Search Results (41)

Brain morphological abnormalities are common in patients with neurodevelopmental disorders (NDDs) and other neuropsychiatric disorders, often reflecting abnormal brain development and function. Genetic studies have found common genetic factors in NDDs and other neuropsychiatric disorders, although the etiology of brain structural changes in these disorders remains poorly understood. In this study, we analyzed magnetic resonance imaging (MRI) and genetic data from more than 30K individuals from the UK Biobank to evaluate whether NDD-risk copy number variants (CNVs) are also associated with neuroanatomical changes in both patients and neurotypical individuals. We found that the size differences in brain regions such as corpus callosum and cerebellum were associated with the deletions of specific areas of the human genome, and that specific neuroanatomical changes confer a risk of neuropsychiatric disorders. Furthermore, we observed that gene sets located in these genomic regions were enriched for pathways crucial for brain development and for phenotypes commonly observed in patients with NDDs. These findings highlight the link between CNVs, brain structure abnormalities, and the shared pathophysiology of NDDs and other neuropsychiatric disorders, providing new insights into the underlying mechanisms of these disorders and the identification of potential biomarkers for better diagnosis. Full article

The increased workload in pathology laboratories today means automated tools such as artificial intelligence models can be useful, helping pathologists with their tasks. In this paper, we propose a segmentation model (DRU-Net) that can provide a delineation of human non-small cell lung carcinomas and an augmentation method that can improve classification results. The proposed model is a fused combination of truncated pre-trained DenseNet201 and ResNet101V2 as a patch-wise classifier, followed by a lightweight U-Net as a refinement model. Two datasets (Norwegian Lung Cancer Biobank and Haukeland University Lung Cancer cohort) were used to develop the model. The DRU-Net model achieved an average of 0.91 Dice similarity coefficient. The proposed spatial augmentation method (multi-lens distortion) improved the Dice similarity coefficient from 0.88 to 0.91. Our findings show that selecting image patches that specifically include regions of interest leads to better results for the patch-wise classifier compared to other sampling methods. A qualitative analysis by pathology experts showed that the DRU-Net model was generally successful in tumor detection. Results in the test set showed some areas of false-positive and false-negative segmentation in the periphery, particularly in tumors with inflammatory and reactive changes. In summary, the presented DRU-Net model demonstrated the best performance on the segmentation task, and the proposed augmentation technique proved to improve the results. Full article

Stroke is a leading cause of death and disability in developed countries. We validated an AI-based prediction model for incident stroke using sensors such as fundus cameras and ophthalmoscopes for retinal images, along with socio-demographic data and traditional risk factors. The model was trained on a proprietary dataset of over 6500 participants, including 171 with 5-year incident strokes and 242 with 10-year incident strokes. The model provides separate 5-year and 10-year risk scores. The model was externally validated on the UK Biobank dataset (3000 subjects with 5-year incident strokes). Using retinal imaging, our models identified individuals with 5-year incident strokes with 80% sensitivity, 82% specificity, and an AUC of 0.83, and predicted 10-year incidents with 72% sensitivity, 78% specificity, and an AUC of 0.79. In comparison, for the 10-year model, the AUC for the Framingham score was 0.73, and the CHADS2 score was 0.74. On the Biobank external dataset, our 5-year model (without retinal features) demonstrated moderate but lower sensitivity (69.3%) and specificity (66.4%) compared to its performance on the proprietary dataset (with retinal features). Using a multi-ethnic dataset, we developed and validated a prediction model that improves stroke risk identification for 5-year and 10-year incidences by incorporating retinal features. Full article

