Background: Maintaining cognitive function is essential for healthy aging and to function autonomously within society. White matter lesions (WMLs) are associated with reduced cognitive function in older adults. However, whether their anatomical location moderates these associations is not well-established. This review systematically evaluates peer-reviewed evidence on the role of anatomical location in the association between WMLs and cognitive function. Methods: In accordance with the preferred reporting items for systematic reviews and meta-analysis (PRISMA) statement, databases of EMBASE, PUBMED, MEDLINE, and CINAHL, and reference lists of selected papers were searched. We limited our search results to adults aged 60 years and older, and studies published in the English language from 2000 to 2011. Studies that investigated the association between cognitive function and WML location were included. Two independent reviewers extracted: 1) study characteristics including sample size, sample characteristic, and study design; 2) WML outcomes including WML location, WML quantification method (scoring or volume measurement), strength of the MRI magnet in Tesla, and MRI sequence used for WML detection; and 3) cognitive function outcomes including cognitive tests for two cognitive domains of memory and executive function/processing speed. Results: Of the 14 studies included, seven compared the association of subcortical versus periventricular WMLs with cognitive function. Seven other studies investigated the association between WMLs in specific brain regions (e.g., frontal, parietal lobes) and cognitive function. Overall, the results show that a greater number of studies have found an association between periventricular WMLs and executive function/processing speed, than subcortical WMLs. However, whether WMLs in different brain regions have a differential effect on cognitive function remains unclear. Conclusions: Evidence suggests that periventricular WMLs may have a significant negative impact on cognitive abilities of older adults. This finding may be influenced by study heterogeneity in: 1) MRI sequences, WML quantification methods, and neuropsychological batteries; 2) modifying effect of cardiovascular risk factors; and 3) quality of studies and lack of sample size calculation.

Although deep learning-based diabetic retinopathy (DR) classification methods typically benefit from well-designed architectures of convolutional neural networks, the training setting also has a non-negligible impact on prediction performance. The training setting includes various interdependent components, such as an objective function, a data sampling strategy, and a data augmentation approach. To identify the key components in a standard deep learning framework (ResNet-50) for DR grading, we systematically analyze the impact of several major components. Extensive experiments are conducted on a publicly available dataset EyePACS. We demonstrate that (1) the DR grading framework is sensitive to input resolution, objective function, and composition of data augmentation; (2) using mean square error as the loss function can effectively improve the performance with respect to a task-specific evaluation metric, namely the quadratically weighted Kappa; (3) utilizing eye pairs boosts the performance of DR grading and; (4) using data resampling to address the problem of imbalanced data distribution in EyePACS hurts the performance. Based on these observations and an optimal combination of the investigated components, our framework, without any specialized network design, achieves a state-of-the-art result (0.8631 for Kappa) on the EyePACS test set (a total of 42,670 fundus images) with only image-level labels. We also examine the proposed training practices on other fundus datasets and other network architectures to evaluate their generalizability. Our codes and pre-trained model are available online.

Self-supervised Learning (SSL) has been widely applied to learn image representations through exploiting unlabeled images. However, it has not been fully explored in the medical image analysis field. In this work, Saliency-guided Self-Supervised image Transformer (SSiT) is proposed for Diabetic Retinopathy (DR) grading from fundus images. We novelly introduce saliency maps into SSL, with a goal of guiding self-supervised pre-training with domain-specific prior knowledge. Specifically, two saliency-guided learning tasks are employed in SSiT: (1) Saliency-guided contrastive learning is conducted based on the momentum contrast, wherein fundus images' saliency maps are utilized to remove trivial patches from the input sequences of the momentum-updated key encoder. Thus, the key encoder is constrained to provide target representations focusing on salient regions, guiding the query encoder to capture salient features. (2) The query encoder is trained to predict the saliency segmentation, encouraging the preservation of fine-grained information in the learned representations. To assess our proposed method, four publicly-accessible fundus image datasets are adopted. One dataset is employed for pre-training, while the three others are used to evaluate the pre-trained models' performance on downstream DR grading. The proposed SSiT significantly outperforms other representative state-of-the-art SSL methods on all downstream datasets and under various evaluation settings. For example, SSiT achieves a Kappa score of 81.88% on the DDR dataset under fine-tuning evaluation, outperforming all other ViT-based SSL methods by at least 9.48%.

