Magnetic Resonance Imaging Data Research Articles

Aberrant connectivity in the hippocampus, bilateral insula and temporal poles precedes treatment resistance in first-episode psychosis: a prospective resting-state functional magnetic resonance imaging study with connectivity concordance mapping.

Functional connectivity resting-state functional magnetic resonance imaging has been proposed to predict antipsychotic treatment response in schizophrenia. However, only a few prospective studies have examined baseline resting-state functional magnetic resonance imaging data in drug-naïve first-episode schizophrenia patients with regard to subsequent treatment response. Data-driven approaches to conceptualize and measure functional connectivity patterns vary broadly, and model-free, voxel-wise, whole-brain analysis techniques are scarce. Here, we apply such a method, called connectivity concordance mapping to resting-state functional magnetic resonance imaging data acquired from an Asian sample (n = 60) with first-episode psychosis, prior to pharmaceutical treatment. Using a longitudinal design, 12 months after the resting-state functional magnetic resonance imaging, we measured and classified patients into two groups based on psychometric testing: treatment responsive and treatment resistant. Next, we compared the two groups' connectivity concordance maps that were derived from the resting-state functional magnetic resonance imaging data at baseline. We have identified consistently higher functional connectivity in the treatment-resistant group in a network including the left hippocampus, bilateral insula and temporal poles. These data-driven novel findings can help researchers to consider new regions of interest and facilitate biomarker development in order to identify treatment-resistant schizophrenia patients early, in advance of treatment and at the time of their first psychotic episode.

Lifestyle management and brain MRI metrics in female Australian adults living with multiple sclerosis: a feasibility and acceptability study

BackgroundLimited studies of multiple sclerosis (MS) exist whereby magnetic resonance imaging (MRI) of the brain with consistent imaging protocols occurs at the same time points as collection of healthy lifestyle measures. The aim of this study was to test the feasibility, acceptability and preliminary efficacy of acquiring MRI data as an objective, diagnostic and prognostic marker of MS, at the same time point as brain-healthy lifestyle measures including diet.MethodsParticipants living with relapsing remitting MS partook in one structural MRI scanning session of the brain, completed two online 24-hour dietary recalls and demographic and self-reported lifestyle questionnaires (e.g. self-reported disability, comorbidities, physical activity, smoking status, body mass index (BMI), stress). Measures of central tenancy and level of dispersion were calculated for feasibility and acceptability of the research protocols. Lesion count was determined by one radiologist and volumetric analyses by a data analysis pipeline based on FreeSurfer software suite. Correlations between white matter lesion count, whole brain volume analyses and lifestyle measures were assessed using Spearman’s rank-order correlation coefficient.ResultsThirteen female participants were included in the study: eligibility rate 90.6% (29/32), recruitment rate 46.9% (15/32) and compliance rate 87% (13/15). The mean time to complete all required tasks, including MRI acquisition was 115.86 minutes (\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$($$\\end{document}± 23.04), over 4 days. Conversion to usual dietary intake was limited by the small sample. There was one strong, negative correlation between BMI and brain volume (rs = −0.643, p = 0.018) and one strong, positive correlation between physical activity and brain volume (rs = 0.670, p = 0.012) that were both statistically significant.ConclusionsAcquiring MRI brain scans at the same time point as lifestyle profiles in adults with MS is both feasible and accepted among adult females living with MS. Quantification of volumetric MRI data support further investigations using semi-automated pipelines among people living with MS, with pre-processing steps identified to increase automated feasibility. This protocol may be used to determine relationships between elements of a brain-healthy lifestyle, including dietary intake, and measures of disease burden and brain health, as assessed by T1-weighted and T2-weighted lesion count and whole brain volume, in an adequately powered sample.Trial registrationThe study protocol was retrospectively registered in the Australia New Zealand Clinical Trials Registry (ACTRN12624000296538).