Background and objectives: Type 2 diabetes mellitus (T2DM) affects brain white matter microstructure. While diffusion tensor imaging (DTI) has been used to study white matter abnormalities in T2DM, it lacks specificity for complex white matter tracts. Neurite orientation dispersion and density imaging (NODDI) offers a more specific approach to characterising white matter microstructures. This study aims to explore white matter alterations in T2DM using both DTI and NODDI and assess their association with disease duration and glycaemic control, as indicated by HbA1c levels. Methods and Materials: We analysed white matter microstructure in 48 tracts using data from the UK Biobank, involving 1023 T2DM participants (39% women, mean age 66) and 30,744 non-T2DM controls (53% women, mean age 64). Participants underwent 3.0T multiparametric brain imaging, including T1-weighted and diffusion imaging for DTI and NODDI. We performed region-of-interest analyses on fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), radial diffusivity (RD), orientation dispersion index (ODI), intracellular volume fraction (ICVF), and isotropic water fraction (IsoVF) to assess white matter abnormalities. Results: We observed reduced FA and ICVF, and increased MD, AD, RD, ODI, and IsoVF in T2DM participants compared to controls (p < 0.05). These changes were associated with longer disease duration and higher HbA1c levels (0 < r ≤ 0.2, p < 0.05). NODDI identified microstructural changes in white matter that were proxies for reduced neurite density and disrupted fibre orientation, correlating with disease progression and poor glucose control. In conclusion, NODDI contributed to DTI in capturing white matter differences in participants with type 2 diabetes, suggesting the feasibility of NODDI in detecting white matter alterations in type 2 diabetes. Type 2 diabetes can cause white matter microstructural abnormalities that have associations with glucose control. Conclusions: The NODDI diffusion model allows the characterisation of white matter neuroaxonal pathology in type 2 diabetes, giving biophysical information for understanding the impact of type 2 diabetes on brain microstructure. Future research should focus on the longitudinal tracking of these microstructural changes to better understand their potential as early biomarkers for cognitive decline in T2DM. Full article

To examine the association between post-stroke depression (PSD) and macrostructural and microstructural brain measures, and to explore whether changes in accelerometer-measured physical activity (PA) are associated with PSD, we conducted an exploratory study in UK Biobank with dementia-free participants diagnosed with at least one prior stroke. Eligible participants (n = 1186) completed an MRI scan. Depression was classified based on positive depression screening scores (PHQ-2 ≥ 3). Multivariate linear regression models assessed the relationships between depression and structural and diffusion measures generated from brain MRI scans. Logistic regression models were used to examine the relationship between accelerometer-measured daily PA and future depression (n = 367). Depression was positively associated with total white matter hyperintensities (WMHs) volume (standardized β [95% CI]—0.1339 [0.012, 0.256]; FDR-adjusted p-value—0.039), periventricular WMHs volume (standardized β [95% CI]—0.1351 [0.020, 0.250]; FDR-adjusted p-value—0.027), and reduced MD for commissural fibers (standardized β [95% CI]—−0.139 [−0.255, −0.024]; adjusted p-value—0.045). The odds of depression decreased by 0.3% for each daily minute spent in objectively measured light PA, while each minute spent in sleep from midnight to 6:00 AM was associated with a 0.9% decrease in the odds of depression. This early-stage analysis using a population cohort offers a scientific rationale for researchers using multimodal data sources to investigate the heterogenous nature of PSD and, potentially, identify stroke patients at risk of poor outcomes. Full article

Background: Despite the widespread use of lipid-lowering agents, the risk of atherosclerotic cardiovascular disease (ASCVD) remains; this residual risk has been attributed to remnant cholesterol (RC) levels. However, the causal associations between RC levels and various atherosclerosis-related cardiometabolic and vascular risk factors for ASCVD remain unclear. Methods: Using genetic and biochemical data of 108,876 Taiwan Biobank study participants, follow-up data of 31,790 participants, and follow-up imaging data of 18,614 participants, we conducted a genome-wide association study, a Functional Mapping and Annotation analysis, and bidirectional Mendelian randomization analyses to identify the genetic determinants of RC levels and the causal associations between RC levels and various cardiometabolic and vascular risk factors. Results: We found that higher RC levels were associated with higher prevalence or incidence of the analyzed risk factors. The genome-wide association study unveiled 61 lead genetic variants determining RC levels. The Functional Mapping and Annotation analysis revealed 21 gene sets exhibiting strong enrichment signals associated with lipid metabolism. Standard Mendelian randomization models adjusted for nonlipid variables and low-density lipoprotein cholesterol levels unraveled forward causal associations of RC levels with the prevalence of diabetes mellitus, hypertension, microalbuminuria, and metabolic liver disease. Reverse Mendelian randomization analysis revealed the causal association of diabetes mellitus with RC levels. Conclusions: RC levels, mainly influenced by genes associated with lipid metabolism, exhibit causal associations with various cardiometabolic risk factors, including diabetes mellitus, hypertension, microalbuminuria, and metabolic liver disease. This study provides further insights into the role of RC levels in predicting the residual risk of ASCVD. Full article

