Application of individual brain connectome in chronic ischemia: mapping symptoms before and after reperfusion.

How brain functions in the distorted ischemic state before and after reperfusion is unclear. It is also uncertain whether there are any indicators within ischemic brain that could predict surgical outcomes. To alleviate these issues, we applied individual brain connectome in chronic steno-occlusive vasculopathy (CSOV) to map both ischemic symptoms and their postbypass changes. A total of 499 bypasses in 455 CSOV patients were collected and followed up for 47.8 ± 20.5 months. Using multimodal parcellation with connectivity-based and pathological distortion-independent approach, areal MR features of brain connectome were generated with three measurements of functional connectivity (FC), structural connectivity, and PageRank centrality at the single-subject level. Thirty-three machine-learning models were then trained with clinical and areal MR features to obtain acceptable classifiers for both ischemic symptoms and their postbypass changes, among which, 11 were deemed acceptable (AUC > 0.7). Notably, the FC feature-based model for long-term neurological outcomes performed very well (AUC > 0.8). Finally, a Shapley additive explanations plot was adopted to extract important individual features in acceptable models to generate "fingerprints" of brain connectome. This study not only establishes brain connectomic fingerprint databases for brain ischemia with distortion, but also provides informative insights for how brain functions before and after reperfusion.

Parcellation-Based Connectivity Model of the Judgement Core.

Judgement is a higher-order brain function utilized in the evaluation process of problem solving. However, heterogeneity in the task methodology based on the many definitions of judgement and its expansive and nuanced applications have prevented the identification of a unified cortical model at a level of granularity necessary for clinical translation. Forty-six task-based fMRI studies were used to generate activation-likelihood estimations (ALE) across moral, social, risky, and interpersonal judgement paradigms. Cortical parcellations overlapping these ALEs were used to delineate patterns in neurocognitive network engagement for the four judgement tasks. Moral judgement involved the bilateral superior frontal gyri, right temporal gyri, and left parietal lobe. Social judgement demonstrated a left-dominant frontoparietal network with engagement of right-sided temporal limbic regions. Moral and social judgement tasks evoked mutual engagement of the bilateral DMN. Both interpersonal and risk judgement were shown to involve a right-sided frontoparietal network with accompanying engagement of the left insular cortex, converging at the right-sided CEN. Cortical activation in normophysiological judgement function followed two separable patterns involving the large-scale neurocognitive networks. Specifically, the DMN was found to subserve judgement centered around social inferences and moral cognition, while the CEN subserved tasks involving probabilistic reasoning, risk estimation, and strategic contemplation.

Functional connectivity of the language area in migraine: a preliminary classification model.

BACKGROUND: Migraine is a complex disorder characterized by debilitating headaches. Despite its prevalence, its pathophysiology remains unknown, with subsequent gaps in diagnosis and treatment. We combined machine learning with connectivity analysis and applied a whole-brain network approach to identify potential targets for migraine diagnosis and treatment. METHODS: Baseline anatomical T1 magnetic resonance imaging (MRI), resting-state functional MRI(rfMRI), and diffusion weighted scans were obtained from 31 patients with migraine, and 17 controls. A recently developed machine learning technique, Hollow Tree Super (HoTS) was used to classify subjects into diagnostic groups based on functional connectivity (FC) and derive networks and parcels contributing to the model. PageRank centrality analysis was also performed on the structural connectome to identify changes in hubness. RESULTS: Our model attained an area under the receiver operating characteristic curve (AUC-ROC) of 0.68, which rose to 0.86 following hyperparameter tuning. FC of the language network was most predictive of the model's classification, though patients with migraine also demonstrated differences in the accessory language, visual and medial temporal regions. Several analogous regions in the right hemisphere demonstrated changes in PageRank centrality, suggesting possible compensation. CONCLUSIONS: Although our small sample size demands caution, our preliminary findings demonstrate the utility of our method in providing a network-based perspective to diagnosis and treatment of migraine.

Connectomics underlying motor functional outcomes in the acute period following stroke.

