Convolutional neural networks (CNNs) have recently been used in biomedical imaging applications with great success. In this paper, we investigated the classification performance of CNN models on diffusion weighted imaging (DWI) streamlines defined by functional MRI (fMRI) and electrical stimulation mapping (ESM). To learn a set of discriminative and interpretable features from the extremely unbalanced dataset, we evaluated different CNN architectures with multiple loss functions (e.g., focal loss and center loss) and a soft attention mechanism and compared our models with current state-of-the-art methods. Through extensive experiments on streamlines collected from 70 healthy children and 70 children with focal epilepsy, we demonstrated that our deep CNN model with focal and central losses and soft attention outperforms all existing models in the literature and provides clinically acceptable accuracy (73%-100%) for the objective detection of functionally important white matter pathways, including ESM determined eloquent areas such as primary motors, aphasia, speech arrest, auditory, and visual functions. The findings of this paper encourage further investigations to determine if DWI-CNN analysis can serve as a noninvasive diagnostic tool during pediatric presurgical planning by estimating not only the location of essential cortices at the gyral level but also the underlying fibers connecting these cortical areas to minimize or predict postsurgical functional deficits. This paper translates an advanced CNN model to clinical practice in the pediatric population where currently available approaches (e.g., ESM and fMRI) are suboptimal. The implementation will be released at https://github.com/HaotianMXu/Brain-fiber-classification-using-CNNs.

To assess, using fractional anisotropy (FA) analysis, alterations of brain network connectivity in adults with obstructive sleep apnea (OSA). Abnormal networks could mediate clinical functional deficits and reflect brain tissue injury. Structural brain networks were constructed using diffusion tensor imaging (DTI) from 165 healthy (age 57.99 ± 6.02 years, male 27.9%) and 135 OSA participants (age 59.01 ± 5.91 years, male 28.9%) and global network properties (strength, global efficiency, and local efficiency) and regional efficiency were compared between groups. We examined MRI biomarkers of brain tissue injury using FA analysis and its effect on the network properties. Differences between groups of interest were noted in global network properties (p-value < 0.05, corrected), and regional efficiency (p-value < 0.05, corrected) in the left middle cingulate and paracingulate gyri, right posterior cingulate gyrus, and amygdala. In FA analysis, OSA participants showed lower FA values in white matter (WM) of the right transverse temporal, anterior cingulate and paracingulate gyri, and left postcentral, middle frontal and medial frontal gyri, and the putamen. After culling fiber tracts through WM which showed significant differences in FA, we observed no group difference in network properties. Changes in WM integrity and structural connectivity are present in OSA participants. We found that the integrity of WM affected brain network properties. Brain network analysis may improve understanding of neurocognitive deficits in OSA, enable longitudinal tracking, and provides explanations for specific symptoms and recovery kinetics.

Objective: To investigate the clinical utility of deep convolutional neural network (DCNN) tract classification as a new imaging tool in the preoperative evaluation of children with focal epilepsy (FE). Methods: A DCNN tract classification deeply learned spatial trajectories of DWI white matter pathways linking electrical stimulation mapping (ESM) findings from 89 children with FE, and then automatically identified white matter pathways associated with eloquent functions (i.e., primary motor, language, and vision). Clinical utility was examined by 1) measuring the nearest distance between DCNN-determined pathways and ESM, 2) evaluating the effectiveness of DCNN-determined pathways to optimize surgical margins via Kalman filter analysis, and 3) evaluating how accurately changes in DCNN-determined language pathway volume can predict changes in language ability via canonical correlation analysis. Results: DCNN tract classification outperformed other existing methods, achieving an excellent accuracy of 98% while non-invasively detecting eloquent areas within the spatial resolution of ESM (i.e., 1 cm). The Kalman filter analysis found that the preservation of brain areas within a surgical margin determined by DCNN tract classification predicted lack of postoperative deficit with a high accuracy of 92%. Postoperative change of DCNN-determined language pathway volume showed a significant correlation with postoperative changes in language ability (R = 0.7, p $<; $ 0.001). Conclusion: Our findings demonstrate that postoperative functional deficits substantially differ according to the extent of resected white matter, and that DCNN tract classification may offer key translational information by identifying these pathways in pediatric epilepsy surgery. Significance: DCNN tract classification may be an effective tool to improve surgical outcome of children with FE.