Objective: Functional magnetic resonance imaging data pose significant challenges due to their inherently noisy and complex nature, making traditional statistical models less effective in capturing predictive features. While deep learning models offer superior performance through their non-linear capabilities, they often lack transparency, reducing trust in their predictions. This study introduces the Time Reversal (TR) pretraining method to address these challenges. TR aims to learn temporal dependencies in data, leveraging large datasets for pretraining and applying this knowledge to improve schizophrenia classification on smaller datasets. Methods: We pretrained an LSTM-based model with attention using the TR approach, focusing on learning the direction of time in fMRI data, achieving over 98 % accuracy on HCP and UK Biobank datasets. For downstream schizophrenia classification, TR-pretrained weights were transferred to models evaluated on FBIRN, COBRE, and B-SNIP datasets. Saliency maps were generated using Integrated Gradients (IG) to provide post hoc explanations for pretraining, while Earth Mover’s Distance (EMD) quantified the temporal dynamics of salient features in the downstream tasks. Results: TR pretraining significantly improved schizophrenia classification performance across all datasets: median AUC scores increased from 0.7958 to 0.8359 (FBIRN), 0.6825 to 0.7778 (COBRE), and 0.6341 to 0.7224 (B-SNIP). The saliency maps revealed more concentrated and biologically meaningful salient features along the time axis, aligning with the episodic nature of schizophrenia. TR consistently outperformed baseline pretraining methods, including OCP and PCL, in terms of AUC, balanced accuracy, and robustness. Conclusions: This study demonstrates the dual benefits of the TR method: enhanced predictive performance and improved interpretability. By aligning model predictions with meaningful temporal patterns in brain activity, TR bridges the gap between deep learning and clinical relevance. These findings emphasize the potential of explainable AI tools for aiding clinicians in diagnostics and treatment planning, especially in conditions characterized by disrupted temporal dynamics. Full article

Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) is the leading cause of liver-related morbidity and mortality. Although the invasive liver biopsy remains the golden standard for MASLD diagnosis, Magnetic Resonance Imaging-derived Proton Density Fat Fraction (MRI-PDFF) is an accurate, non-invasive method for the assessment of treatment response. This study aimed at developing a Polygenic Risk Score (PRS) to improve MRI-PDFF prediction using UK Biobank data to assess an individual’s genetic liability to MASLD. Methods: We iteratively sequestered 10% of MRI-PDFF samples as a validation set and split the rest of each dataset into base and target partitions, containing GWAS summary statistics and raw genotype data, respectively. PRSice2 was deployed to derive PRS candidates. Based on the frequency of SNP appearances along the PRS candidates, we generated different SNP sets according to variable frequency cutoffs. By applying the PRSs to the validation set, we identified the optimal SNP set, which was then applied to a Greek nonalcoholic fatty liver disease (NAFLD) study. Results: Data from 3553 UK Biobank participants yielded 49 different SNP sets. After calculating the PRS on the validation set for every SNP set, an optimal PRS with 75 SNPs was selected (incremental R² = 0.025, p-value = 0.00145). Interestingly, 43 SNPs were successfully mapped to MASLD-related known genes. The selected PRS could predict traits, like LDL cholesterol and diastolic blood pressure in the UK Biobank, as also disease outcome in the Greek NAFLD study. Conclusions: Our findings provide strong evidence that PRS is a powerful prediction model for MASLD, while it can also be applied on populations of different ethnicity. Full article

Background: Dietary intake of polyunsaturated fatty acids (PUFA) plays a significant role in the onset and progression of neurodegenerative diseases. Since the neuroprotective effects of n-3 PUFA have been widely validated, the role of n-6 PUFA remains debated, with their underlying mechanisms still not fully understood. Methods: In this study, 169,295 participants from the UK Biobank were included to analyze the associations between dietary n-6 PUFA intake and neurodegenerative diseases using Cox regression models with full adjustments for potential confounders. In addition, multiple linear regression models were utilized to estimate the impact of n-6 PUFA intake on brain imaging phenotypes. Results: Results indicated that low dietary n-6 PUFA intake was associated with increased risks of incident dementia (hazard ratio [95% confidence interval] = 1.30 [1.13, 1.49]), Parkinson’s disease (1.42 [1.16, 1.74]), and multiple sclerosis (1.65 [1.03, 2.65]). Moreover, the low intake was linked to diminished volumes of various brain structures, including the hippocampus (β [95% confidence interval] = −0.061 [−0.098, −0.025]), thalamus (−0.071 [−0.105, −0.037]), and others. White matter integrity was also found to be compromised in individuals with low n-6 PUFA intake. Conclusions: These findings enhanced our understanding of how dietary n-6 PUFA intake might affect neurological health, thereby providing epidemiological evidence for future clinical and public health interventions. Full article