Objective: Stroke remains the number one cause of morbidity in many developing countries, and while effective neurorehabilitation strategies exist, it remains difficult to predict the individual trajectories of patients in the acute period, making personalized therapies difficult. Sophisticated and data-driven methods are necessary to identify markers of functional outcomes. Methods: Baseline anatomical T1 magnetic resonance imaging (MRI), resting-state functional MRI (rsfMRI), and diffusion weighted scans were obtained from 79 patients following stroke. Sixteen models were constructed to predict performance across six tests of motor impairment, spasticity, and activities of daily living, using either whole-brain structural or functional connectivity. Feature importance analysis was also performed to identify brain regions and networks associated with performance in each test. Results: The area under the receiver operating characteristic curve ranged from 0.650 to 0.868. Models utilizing functional connectivity tended to have better performance than those utilizing structural connectivity. The Dorsal and Ventral Attention Networks were among the top three features in several structural and functional models, while the Language and Accessory Language Networks were most commonly implicated in structural models. Conclusions: Our study highlights the potential of machine learning methods combined with connectivity analysis in predicting outcomes in neurorehabilitation and disentangling the neural correlates of functional impairments, though further longitudinal studies are necessary.

Symptom-circuit mappings of the schizophrenia connectome.

OBJECTIVE: This paper aims to model the anatomical circuits underlying schizophrenia symptoms, and to explore patterns of abnormal connectivity among brain networks affected by psychopathology. METHODS: T1 magnetic resonance imaging (MRI), diffusion weighted imaging (DWI), and resting-state functional MRI (rsfMRI) were obtained from a total of 126 patients with schizophrenia who were recruited for the study. The images were processed using the Omniscient software (https://www.o8t. com). We further apply the use of the Hollow-tree Super (HoTS) method to gain insights into what brain regions had abnormal connectivity that might be linked to the symptoms of schizophrenia. RESULTS: The Positive and Negative Symptom Scale is characterised into 6 factors. Each symptom is mapped with specific anatomical abnormalities and circuits. Comparison between factors reveals co-occurrence in parcels in Factor 1 and Factor 2. Multiple large-scale networks are involved in SCZ symptomatology, with functional connectivity within Default Mode Network (DMN) and Central Executive Network (CEN) regions most frequently associated with measures of psychopathology. CONCLUSION: We present a summary of the relevant anatomy for regions of the cortical areas as part of a larger effort to understand its contribution in schizophrenia. This unique machine learning-type approach maps symptoms to specific brain regions and circuits by bridging the diagnostic subtypes and analysing the features of the connectome.

An agile, data-driven approach for target selection in rTMS therapy for anxiety symptoms: Proof of concept and preliminary data for two novel targets.

INTRODUCTION: Data-driven approaches to transcranial magnetic stimulation (TMS) might yield more consistent and symptom-specific results based on individualized functional connectivity analyses compared to previous traditional approaches due to more precise targeting. We provide a proof of concept for an agile target selection paradigm based on using connectomic methods that can be used to detect patient-specific abnormal functional connectivity, guide treatment aimed at the most abnormal regions, and optimize the rapid development of new hypotheses for future study. METHODS: We used the resting-state functional MRI data of 28 patients with medically refractory generalized anxiety disorder to perform agile target selection based on abnormal functional connectivity patterns between the Default Mode Network (DMN) and Central Executive Network (CEN). The most abnormal areas of connectivity within these regions were selected for subsequent targeted TMS treatment by a machine learning based on an anomalous functional connectivity detection matrix. Areas with mostly hyperconnectivity were stimulated with continuous theta burst stimulation and the converse with intermittent theta burst stimulation. An image-guided accelerated theta burst stimulation paradigm was used for treatment. RESULTS: Areas 8Av and PGs demonstrated consistent abnormalities, particularly in the left hemisphere. Significant improvements were demonstrated in anxiety symptoms, and few, minor complications were reported (fatigue (n = 2) and headache (n = 1)). CONCLUSIONS: Our study suggests that a left-lateralized DMN is likely the primary functional network disturbed in anxiety-related disorders, which can be improved by identifying and targeting abnormal regions with a rapid, data-driven, agile aTBS treatment on an individualized basis.

Real world demonstration of hand motor mapping using the structural connectivity atlas.