Constructing personalized characterizations of structural brain aberrations in patients with dementia using explainable artificial intelligence

Deep learning approaches for clinical predictions based on magnetic resonance imaging data have shown great promise as a translational technology for diagnosis and prognosis in neurological disorders, but its clinical impact has been limited. This is partially attributed to the opaqueness of deep learning models, causing insufficient understanding of what underlies their decisions. To overcome this, we trained convolutional neural networks on structural brain scans to differentiate dementia patients from healthy controls, and applied layerwise relevance propagation to procure individual-level explanations of the model predictions. Through extensive validations we demonstrate that deviations recognized by the model corroborate existing knowledge of structural brain aberrations in dementia. By employing the explainable dementia classifier in a longitudinal dataset of patients with mild cognitive impairment, we show that the spatially rich explanations complement the model prediction when forecasting transition to dementia and help characterize the biological manifestation of disease in the individual brain. Overall, our work exemplifies the clinical potential of explainable artificial intelligence in precision medicine.

In vivo neuropil density from anatomical MRI and machine learning.

Brain energy budgets specify metabolic costs emerging from underlying mechanisms of cellular and synaptic activities. While current bottom-up energy budgets use prototypical values of cellular density and synaptic density, predicting metabolism from a person's individualized neuropil density would be ideal. We hypothesize that in vivo neuropil density can be derived from magnetic resonance imaging (MRI) data, consisting of longitudinal relaxation (T1) MRI for gray/white matter distinction and diffusion MRI for tissue cellularity (apparent diffusion coefficient, ADC) and axon directionality (fractional anisotropy, FA). We present a machine learning algorithm that predicts neuropil density from in vivo MRI scans, where ex vivo Merker staining and in vivo synaptic vesicle glycoprotein 2A Positron Emission Tomography (SV2A-PET) images were reference standards for cellular and synaptic density, respectively. We used Gaussian-smoothed T1/ADC/FA data from 10 healthy subjects to train an artificial neural network, subsequently used to predict cellular and synaptic density for 54 test subjects. While excellent histogram overlaps were observed both for synaptic density (0.93) and cellular density (0.85) maps across all subjects, the lower spatial correlations both for synaptic density (0.89) and cellular density (0.58) maps are suggestive of individualized predictions. This proof-of-concept artificial neural network may pave the way for individualized energy atlas prediction, enabling microscopic interpretations of functional neuroimaging data.

Development of neonatal connectome dynamics and its prediction for cognitive and language outcomes at age 2.

The functional brain connectome is highly dynamic over time. However, how brain connectome dynamics evolves during the third trimester of pregnancy and is associated with later cognitive growth remains unknown. Here, we use resting-state functional Magnetic Resonance Imaging (MRI) data from 39 newborns aged 32 to 42 postmenstrual weeks to investigate the maturation process of connectome dynamics and its role in predicting neurocognitive outcomes at 2years of age. Neonatal brain dynamics is assessed using a multilayer network model. Network dynamics decreases globally but increases in both modularity and diversity with development. Regionally, module switching decreases with development primarily in the lateral precentral gyrus, medial temporal lobe, and subcortical areas, with a higher growth rate in primary regions than in association regions. Support vector regression reveals that neonatal connectome dynamics is predictive of individual cognitive and language abilities at 2 years of age. Our findings highlight network-level neural substrates underlying early cognitive development.

Sports promote brain evolution: a resting-state fMRI study of volleyball athlete.