Background: Stimulated Raman histology (SRH) is a label-free optical imaging method for rapid intraoperative analysis of fresh tissue samples. Analysis of SRH images using Convolutional Neural Networks (CNN) has shown promising results for predicting the main histopathological classes of neurooncological tumors. Due to the relatively low number of rare tumor representations in CNN training datasets, a valid prediction of rarer entities remains limited. To develop new reliable analysis tools, larger datasets and greater tumor variety are crucial. One way to accomplish this is through research biobanks storing frozen tumor tissue samples. However, there is currently no data available regarding the pertinency of previously frozen tissue samples for SRH analysis. The aim of this study was to assess image quality and perform a comparative reliability analysis of artificial intelligence-based tumor classification using SRH in fresh and frozen tissue samples. Methods: In a monocentric prospective study, tissue samples from 25 patients undergoing brain tumor resection were obtained. SRH was acquired in fresh and defrosted samples of the same specimen after varying storage durations at −80 °C. Image quality was rated by an experienced neuropathologist, and prediction of histopathological diagnosis was performed using two established CNNs. Results: The image quality of SRH in fresh and defrosted tissue samples was high, with a mean image quality score of 1.96 (range 1–5) for both groups. CNN analysis showed high internal consistency for histo-(Cα 0.95) and molecular (Cα 0.83) pathological tumor classification. The results were confirmed using a dataset with samples from the local tumor biobank (Cα 0.91 and 0.53). Conclusions: Our results showed that SRH appears comparably reliable in fresh and frozen tissue samples, enabling the integration of tumor biobank specimens to potentially improve the diagnostic range and reliability of CNN prediction tools. Full article

Background: Polyunsaturated fatty acids (PUFAs) are promising nutrients for the prevention and management of psychiatric disorders. Both animal experiments and cohort studies have demonstrated the antidepressant effects of PUFAs, especially omega-3 PUFAs. However, inconsistent reports about specific types of PUFAs, such as the omega-3 and omega-6 PUFAs, still exist. Objectives: To assess the effects of specific PUFAs on mental disorders and related symptoms and explore the potential mechanisms involving white matter microstructure. Methods: Leveraging 102,252 residents from the UK Biobank, the effects of five PUFA measures on depressive disorder and anxiety disorder were explored through Cox regression models with full adjustment for possible confounders. Furthermore, the effects on related psychiatric symptoms and brain white matter microstructures were also estimated using logistic regression models and multiple linear regression models, respectively. Results: In this study, plasma levels of five PUFAs measured in quartile 4 were associated with lower risks of incident depressive disorder compared with the lowest quartile, with hazard ratios of 0.80 [95% confidence interval] = [0.71, 0.90] for total PUFAs, 0.86 [0.76, 0.97] for omega-3 PUFAs, 0.80 [0.71, 0.91] for docosahexaenoic acid, 0.79 [0.70, 0.89] for omega-6 PUFAs, and 0.77 [0.69, 0.87] for linoleic acid. Similar associations were observed between PUFAs and the incident risk of anxiety disorder. In addition, high plasma PUFA levels were also related to lower risks of occurrence of several adverse psychological symptoms, especially omega-3 PUFAs and DHA. Among the included participants, 8780 individuals with brain imaging information were included in further neuroimaging analyses, and significant associations with white matter microstructures were observed. Conclusions: Thus, this study provides population-based evidence to support the value of interventions to target PUFAs (specifically omega-3 PUFAs) for the prevention and improvement of mental health. Full article

Metabolic dysfunction-associated steatotic liver disease (MASLD) is often asymptomatic and underdiagnosed; consequently, there is a demand for simple, non-invasive diagnostic tools. In this study, we developed a method to quantify liver steatosis based on miRNAs, present in liver and serum, that correlate with liver fat. The miRNAs were analyzed by miRNAseq in liver samples from two cohorts of patients with a precise quantification of liver steatosis. Common miRNAs showing correlation with liver steatosis were validated by RT-qPCR in paired liver and serum samples. Multivariate models were built using partial least squares (PLS) regression to predict the percentage of liver steatosis from serum miRNA levels. Leave-one-out cross validation and external validation were used for model selection and to estimate predictive performance. The miRNAseq results disclosed (a) 144 miRNAs correlating with triglycerides in a set of liver biobank samples (n = 20); and (b) 124 and 102 miRNAs correlating with steatosis by biopsy digital image and MRI analyses, respectively, in liver samples from morbidly obese patients (n = 24). However, only 35 miRNAs were common in both sets of samples. RT-qPCR allowed to validate the correlation of 10 miRNAs in paired liver and serum samples. The development of PLS models to quantitatively predict steatosis demonstrated that the combination of serum miR-145-3p, 122-5p, 143-3p, 500a-5p, and 182-5p provided the lowest root mean square error of cross validation (RMSECV = 1.1, p-value = 0.005). External validation of this model with a cohort of mixed MASLD patients (n = 25) showed a root mean squared error of prediction (RMSEP) of 5.3. In conclusion, it is possible to predict the percentage of hepatic steatosis with a low error rate by quantifying the serum level of five miRNAs using a cost-effective and easy-to-implement RT-qPCR method. Full article