Background: Obstructive sleep apnoea (OSA) is associated with increased risk of type 2 diabetes. However, results from large population-based prospective cohort studies are rare. The main aim of the present study was to investigate the relative risk of 8-year incident type 2 diabetes in relation to OSA severity in a prospective cohort study of middle-aged and older adults.Methods: A total of 2918 participants (mean age 59 years) of the Korean Genome and Epidemiology Study (KoGES), who underwent home-based overnight polysomnography at baseline examination between 2011 and 2014, were followed up 4-yearly between 2015–2018 and 2019–2021. A total of 1697 participants were present in both follow-ups. After excluding participants who had diabetes at baseline (n=481), a total of 1216 participants were eligible for the analyses.Results: OSA at baseline was categorised by apnoea–hypopnoea index levels as non-OSA (0–4.9 events·h⁻¹), mild OSA (5.0–14.9 events·h⁻¹) and moderate–severe OSA (≥15.0 events·h⁻¹). Incident type 2 diabetes was identified at each follow-up. Compared with non-OSA, participants with moderate–severe OSA had 1.5 times higher risk of developing type 2 diabetes at the end of the 8-year follow-up after adjusting for potential covariates (relative risk 1.50, 95% CI 1.02–2.21). In the same analytical models for 4-year relative risk of incident type 2 diabetes, none of the OSA groups were at significantly higher risk compared with the non-OSA group.Conclusion: Moderate–severe OSA, a modifiable risk factor, poses a higher risk of incident type 2 diabetes compared with non-OSA over an 8-year period in general middle-aged and older adults.

Intermittent hypoxia and sleep fragmentation due to obstructive sleep apnea (OSA) may contribute to oxidative tissue damage and apoptotic neuronal cell death, inflammation, and intracellular edema in the brain. We examined whether OSA in overweight and obese adolescent children is associated with cortical thickness and hippocampal structure compared to overweight and obese controls and whether OSA severity is associated with measures of brain integrity. We calculated cortical thickness and hippocampal subfield volumes from T1-weighted images of 45 controls (age 15.43 ± 1.73 years, 21 male) and 53 adolescent children with OSA (age 15.26 ± 1.63 years, 32 male) to investigate the association of childhood OSA with the alteration of cortical structure and hippocampal subfield structural changes. In addition, we investigated the correlation between OSA severity and cortical thickness or hippocampal subfield volume using Pearson’s correlation analysis. We found cortical thinning in the right superior parietal area of adolescent children with OSA (cluster size 32.29 mm², cluster-wise corrected p-value = .030) that was negatively correlated with apnea-hypopnea index (AHI) (R=−0.27, p-value = .009) and arousal index (R=−0.25, p-value = .014). In addition, the volume of the right subiculum-head area of the hippocampus of adolescent children with OSA was larger than controls (0.19 ± 0.02 ml vs. 0.18 ± 0.02 ml, β = 13.79, false discovery rate corrected p-value = .044), and it was positively correlated with AHI (R = 0.23, p-value = .026) and arousal index (R = 0.31, p-value = .002). Our findings provide evidence for OSA-associated brain structure alterations in adolescent children prior to the onset of treatment that likely have important implications for timely intervention and continued monitoring of health outcomes.

Obstructive sleep apnea (OSA) is characterized by recurrent airway collapse during sleep, resulting in intermittent hypoxia and sleep fragmentation that may contribute to alternations in brain structure and function. We hypothesized that OSA in children reorganizes and alters cortical structure, which can cause changes in cortical thickness correlation between brain regions across subjects. We constructed cortical structure networks based on cortical thickness measurements from 41 controls (age 15.54 ± 1.66 years, male 19) and 50 children with OSA (age 15.32 ± 1.65 years, male 29). The global (clustering coefficient [CC], path length, and small-worldness) and regional (nodal betweenness centrality, NBC) network properties and hub region distributions were examined between groups. We found increased CCs in OSA compared to controls across a wide range of network densities (p-value < .05) and lower NBC area under the curve in left caudal anterior cingulate, left caudal middle frontal, left fusiform, left transverse temporal, right pars opercularis, and right precentral gyri (p-value < .05). In addition, while most of the hub regions were the same between groups, the OSA group had fewer hub regions and a different hub distribution compared to controls. Our findings suggest that children with OSA exhibit altered global and regional network characteristics compared to healthy controls. Our approach to the investigation of cortical structure in children with OSA could prove useful in understanding the etiology of OSA-related brain functional disorders.