Long-term skill learning can lead to structure and function changes in the brain. Different sports can trigger neuroplasticity in distinct brain regions. Volleyball, as one of the most popular team sports, heavily relies on individual abilities such as perception and prediction for high-level athletes to excel. However, the specific brain mechanisms that contribute to the superior performance of volleyball athletes compared to non-athletes remain unclear. We conducted a study involving the recruitment of ten female volleyball athletes and ten regular female college students, forming the athlete and novice groups, respectively. Comprehensive behavioral assessments, including Functional Movement Screen and audio-visual reaction time tests, were administered to both groups. Additionally, resting-state magnetic resonance imaging (MRI) data were acquired for both groups. Subsequently, we conducted in-depth analyses, focusing on the amplitude of low-frequency fluctuations (ALFF), regional homogeneity (ReHo), and functional connectivity (FC) in the brain for both the athlete and novice groups. No significant differences were observed in the behavioral data between the two groups. However, the athlete group exhibited noteworthy enhancements in both the ALFF and ReHo within the visual cortex compared to the novice group. Moreover, the functional connectivity between the visual cortex and key brain regions, including the left primary sensory cortex, left supplementary motor cortex, right insula, left superior temporal gyrus, and left inferior parietal lobule, was notably stronger in the athlete group than in the novice group. This study has unveiled the remarkable impact of volleyball athletes on various brain functions related to vision, movement, and cognition. It indicates that volleyball, as a team-based competitive activity, fosters the advancement of visual, cognitive, and motor skills. These findings lend additional support to the early cultivation of sports talents and the comprehensive development of adolescents. Furthermore, they offer fresh perspectives on preventing and treating movement-related disorders. Registration number: ChiCTR2400079602. Date of Registration: January 8, 2024.

Stimulating prefrontal cortex facilitates training transfer by increasing representational overlap.

A recent hypothesis characterizes difficulties in multitasking as being the price humans pay for our ability to generalize learning across tasks. The mitigation of these costs through training has been associated with reduced overlap of constituent task representations within frontal, parietal, and subcortical regions. Transcranial direct current stimulation, which can modulate functional brain activity, has shown promise in generalizing performance gains when combined with multitasking training. However, the relationship between combined transcranial direct current stimulation and training protocols with task-associated representational overlap in the brain remains unexplored. Here, we paired prefrontal cortex transcranial direct current stimulation with multitasking training in 178 individuals and collected functional magnetic resonance imaging data pre- and post-training. We found that 1mA transcranial direct current stimulation applied to the prefrontal cortex paired with multitasking training enhanced training transfer to spatial attention, as assessed via a visual search task. Using machine learning to assess the overlap of neural activity related to the training task in task-relevant brain regions, we found that visual search gains were predicted by changes in classification accuracy in frontal, parietal, and cerebellar regions for participants that received left prefrontal cortex stimulation. These findings demonstrate that prefrontal cortex transcranial direct current stimulation may interact with training-related changes to task representations, facilitating the generalization of learning.

Spatial functional reorganizations can serve as potential biomarkers of post facial palsy synkinesis.

Facial palsy can result in a serious complication known as facial synkinesis, causing both physical and psychological harm to the patients. There is growing evidence that patients with facial synkinesis have brain abnormalities, but the brain mechanisms and underlying imaging biomarkers remain unclear. Here, we employed functional magnetic resonance imaging (fMRI) to investigate brain function in 31 unilateral post facial palsy synkinesis patients and 25 healthy controls during different facial expression movements and at rest. Combining surface-based mass-univariate analysis and multivariate pattern analysis, we identified diffused activation and intrinsic connection patterns in the primary motor cortex and the somatosensory cortex on the patient's affected side. Further, we classified post facial palsy synkinesis patients from healthy subjects with favorable accuracy using the support vector machine based on both task-related and resting-state functional magnetic resonance imaging data. Together, these findings indicate the potential of the identified functional reorganizations to serve as neuroimaging biomarkers for facial synkinesis diagnosis.

Bridging Modalities: A Multimodal Machine Learning Approach for Parkinson’s Disease Diagnosis Using EEG and MRI Data

