) validation cohorts. Model performance was evaluated by the area under the receiver operating characteristic curve (AUC).

Results: In the training cohort, the 2D-RF and 3D-RF models showed comparable performance for LVSI prediction (AUC: 0.775 vs. 0.772, P = 0.89). Among the deep learning models, DenseNet121-DF achieved the highest AUC (0.757), which was significantly higher than ResNet-101-DF (AUC: 0.671; P = 0.01) and not statistically different from VGG11-DF (AUC: 0.720, P = 0.20). The optimal combined model, integrating features from 2D-RF and DenseNet121-DF, yielded the highest performance in the training cohort (AUC: 0.796). These findings were confirmed in both the internal and external validation cohorts.

Conclusions: A multimodal MRI-based model integrating both RFs and DFs achieved superior performance for noninvasive prediction of LVSI in patients with EC. This approach holds potential to enhance preoperative risk stratification and guide personalized treatment planning.

Despite advancements in skin cancer diagnosis procedures, misclassification rates in early detection remain high, leading to delayed treatments and reduced survival rates. Existing manual diagnostic methods are often prone to inter-observer variability and human error, while traditional machine learning models struggle with imbalanced datasets and insufficient feature generalization. To address these challenges, this work proposes an Optimal Skin Cancer Classification Network (OSCC-Net), developed on the International Skin Imaging Collaboration-2019 (ISIC-2019) dataset. The model integrates an Adaptive Minority Over-Sampling Procedure (AMOP) to balance under-represented lesion classes, ensuring robust learning for minority lesion classes. The Stochastic Neighbourhood T-Distilling driven Score-Weighted Class Activation Mapping (STND-SWCAM) framework is introduced for feature analysis. It performs fine-grained lesion localization and interpretability, enabling better understanding of decisions. In the feature selection stage, a Grizzly Bear Fat Increase Optimizer with Density-Based Spatial Neighbourhood Discovery Algorithm (GBFIO-DSNDA) is employed to enhance discriminative feature extraction by eliminating redundant and noisy features. Finally, classification is performed using a Graph Convolutional Vision Neural Network (GC-VNN), which leverages spatial dependencies among lesion attributes for improved decision-making. Experimental evaluation reveals that, OSCC-Net achieves 98.32% accuracy, 98.43% precision, 98.40% recall, and 98.39% F1-Score, marking a substantial improvement over baselines shown in our experiments.

Background: This study explores the feasibility and effectiveness of an interpretable machine learning model for assessing the pathological grading of pancreatic ductal adenocarcinoma (PDAC) using radiomics and topological features derived from contrast-enhanced CT habitat subregions.

Methods: A retrospective study was conducted on a total of 306 patients with PDAC from two hospitals: a training cohort (n = 176), a validation cohort (n = 76), and a test cohort (n = 54). K-means clustering analysis was first used to segment portal venous phase CT images into three habitat regions. Radiomics features of the whole-tumour region, along with radiomics and topological features of each habitat region, were extracted respectively. LASSO regression was applied for feature dimensionality reduction to construct the radiomics score (Rad-score) for the whole-tumour region and the habitat score (H-score) for each habitat region. Meanwhile, logistic regression was used to identify statistically significant predictors from clinical and semantic features. Five machine learning algorithms were used to construct Habitat-TDA models, with interpretability analysis performed via SHAP analysis.

Results: Total volume, diabetes, and M staging were identified as independent risk factors for predicting the pathological grading of PDAC, and were used to construct the Clinical model. 6 radiomics features with non-zero coefficients were selected to calculate the Rad-score, which was further used to construct the WholeRad model. In the three habitat regions, 6, 5, and 6 topological and radiomics features were included to generate the H-score. The logistic regression algorithm performed best in the validation and test cohorts and was ultimately selected as the classifier for constructing the Habitat-TDA model. SHAP analysis showed that H-score1, derived from Habitat Region 1 (the habitat region with the lowest average CT value), has the most significant average impact on the model output intensity. The AUC values of the Habitat-TDA model in the training, validation, and test cohorts were 0.894, 0.872, and 0.829, all outperforming the clinical model (0.784, 0.765, 0.731) and WholeRad model (0.817, 0.810, 0.773).

Conclusions: The Habitat-TDA model improves the accuracy and interpretability of preoperative predictions of PDAC grading, providing a promising tool for personalised management.

Background: Studies reported that sphenoid sinus pneumatization (SSP) affects paraclinoid structures, including the optic canal (OC), the anterior clinoid process (ACP), the optic strut (OS), and the sella turcica (ST). We aimed to analyze this assesment based on sagittal and coronal SSP (SSSP and CSSP) patterns considering gender, laterality, and age.

Methods: Computed tomography (CT) images of 154 patients (78 males and 76 females), the dimensions of ST, OC, and ACP were measured, and caroticoclinoid foramen (CCF) and OC protrusion (OCP) were detected, as well as the prevalence of SSSP, CSSP, and ACP pneumatization (ACPP).

Results: The prevalence of ACPP and OCP significantly increased with the degree of SSSP and CSSP. The ACPP was found to be linked to the OCP and sulcal/postsulcal OS (p < 0.05). Also, CCF types were more common in sellar and postsellar SSSP (p = 0.041). The ST and OC dimensions were found to be influenced negatively by an increased degree of SSSP. As the degree of ACPP increased, the OC diameters and ST height decreased, while the OC and ACP lengths increased. The probability of having postsellar SSSP (p = 0.029), postrotundum CSSP (p = 0.000), and ACPP (p = 0.036) decreased with ageing. We found that the OC diameters and ST dimensions increased, while the lengths of the OC and ACP decreased with age.

Conclusion: Our results suggest that morphology and dimensions of paraclinoid structures can be predicted based on SSSP and CSSP in relation to gender and age. This is essential for improved treatment planning and avoidance of iatrogenic injury during surgery.

Objective: This study seeks to identify brain regions with atypical neural connectivity in individuals suffering from arthritis-related chronic pain, compared to healthy controls, using resting-state functional magnetic resonance imaging (rs-fMRI).

Methods: A seed-based connectivity analysis was conducted between the known pain-related regions of interest (ROIs), derived from the MNI (n = 76) and the Automated Anatomical Labeling (AAL) whole brain atlas (n = 116). We examined the connectivity differences in a cohort of 56 osteoarthritis patients and 20 healthy controls. Connectivity matrices were compared using permutation tests corrected for multiple comparisons, identifying statistically significant differences (p < 0.05). Subsequent network analysis resulted in hub scores, identifying the most central and influential brain regions within the altered connectivity network in patients experiencing pain.

Results: The most significant atypical neural connections in osteoarthritis patients were identified in the cingulate gyrus, insula, inferior parietal lobe, and thalamus, with notable involvement of the occipital lobe, postcentral gyrus, inferior frontal gyrus, orbitofrontal cortex, temporal lobe, hippocampus, and basal ganglia. The thalamus, cingulate gyrus, and insula emerged as key hubs in the chronic pain network, reflecting disrupted sensory, emotional, and cognitive pain processing. No significant connectivity differences were found in the brainstem, cerebellum, superior parietal lobe, precentral gyrus, superior and middle frontal gyri, or amygdala.

Conclusion: Our data-driven approach reveals specific neural connectivity disruptions in OA, highlighting connections between the cingulate gyrus, temporal lobe, and thalamus. These findings identify specific network disruptions in OA-related pain, offering insight into altered brain connectivity and potential avenues for targeted interventions.