Altered Callosal Morphology and Connectivity in Asymptomatic Carotid Stenosis.

BACKGROUND: Carotid stenosis, even in the clinically asymptomatic stage, causes cognitive impairment, silent lesions, and hemispheric changes. The corpus callosum (CC) is crucial for hemispheric cortical integration and specialization. PURPOSE: To examine if CC morphology and connectivity relate to cognitive decline and lesion burden in asymptomatic carotid stenosis (ACS). STUDY TYPE: Retrospective, cross-sectional. POPULATION: 33 patients with unilaterally severe (70%) ACS and 28 demographically and comorbidity-matched controls. A publicly available healthy adult lifespan (ages between 18 and 80; n = 483) MRI dataset was also included. FIELD STRENGTH/SEQUENCE: A 3.0 T; T1 MPRAGE and diffusion weighted gradient echo-planar imaging sequences. ASSESSMENT: Structural MRI and multidomain cognitive data were obtained. Midsagittal CC area, circularity, thickness, integrity, and probabilistic tractography were calculated and correlated with cognitive tests and white matter hyperintensity. Fractional anisotropy, mean diffusivity (MD), and radial diffusivity were determined from DTI. STATISTICAL TESTS: Independent two-sample t-tests, χ2 tests, Mann-Whitney U, locally weighted scatterplot smoothing (LOWESS) curve fit, and Pearson correlation. A P value < 0.05 was considered statistically significant. RESULTS: Patients with ACS demonstrated significant reductions in callosal area, circularity, and thickness compared to controls. The callosal atrophy was significantly correlated with white matter hyperintensity size (r = -0.629, P < 0.001). Voxel-wise analysis of diffusion measures in the volumetric CC showed that ACS patients exhibited significantly lower fractional anisotropy and higher MD and radial diffusivity in the genu and splenium of the CC than controls. Further lifespan trajectory analysis showed that although the midsagittal callosal area, circularity, and thickness exhibited age-related decreases, the values in the ACS patients were significantly lower in all age groups. DATA CONCLUSION: Midsagittal callosal atrophy and connectivity reflect the load of silent lesions and the severity of cognitive decline, respectively, suggesting that CC degeneration has potential to serve as an early marker in ACS. LEVEL OF EVIDENCE: 3 TECHNICAL EFFICACY STAGE: 2.

Hemispheric coupling between structural and functional asymmetries in clinically asymptomatic carotid stenosis with cognitive impairment.

Advanced carotid stenosis is a known risk factor for ischemic stroke and vascular dementia, and it is associated with multidomain cognitive impairment as well as asymmetric alterations in hemispheric structure and function. Here we introduced a novel measure-the asymmetry index of amplitude of low-frequency fluctuations (ALFF_AI)-derived from resting-state functional magnetic resonance imaging. This measure captures the hemispheric asymmetry of intrinsic brain activity using high-dimensional registration. We aimed to investigate functional brain asymmetric alterations in patients with severe asymptomatic carotid stenosis (SACS). Furthermore, we extended the analyses of ALFF_AI to different frequencies to detect frequency-specific alterations. Finally, we examined the coupling between hemispheric asymmetric structure and function and the relationship between these results and cognitive tests, as well as the white matter hyperintensity burden. SACS patients presented significantly decreased ALFF_AI in several clusters, including the visual, auditory, parahippocampal, Rolandic, and superior parietal regions. At low frequencies (0.01-0.25 Hz), the ALFF_AI exhibited prominent group differences as frequency increased. Further structure-function coupling analysis indicated that SACS patients had lower coupling in the lateral prefrontal, superior medial frontal, middle temporal, superior parietal, and striatum regions but higher coupling in the lateral occipital regions. These findings suggest that, under potential hemodynamic burden, SACS patients demonstrate asymmetric hemispheric configurations of intrinsic activity patterns and a decoupling between structural and functional asymmetries.

Neuroanatomical correlates of cognitive impairment following basal ganglia-thalamic post-hemorrhagic stroke: Uncovering network-wide alterations in hemispheric gray matter asymmetry.

Cognitive impairment and recovery are central issues in hemorrhagic stroke. This study aimed to investigate whether post-hemorrhagic stroke cognitive impairment (PhSCI) is associated with cortical gray matter (GM) loss and hemispheric asymmetry changes and whether these changes could predict improvements in cognitive function during the recovery. Nineteen patients with PhSCI, comprising 10 with basal ganglia hemorrhage and 9 with thalamic hemorrhage, were recruited. Among them, 9 completed a course of repetitive transcranial magnetic stimulation (rTMS). Additionally, 19 demographically and comorbidity-matched healthy controls were also included. Structural brain MRI and cognitive assessments were performed. Voxel-wise GM volume and hemispheric asymmetry were analyzed. The PhSCI patients exhibited bilateral, yet asymmetric, GM losses in the hippocampus, fusiform, lateral temporal, prefrontal, somatomotor, and inferior parietal regions. The analysis of GM asymmetry revealed that patients showed rightward GM in the lateral temporal, somatomotor, and inferior parietal regions. Among the 9 PhSCI patients who completed rTMS, there was a marginal trend of regional GM increase and leftward GM, and these changes were in parallel with the improvements in cognitive tests. Further lesion connectivity and metanalytic mapping identified two interconnected systems linked to the lesions, which were anchored in the default mode, somatomotor, and salience/cognitive control networks and in the cognitive domains of memory, language, decision-making, and executive function. In conclusion, PhSCI patients exhibited network-wide cortical GM losses, distal to subcortical hemorrhagic lesions, and hemisphere asymmetry changes. These changes appear to predict rTMS-related cognitive improvements, suggesting that even subcortical focal lesions can lead to alterations in distal cortical neuroanatomical architecture. Our preliminary findings provide new insights into the neuroanatomical basis of PhSCI.