Parkinson’s disease (PD) is a slowly progressing neurological disorder with symptoms that overlap with those of other conditions, making early detection and accurate diagnosis vital for effective treatment and a patient’s quality of life. Symptoms such as tremors, stiffness, slow movements, and balance issues, along with psychiatric manifestations, are typical of PD. This study introduces a groundbreaking approach to PD diagnosis, utilizing a multimodal machine learning framework that integrates Electroencephalography (EEG) and Magnetic Resonance Imaging (MRI) data. Focusing on the early detection and accurate classification of PD, the proposed research leverages the distinct yet complementary nature of EEG and MRI datasets to enhance diagnostic precision. We employed a robust algorithmic strategy, including LightGBM and machine learning techniques, to analyze the complex patterns inherent in neurological data. The key steps of the proposed research are preprocessing and feature extraction from both EEG and MRI modalities, followed by their fusion using Principal Component Analysis (PCA) for dimensionality reduction. The fused dataset was then analyzed using a LightGBM model and validated through a 10-fold cross-validation process to ensure reliability and stability. The model’s efficacy was further tested on independent datasets, demonstrating its robustness across diverse patient demographics. The obtained results showcased an accuracy of 97.17%, sensitivity of 96.58%, and specificity of 96.82% in PD classification, outperforming traditional multimodal as well as single-modality diagnostic methods. The integration of EEG and MRI data provided a more comprehensive view of the neurophysiological and neuroanatomical changes associated with PD. Additionally, the use of advanced machine learning algorithms allowed for a nuanced analysis, capturing subtle patterns indicative of early PD stages.

Multimodal MRI reveals brainstem connections that sustain wakefulness in human consciousness.

Consciousness is composed of arousal (i.e., wakefulness) and awareness. Substantial progress has been made in mapping the cortical networks that underlie awareness in the human brain, but knowledge about the subcortical networks that sustain arousal in humans is incomplete. Here, we aimed to map the connectivity of a proposed subcortical arousal network that sustains wakefulness in the human brain, analogous to the cortical default mode network (DMN) that has been shown to contribute to awareness. We integrated data from ex vivo diffusion magnetic resonance imaging (MRI) of three human brains, obtained at autopsy from neurologically normal individuals, with immunohistochemical staining of subcortical brain sections. We identified nodes of the proposed default ascending arousal network (dAAN) in the brainstem, hypothalamus, thalamus, and basal forebrain. Deterministic and probabilistic tractography analyses of the ex vivo diffusion MRI data revealed projection, association, and commissural pathways linking dAAN nodes with one another and with DMN nodes. Complementary analyses of in vivo 7-tesla resting-state functional MRI data from the Human Connectome Project identified the dopaminergic ventral tegmental area in the midbrain as a widely connected hub node at the nexus of the subcortical arousal and cortical awareness networks. Our network-based autopsy methods and connectivity data provide a putative neuroanatomic architecture for the integration of arousal and awareness in human consciousness.

Expert-level diagnosis of pediatric posterior fossa tumors via consistency calibration

Accurate diagnosis of pediatric posterior fossa tumors (PFTs) is critical for saving lives; however, the limited number of specialists makes accurate diagnostics scarce. To make the diagnosis of PFTs accurate, automatic, and noninvasive, scholars have proposed employing deep neural networks (DNNs) to predict tumor types using magnetic resonance imaging data. Advanced methods primarily focus on fine-tuning DNNs pre-trained on large-scale datasets of natural images, e.g., ImageNet. However, the existing methods overlook the priors of human experts. Human experts typically recheck whether images predicted as a particular class are similar to those predicted as the same class to ensure prediction consistency. Therefore, the predicted results of an intelligent system should be consistent. Inspired by the rechecking process, we propose a novel learning paradigm called Consistency calibration (Coca). Within the Coca framework, the output predicted by DNNs is guided by two objective functions: (i) the task-specific objective of making the predicted results the same as the groundtruth, and (ii) an auxiliary objective of rechecking the prediction consistency. Coca is developed by defining the inconsistency for each sample by inconsistent risks: the auxiliary risk is small (large), but the task-specific risk is large (small). Building on the inconsistency definition, Coca identifies inconsistencies for each sample using an adversarial attack. Subsequently, these inconsistencies are leveraged to tune DNNs in an adversarial training manner for consistency calibration. To verify the efficacy of Coca, we conduct comprehensive experiments using a large-scale PBT dataset, and the results show that Coca significantly outperforms state-of-the-art methods. Moreover, Coca has improved performance over human experts as demonstrated by expert-level diagnostic performance in real-world PBT scenarios for the first time.