Background: The natural history of autonomic alterations following catheter ablation of drug‐refractory paroxysmal atrial fibrillation is poorly defined, largely because of the historical reliance on non‐invasive intermittent rhythm monitoring for outcome ascertainment.Methods and Results: The study included 346 patients with drug‐refractory paroxysmal atrial fibrillation undergoing pulmonary vein isolation using contemporary advanced‐generation ablation technologies. All patients underwent insertion of a Reveal LINQ (Medtronic) implantable cardiac monitor before ablation. The implantable cardiac monitor continuously recorded physical activity, heart rate variability (measured as the SD of the average normal‐to‐normal), daytime heart rate, and nighttime heart rate. Longitudinal autonomic data in the 2‐month period leading up to the date of ablation were compared with the period from 91 to 365 days following ablation. Following ablation there was a significant decrease in SD of the average normal‐to‐normal (mean difference versus baseline of 19.3 ms; range, 12.9–25.7;P<0.0001), and significant increases in daytime and nighttime heart rates (mean difference versus baseline of 9.6 bpm; range, 7.4–11.8;P<0.0001, and 7.4 bpm; range, 5.4–9.3;P<0.0001, respectively). Patients free of arrhythmia recurrence had significantly faster daytime (11±11 versus 8±12 bpm,P=0.001) and nighttime heart rates (8±9 versus 6±8 bpm,P=0.049), but no difference in SD of the average normal‐to‐normal (P=0.09) compared with those with atrial fibrillation recurrence. Ablation technology and cryoablation duration did not influence these autonomic nervous system effects.Conclusions: Pulmonary vein isolation results in significant sustained changes in the heart rate parameters related to autonomic function. These changes are correlated with procedural outcome and are independent of the ablation technology used.Registration: URL:https://www.clinicaltrials.gov; Unique identifier: NCT01913522.

White matter hyperintensities (WMH) are a prominent feature of cerebral small vessel disease and are associated with cognitive impairment. These deficits in cognition may be caused by the disruption of large-scale functional networks due to the presence of WMHs. However, knowledge regarding the relevance of these lesions on functional networks remains inconclusive. These inconsistencies may derive from issues with interpreting functional imaging data from clinical populations. Lesion network mapping is a technique that allows the overlaying of lesions from a patient population to the functional connectivity of a human connectome derived from healthy adults. This allows researchers to identify functional networks that would be disrupted in a healthy population should the WMHs seen in cerebral small vessel disease be present. We hypothesized that the extent to which these functional networks are disrupted by WMHs is associated with cognitive performance in older adults with cerebral small vessel disease. This cross-sectional study combined baseline data from four studies to create a total sample of 164 older adults (aged ≥55) from metropolitan Vancouver with cerebral small vessel disease. Using lesion network mapping, we assessed the percentage overlap between voxels functionally connected with both the WMHs (lesion network) and five common functional networks: (1) visual; (2) dorsal attention; (3) ventral attention; (4) sensorimotor; and (5) frontoparietal. Cognition was assessed using: (1) Montreal Cognitive Assessment (MoCA); (2) Stroop Colour Word Test (3-2); (3) Trail Making Tests (Part B-A); and (4) Digit Symbol Substitution Test. A One-Way ANOVA and Tukey post-hoc tests were performed to identify the functional networks with greatest percentage overlap with the lesion network. Partial correlations controlling for age were used to analyse whether the extent of the overlap between the lesion and functional networks was associated with poorer cognition. The visual, ventral attention, and frontoparietal networks had significantly greater overlap with the lesion network. After controlling for multiple comparisons, level of lesion network overlap with both the sensorimotor network (p<.001) and ventral attention network (p <. 001) was significantly correlated with MoCA score. Thus, the greater the disruption to the sensorimotor and ventral attention networks, the poorer the global cognition. Our results reveal that the visual, ventral attention, and frontoparietal networks are most vulnerable to disruptions stemming from WMHs. Additionally, we identified that disruption to the sensorimotor and ventral attention networks, as a result of WMHs, may underlie deficits in global cognition in older adults with cerebral small vessel disease.

BACKGROUND AND PURPOSE: Dose-dependent association between hyperintensity in deep brain structures on unenhanced T1WIs and gadolinium-based contrast agent administrations has been demonstrated with subsequent histopathological confirmation of gadolinium deposition. Our aim was to determine whether greater exposure to linear gadolinium-based contrast agent administration is associated with higher signal intensity in deep brain structures on unenhanced T1-weighted MR imaging. Secondary objective was to compare signal intensity differences between ionic and nonionic linear gadolinium-based contrast agents. MATERIALS AND METHODS: Subjects with secondary-progressive MS originally enrolled in a multicenter clinical trial were studied retrospectively. Eighty subjects (high-exposure cohort) received 9 linear gadolinium-based contrast agent administrations (30 nonionic/50 ionic) between week −4 and year 1 and a tenth administration by year 2. One hundred fifteen subjects (low-exposure cohort) received 2 administrations (40 nonionic/75 ionic) between week −4 and year 1 and a third administration by year 2. Signal intensities were measured on unenhanced T1WIs by placing sample-points on the dentate nucleus, globus pallidus, caudate, thalamus, pons, and white matter, and they were normalized using the following ratios: dentate/pons, globus pallidus/white matter, caudate/white matter, and thalamus/white matter. RESULTS: Between week −4 and year 1, subjects in the high-exposure cohort showed increased signal intensity ratios in all regions (P < .01), while the low-exposure cohort showed only an increase in the dentate nucleus (P = .003). Between years 1 and 2, when both cohorts received only 1 additional gadolinium-based contrast agent, no significant changes were observed. In the high-exposure cohort, significantly higher changes in signal intensity ratios were observed in subjects receiving linear nonionic than in those receiving linear ionic gadolinium-based contrast agents. CONCLUSIONS: Hyperintensity in deep brain structures from gadolinium deposition is related to the number of doses and the type of linear gadolinium-based contrast agent (nonionic greater than ionic) administration.