Multi-view radiomics and deep learning modeling for prostate cancer detection based on multi-parametric MRI.

Introduction: This study aims to develop an imaging model based on multi-parametric MR images for distinguishing between prostate cancer (PCa) and prostate hyperplasia. Methods: A total of 236 subjects were enrolled and divided into training and test sets for model construction. Firstly, a multi-view radiomics modeling strategy was designed in which different combinations of radiomics feature categories (original, LoG, and wavelet) were compared to obtain the optimal input feature sets. Minimum-redundancy maximum-relevance (mRMR) selection and least absolute shrinkage selection operator (LASSO) were used for feature reduction, and the next logistic regression method was used for model construction. Then, a Swin Transformer architecture was designed and trained using transfer learning techniques to construct the deep learning models (DL). Finally, the constructed multi-view radiomics and DL models were combined and compared for model selection and nomogram construction. The prediction accuracy, consistency, and clinical benefit were comprehensively evaluated in the model comparison. Results: The optimal input feature set was found when LoG and wavelet features were combined, while 22 and 17 radiomic features in this set were selected to construct the ADC and T2 multi-view radiomic models, respectively. ADC and T2 DL models were built by transferring learning from a large number of natural images to a relatively small sample of prostate images. All individual and combined models showed good predictive accuracy, consistency, and clinical benefit. Compared with using only an ADC-based model, adding a T2-based model to the combined model would reduce the model's predictive performance. The ADCCombinedScore model showed the best predictive performance among all and was transformed into a nomogram for better use in clinics. Discussion: The constructed models in our study can be used as a predictor in differentiating PCa and BPH, thus helping clinicians make better clinical treatment decisions and reducing unnecessary prostate biopsies.

Asymptomatic carotid stenosis is associated with both edge and network reconfigurations identified by single-subject cortical thickness networks.

Background and purpose: Patients with asymptomatic carotid stenosis, even without stroke, are at high risk for cognitive impairment, and the neuroanatomical basis remains unclear. Using a novel edge-centric structural connectivity (eSC) analysis from individualized single-subject cortical thickness networks, we aimed to examine eSC and network measures in severe (> 70%) asymptomatic carotid stenosis (SACS). Methods: Twenty-four SACS patients and 24 demographically- and comorbidities-matched controls were included, and structural MRI and multidomain cognitive data were acquired. Individual eSC was estimated via the Manhattan distances of pairwise cortical thickness histograms. Results: In the eSC analysis, SACS patients showed longer interhemispheric but shorter intrahemispheric Manhattan distances seeding from left lateral temporal regions; in network analysis the SACS patients had a decreased system segregation paralleling with white matter hyperintensity burden and recall memory. Further network-based statistic analysis identified several eSC and subgraph features centred around the Perisylvian regions that predicted silent lesion load and cognitive tests. Conclusion: We conclude that SACS exhibits abnormal eSC and a less-optimized trade-off between physical cost and network segregation, providing a reference and perspective for identifying high-risk individuals.

Circulating Tumor Cell and Metabolites as Novel Biomarkers for Early-Stage Lung Cancer Diagnosis.

BACKGROUND: Lung cancer is a malignant tumor that has the highest morbidity and mortality rate among all cancers. Early diagnosis of lung cancer is a key factor in reducing mortality and improving prognosis. METHODS: In this study, we performed CTC next-generation sequencing (NGS) in early-stage lung cancer patients to identify lung cancer-related gene mutations. Meanwhile, a serum liquid chromatography-tandem mass spectrometry (LC-MS) untargeted metabolomics analysis was performed in the CTC-positive patients. To screen potential diagnostic markers for early lung cancer. RESULTS: 62.5% (30/48) of lung cancer patients had ≥1 CTC. By CTC NGS, we found that > 50% of patients had 4 commonly mutated genes, namely, NOTCH1, IGF2, EGFR, and PTCH1. 47.37% (9/19) patients had ARIDH1 mutations. Additionally, 30 CTC-positive patients and 30 healthy volunteers were subjected to LC-MS untargeted metabolomics analysis. We found 100 different metabolites, and 10 different metabolites were identified through analysis, which may have potential clinical application value in the diagnosis of CTC-positive early-stage lung cancer (AUC >0.9). CONCLUSIONS: Our results indicate that NGS of CTC and metabolomics may provide new tumor markers for the early diagnosis of lung cancer.

Spiking Cortical Model Based Multimodal Medical Image Fusion by Combining Entropy Information with Weber Local Descriptor.

Multimodal medical image fusion (MIF) plays an important role in clinical diagnosis and therapy. Existing MIF methods tend to introduce artifacts, lead to loss of image details or produce low-contrast fused images. To address these problems, a novel spiking cortical model (SCM) based MIF method has been proposed in this paper. The proposed method can generate high-quality fused images using the weighting fusion strategy based on the firing times of the SCM. In the weighting fusion scheme, the weight is determined by combining the entropy information of pulse outputs of the SCM with the Weber local descriptor operating on the firing mapping images produced from the pulse outputs. The extensive experiments on multimodal medical images show that compared with the numerous state-of-the-art MIF methods, the proposed method can preserve image details very well and avoid the introduction of artifacts effectively, and thus it significantly improves the quality of fused images in terms of human vision and objective evaluation criteria such as mutual information, edge preservation index, structural similarity based metric, fusion quality index, fusion similarity metric and standard deviation.