Whole-Body Hypothermia vs Targeted Normothermia for Neonates With Mild Encephalopathy

Although whole-body hypothermia is widely used after mild neonatal hypoxic-ischemic encephalopathy (HIE), safety and efficacy have not been evaluated in randomized clinical trials (RCTs), to our knowledge. To examine the effect of 48 and 72 hours of whole-body hypothermia after mild HIE on cerebral magnetic resonance (MR) biomarkers. This open-label, 3-arm RCT was conducted between October 31, 2019, and April 28, 2023, with masked outcome analysis. Participants were neonates at 6 tertiary neonatal intensive care units in the UK and Italy born at or after 36 weeks' gestation with severe birth acidosis, requiring continued resuscitation, or with an Apgar score less than 6 at 10 minutes after birth and with evidence of mild HIE on modified Sarnat staging. Statistical analysis was per intention to treat. Random allocation to 1 of 3 groups (1:1:1) based on age: neonates younger than 6 hours were randomized to normothermia or 72-hour hypothermia (33.5 °C), and those 6 hours or older and already receiving whole-body hypothermia were randomized to rewarming after 48 or 72 hours of hypothermia. Thalamic N-acetyl aspartate (NAA) concentration (mmol/kg wet weight), assessed by cerebral MR imaging and thalamic spectroscopy between 4 and 7 days after birth using harmonized sequences. Of 225 eligible neonates, 101 were recruited (54 males [53.5%]); 48 (47.5%) were younger than 6 hours and 53 (52.5%) were 6 hours or older at randomization. Mean (SD) gestational age and birth weight were 39.5 (1.1) weeks and 3378 (380) grams in the normothermia group (n = 34), 38.7 (0.5) weeks and 3017 (338) grams in the 48-hour hypothermia group (n = 31), and 39.0 (1.1) weeks and 3293 (252) grams in the 72-hour hypothermia group (n = 36). More neonates in the 48-hour (14 of 31 [45.2%]) and 72-hour (13 of 36 [36.1%]) groups required intubation at birth than in the normothermic group (3 of 34 [8.8%]). Ninety-nine neonates (98.0%) had MR imaging data and 87 (86.1%), NAA data. Injury scores on conventional MR biomarkers were similar across groups. The mean (SD) NAA level in the normothermia group was 10.98 (0.92) mmol/kg wet weight vs 8.36 (1.23) mmol/kg wet weight (mean difference [MD], -2.62 [95% CI, -3.34 to -1.89] mmol/kg wet weight) in the 48-hour and 9.02 (1.79) mmol/kg wet weight (MD, -1.96 [95% CI, -2.66 to -1.26] mmol/kg wet weight) in the 72-hour hypothermia group. Seizures occurred beyond 6 hours after birth in 4 neonates: 1 (2.9%) in the normothermia group, 1 (3.2%) in the 48-hour hypothermia group, and 2 (5.6%) in the 72-hour hypothermia group. In this pilot RCT, whole-body hypothermia did not improve cerebral MR biomarkers after mild HIE, although neonates in the hypothermia groups were sicker at baseline. Safety and efficacy of whole-body hypothermia should be evaluated in RCTs. ClinicalTrials.gov Identifier: NCT03409770.