Background/Objectives: Human brain aging is a complex process that affects various aspects of brain function and structure, increasing susceptibility to neurological and psychiatric disorders. A number of nongenetic (e.g., environmental and lifestyle) and genetic risk factors are found to contribute to the varying rates at which the brain ages among individuals. Methods: In this paper, we conducted both an exposome-wide association study (XWAS) and a genome-wide association study (GWAS) on white matter brain aging in the UK Biobank, revealing the multifactorial nature of brain aging. We applied a machine learning algorithm and leveraged fractional anisotropy tract measurements from diffusion tensor imaging data to predict the white matter brain age gap (BAG) and treated it as the marker of brain aging. For XWAS, we included 107 variables encompassing five major categories of modifiable exposures that potentially impact brain aging and performed both univariate and multivariate analysis to select the final set of nongenetic risk factors. Results: We found current tobacco smoking, dietary habits including oily fish, beef, lamb, cereal, and coffee intake, length of mobile phone use, use of UV protection, and frequency of solarium/sunlamp use were associated with the BAG. In genetic analysis, we identified several SNPs on chromosome 3 mapped to genes IP6K1, GMNC, OSTN, and SLC25A20 significantly associated with the BAG, showing the high heritability and polygenic architecture of human brain aging. Conclusions: The critical nongenetic and genetic risk factors identified in our study provide insights into the causal relationship between white matter brain aging and neurodegenerative diseases. Full article

As medical imaging technologies advance, these tools are playing a more and more important role in assisting clinical disease diagnosis. The fusion of biomedical imaging and multi-modal information is profound, as it significantly enhances diagnostic precision and comprehensiveness. Integrating multi-organ imaging with genomic information can significantly enhance the accuracy of disease prediction because many diseases involve both environmental and genetic determinants. In the present study, we focused on the fusion of imaging-derived phenotypes (IDPs) and polygenic risk score (PRS) of diseases from different organs including the brain, heart, lung, liver, spleen, pancreas, and kidney for the prediction of the occurrence of nine common diseases, namely atrial fibrillation, heart failure (HF), hypertension, myocardial infarction, asthma, type 2 diabetes, chronic kidney disease, coronary artery disease (CAD), and chronic obstructive pulmonary disease, in the UK Biobank (UKBB) dataset. For each disease, three prediction models were developed utilizing imaging features, genomic data, and a fusion of both, respectively, and their performances were compared. The results indicated that for seven diseases, the model integrating both imaging and genomic data achieved superior predictive performance compared to models that used only imaging features or only genomic data. For instance, the Area Under Curve (AUC) of HF risk prediction was increased from 0.68 ± 0.15 to 0.79 ± 0.12, and the AUC of CAD diagnosis was increased from 0.76 ± 0.05 to 0.81 ± 0.06. Full article

Low-grade inflammation (LGI) mainly acted as the mediator of the association of obesity and inflammatory diet with numerous chronic diseases, including neuropsychiatric diseases. However, the evidence about the effect of LGI on brain structure is limited but important, especially in the context of accelerating aging. This study was then designed to close the gap, and we leveraged a total of 37,699 participants from the UK Biobank and utilized inflammation score (INFLA-score) to measure LGI. We built the longitudinal relationships of INFLA-score with brain imaging phenotypes using multiple linear regression models. We further analyzed the interactive effects of specific covariates. The results showed high level inflammation reduced the volumes of the subcortex and cortex, especially the globus pallidus (β [95% confidence interval] = −0.062 [−0.083, −0.041]), thalamus (−0.053 [−0.073, −0.033]), insula (−0.052 [−0.072, −0.032]), superior temporal gyrus (−0.049 [−0.069, −0.028]), lateral orbitofrontal cortex (−0.047 [−0.068, −0.027]), and others. Most significant effects were observed among urban residents. Furthermore, males and individuals with physical frailty were susceptive to the associations. The study provided potential insights into pathological changes during disease progression and might aid in the development of preventive and control targets in an age-friendly city to promote great health and well-being for sustainable development goals. Full article