Neural mechanism of non-adaptive cognitive emotion regulation in patients with non-suicidal self-injury

BackgroundThe incidence of non-suicidal self-injury (NSSI) has been on the rise in recent years. Studies have shown that people with NSSI have difficulties in emotion regulation and cognitive control. In addition, some studies have investigated the cognitive emotion regulation of people with NSSI which found that they have difficulties in cognitive emotion regulation, but there was a lack of research on cognitive emotion regulation strategies and related neural mechanisms. MethodsThis study included 117 people with NSSI (age = 19.47 ± 5.13, male = 17) and 84 non-NSSI participants (age = 19.86 ± 4.14, male = 16). People with NSSI met the DSM-5 diagnostic criteria, and non-NSSI participants had no mental or physical disorders. The study collected all participants' data of Cognitive Emotion Regulation Questionnaire (CERQ) and functional magnetic resonance imaging (fMRI) to explore the differences in psychological performance and brain between two groups. Afterwards, Machine learning was used to select the found differential brain regions to obtain the highest correlation regions with NSSI. Then, Allen's Human Brain Atlas database was used to compare with the information on the abnormal brain regions of people with NSSI to find the genetic information related to NSSI. In addition, gene enrichment analysis was carried out to find the related pathways and specific cells that may have differences. ResultsThe differences between NSSI participants and non-NSSI participants were as follows: positive refocusing (t = −4.74, p < 0.01); refocusing on plans (t = −4.11, p < 0.01); positive reappraisal (t = −9.22, p < 0.01); self-blame (t = 6.30, p < 0.01); rumination (t = 3.64, p < 0.01); catastrophizing (t = 9.10, p < 0.01), and blaming others (t = 2.52, p < 0.01), the precentral gyrus (t = 6.04, pFDR < 0.05) and the rolandic operculum (t = −4.57, pFDR < 0.05). Rolandic operculum activity was negatively correlated with blaming others (r = −0.20, p < 0.05). Epigenetic results showed that excitatory neurons (p < 0.01) and inhibitory neurons (p < 0.01) were significant differences in two pathways, “trans-synaptic signaling” (p < −log108) and “modulation of chemical synaptic transmission” (p < −log108) in both cells. ConclusionsPeople with NSSI are more inclined to adopt non-adaptive cognitive emotion regulation strategies. Rolandic operculum is also abnormally active. Abnormal changes in the rolandic operculum of them are associated with non-adaptive cognitive emotion regulation strategies. Changes in the excitatory and inhibitory neurons provide hints to explore the abnormalities of the neurological mechanisms at the cellular level of them.Trial registration number NCT04094623

In Vivo Biomechanical Dynamic Simulation of a Healthy Knee During the Single-leg Lunge and Its In Vitro Experiment Validation

Biomechanical modelling of the knee during motion is a pivotal component in disease treatment, implant designs, and rehabilitation strategies. Historically, dynamic simulations of the knee have been scant. This study uniquely integrates a dual fluoroscopic imaging system (DFIS) to investigate the in vivo dynamic behavior of the meniscus during functional activities using a finite element (FE) model. The model was subsequently validated through in vitro experiments. Motion capture of a single-leg lunge was executed by DFIS. The motion model was reconstructed using 2D-to-3D registration in conjunction with computed tomography (CT) scans. Both CT and magnetic resonance imaging (MRI) data facilitated the development of the knee FE model. In vivo knee displacements and rotations were utilized as driving conditions for the FE model. Moreover, a 3D-printed model, accompanied with digital imaging correlation (DIC), was used to evaluate the accuracy of the FE model. To a better inner view of knees during the DIC analysis, tibia and femur were crafted by transparent resin. The availability of the FE model was guaranteed by the similar strain distribution of the DIC and FE simulation. Subsequent modeling revealed that the compressive stress distribution between the medial and lateral menisci was balanced in the standing posture. As the flexion angle increased, the medial meniscus bore the primary compressive load, with peak stresses occurring between 60 and 80 degrees of flexion. The simulation of a healthy knee provides a critical theoretical foundation for addressing knee pathologies and advancing prosthetic designs.

Deep learning model based on contrast-enhanced MRI for predicting post-surgical survival in patients with hepatocellular carcinoma

ObjectiveTo develop a deep learning model based on contrast-enhanced magnetic resonance imaging (MRI) data to predict post-surgical overall survival (OS) in patients with hepatocellular carcinoma (HCC). MethodsThis bi-center retrospective study included 564 surgically resected patients with HCC and divided them into training (326), testing (143), and external validation (95) cohorts. This study used a three-dimensional convolutional neural network (3D-CNN) ResNet to learn features from the pretreatment MR images (T1WIpre, late arterial phase, and portal venous phase) and got the deep learning score (DL score). Three cox regression models were established separately using the DL score (3D-CNN model), clinical features (clinical model), and a combination of above (combined model). The concordance index (C-index) was used to evaluate model performance. ResultsWe trained a 3D-CNN model to get DL score from samples. The C-index of the 3D-CNN model in predicting 5-year OS for the training, testing, and external validation cohorts were 0.746, 0.714, and 0.698, respectively, and were higher than those of the clinical model, which were 0.675, 0.674, and 0.631, respectively (P = 0.009, P = 0.204, and P = 0.092, respectively). The C-index of the combined model for testing and external validation cohorts was 0.750 and 0.723, respectively, significantly higher than the clinical model (P = 0.017, P = 0.016) and the 3D-CNN model (P = 0.029, P = 0.036). ConclusionsThe combined model integrating the DL score and clinical factors showed a higher predictive value than the clinical and 3D-CNN models and may be more useful in guiding clinical treatment decisions to improve the prognosis of patients with HCC.

Multi-teacher cross-modal distillation with cooperative deep supervision fusion learning for unimodal segmentation

Accurate brain tumor segmentation is a labor-intensive and time-consuming task that requires automation to enhance its efficacy. Recent advanced techniques have shown promising results in segmenting brain tumors; however, their dependency on extensive multimodal magnetic resonance imaging (MRI) data limits their practicality in clinical environments where such data may not be readily available. To address this, we propose a novel multi-teacher cross-modal knowledge distillation framework, which utilizes the privileged multimodal data during training while relying solely on unimodal data for inference. Our framework is tailored to the unimodal segmentation of the T1ce MRI sequence, which is prevalently available in clinical practice and structurally akin to the T1 modality, providing ample information for the segmentation task. Our framework introduces two learning strategies for knowledge distillation (KD): (1) performance-aware response-based KD and (2) cooperative deep supervision fusion learning (CDSFL). The first strategy involves dynamically assigning confidence weights to each teacher model based on its performance, ensuring that the KD is performance-driven, and the CDSFL module augments the learning capabilities of the multi-teacher models by fostering mutual learning. Moreover, the fused information is distilled into the student model to improve its deep supervision. Extensive experiments on BraTS datasets demonstrate that our framework achieves promising unimodal segmentation results on the T1ce and T1 modalities and outperforms previous state-of-the-art methods. Code is available at https://github.com/ami-lab-knut/mtcm_kd.

The effect of head motion on brain age prediction using deep convolutional neural networks

Deep learning can be used effectively to predict participants' age from brain magnetic resonance imaging (MRI) data, and a growing body of evidence suggests that the difference between predicted and chronological age—referred to as brain-predicted age difference (brain-PAD)—is related to various neurological and neuropsychiatric disease states. A crucial aspect of the applicability of brain-PAD as a biomarker of individual brain health is whether and how brain-predicted age is affected by MR image artifacts commonly encountered in clinical settings. To investigate this issue, we trained and validated two different 3D convolutional neural network architectures (CNNs) from scratch and tested the models on a separate dataset consisting of motion-free and motion-corrupted T1-weighted MRI scans from the same participants, the quality of which were rated by neuroradiologists from a clinical diagnostic point of view. Our results revealed a systematic increase in brain-PAD with worsening image quality for both models. This effect was also observed for images that were deemed usable from a clinical perspective, with brains appearing older in medium than in good quality images. These findings were also supported by significant associations found between the brain-PAD and standard image quality metrics indicating larger brain-PAD for lower-quality images. Our results demonstrate a spurious effect of advanced brain aging as a result of head motion and underline the importance of controlling for image quality when using brain-predicted age based on structural neuroimaging data as a proxy measure for brain health.

Automated, High-Throughput Platform to Generate a High-Reliability, Comprehensive Rectal Cancer Database.

Dynamic operations platforms allow for cross-platform data extraction, integration, and analysis, although application of these platforms to large-scale oncology enterprises has not been described. This study presents a pipeline for automated, high-fidelity extraction, integration, and validation of cross-platform oncology data in patients undergoing treatment for rectal cancer at a single, high-volume institution. A dynamic operations platform was used to identify patients with rectal cancer treated at MD Anderson Cancer Center between 2016 and 2022 who had magnetic resonance imaging (MRI) imaging and preoperative treatment details available in the electronic health record (EHR). Demographic, clinicopathologic, tumor mutation, radiographic, and treatment data were extracted from the EHR using a methodology adaptable to any disease site. Data accuracy was assessed by manual review. Accuracy before and after implementation of synoptic reporting was determined for MRI data. A total of 516 patients with localized rectal cancer were included. In the era after institutional adoption of synoptic reports, the dynamic operations platform extracted T (tumor) category data from the EHR with 95% accuracy compared with 87% before the use of synoptic reports, and N (lymph node) category with 88% compared with 58%. Correct extraction of pelvic sidewall adenopathy was 94% compared with 78%, and extramural vascular invasion accuracy was 99% compared with 89%. Neoadjuvant chemotherapy and radiation data were 99% accurate for patients who had synoptic data sources. Using dynamic operations platforms enables automated cross-platform integration of multiparameter oncology data with high fidelity in patients undergoing multimodality treatment for rectal cancer. These pipelines can be adapted to other solid tumors and, together with standardized reporting, can increase efficiency in clinical research and the translation of actionable findings toward optimizing patient outcomes.

Development of segregation and integration of functional connectomes during the first 1,000 days

The first 1,000days of human life lay the foundation for brain development and later cognitive growth. However, the developmental rules of the functional connectome during this critical period remain unclear. Using high-resolution, longitudinal, task-free functional magnetic resonance imaging data from 930 scans of 665 infants aged 28 postmenstrual weeks to 3 years, we report the early maturational process of connectome segregation and integration. We show the dominant development of local connections alongside a few global connections, the shift of brain hubs from primary regions to high-order association cortices, the developmental divergence of network segregation and integration along the anterior-posterior axis, the prediction of neurocognitive outcomes, and their associations with gene expression signatures of microstructural development and neuronal metabolic pathways. These findings advance our understanding of the principles of connectome remodeling during early life and its neurobiological underpinnings and have implications for studying typical and atypical development.

Unveiling convergent and divergent intrinsic brain network alternations in depressed adolescents engaged in non-suicidal self-injurious behaviors with and without suicide attempts.

Limited understanding exists regarding the neurobiological mechanisms underlying non-suicidal self-injury (NSSI) and suicide attempts (SA) in depressed adolescents. The maturation of brain network is crucial during adolescence, yet the abnormal alternations in depressed adolescents with NSSI or NSSI+SA remain poorly understood. Resting-state functional magnetic resonance imaging data were collected from 114 depressed adolescents, classified into three groups: clinical control (non-self-harm), NSSI only, and NSSI+SA based on self-harm history. The alternations of resting-state functional connectivity (RSFC) were identified through support vector machine-based classification. Convergent alterations in NSSI and NSSI+SA predominantly centered on the inter-network RSFC between the Limbic network and the three core neurocognitive networks (SalVAttn, Control, and Default networks). Divergent alterations in the NSSI+SA group primarily focused on the Visual, Limbic, and Subcortical networks. Additionally, the severity of depressive symptoms only showed a significant correlation with altered RSFCs between Limbic and DorsAttn or Visual networks, strengthening the fact that increased depression severity alone does not fully explain observed FC alternations in the NSSI+SA group. Convergent alterations suggest a shared neurobiological mechanism along the self-destructiveness continuum. Divergent alterations may indicate biomarkers differentiating risk for SA, informing neurobiologically guided interventions.