BACKGROUND: Locating the hand-motor-cortex (HMC) is an essential component within many neurosurgeries. Despite advancements in these localization methods there are still downfalls for each. Additionally, the importance of presurgical planning calls for increasingly accurate and efficient methods of locating specific cortical regions. OBJECTIVE: In this study we aimed to test the ability of the Structural Connectivity Atlas (SCA), a machine-learning based method to parcellate the human cortex, to locate the HMC in a small cohort study. METHODS: Using MRI and DTI images obtained from adult subjects (n = 11), personalized brain maps were created for each individual based on a SCA paired with the Brainnetome region for the HMC. Subjects received single pulse TMS, over the HMC region through the use of a neuronavigation system. If they responded with motor movement, this was recorded. The SCA identified HMC region was compared to the visual-determined HMC through identifying the Omega fold on the Precentral Gyrus, which was completed by a trained neuroanatomist. A Kendall's Tau B correlation was conducted between anatomical match and visual movement. RESULTS: This study concluded that the SCA was capable of locating the HMC in healthy and distorted brains. Overall, the SCA defined the anatomical area of the HMC in 90 % of subjects and triggered a motor response in 61 %. CONCLUSION: The SCA could be suitable for incorporation into presurgical planning practices due to its ability to map anatomically abnormal brains. Further studies on larger cohorts and targeting different areas of cortex could be beneficial.

Functional connectivity analysis of the depression connectome provides potential markers and targets for transcranial magnetic stimulation.

BACKGROUND: Despite efforts to improve targeting accuracy of the dorsolateral prefrontal cortex (DLPFC) as a repetitive transcranial magnetic stimulation (rTMS) target for Major Depressive Disorder (MDD), the heterogeneity in clinical response remains unexplained. OBJECTIVE: We sought to compare the patterns of functional connectivity from the DLPFC treatment site in patients with MDD who were TMS responders to those who were TMS non-responders. METHODS: Baseline anatomical T1 magnetic resonance imaging (MRI), resting-state functional MRI, and diffusion weighted imaging scans were obtained from 37 participants before they underwent a course of rTMS to left Brodmann area 46. A novel machine learning method was utilized to identify brain regions associated with each item of the Beck's Depression Inventory II (BDI-II), and for 26 participants who underwent rTMS treatment over the left Brodmann area 46, identify regions differentiating rTMS responders and non-responders. RESULTS: Nine parcels of the Human Connectome Project Multimodal Parcellation Atlas matched to at least three items of the Beck's Depression Inventory II (BDI-II) as predictors of response to rTMS, with many in the temporal, parietal and cingulate cortices. Additionally, pre-treatment mapping for 17 items of the BDI-II demonstrated significant variability in symptom to parcel mapping. When parcels associated with symptom presence and symptom resolution were compared, 15 parcels were uniquely associated with resolution (potential targets), and 12 parcels were associated with both symptom presence and resolution (blockers or biomarkers). CONCLUSIONS: Machine learning approaches show promise for the development of pathoanatomical diagnosis and treatment algorithms for MDD. Prospective studies are required to facilitate clinical translation.

Functional connectivity in ADHD children doing Go/No-Go tasks: An fMRI systematic review and meta-analysis.

Attention deficit hyperactivity disorder (ADHD) is one of the most common neurodevelopmental disorders diagnosed in childhood. Two common features of ADHD are impaired behavioural inhibition and sustained attention. The Go/No-Go experimental paradigm with concurrent functional magnetic resonance imaging (fMRI) scanning has previously revealed important neurobiological correlates of ADHD such as the supplementary motor area and the prefrontal cortex. The coordinate-based meta-analysis combined with quantitative techniques, such as activation likelihood estimate (ALE) generation, provides an unbiased and objective method of summarising these data to understand the brain network architecture and connectivity in ADHD children. Go/No-Go task-based fMRI studies involving children and adolescent subjects were selected. Coordinates indicating foci of activation were collected to generate ALEs using threshold values (voxel-level: p < 0.001; cluster-level: p < 0.05). ALEs were matched to one of seven canonical brain networks based on the cortical parcellation scheme derived from the Human Connectome Project. Fourteen studies involving 457 children met the eligibility criteria. No significant convergence of Go/No-Go related brain activation was found for ADHD groups. Three significant ALE clusters were detected for brain activation relating to controls or ADHD < controls. Significant clusters were related to specific areas of the default mode network (DMN). Network-based analysis revealed less extensive DMN, dorsal attention network, and limbic network activation in ADHD children compared to controls. The presence of significant ALE clusters may be due to reduced homogeneity in the selected sample demographic and experimental paradigm. Further investigations regarding hemispheric asymmetry in ADHD subjects would be beneficial.

Use of machine learning to identify functional connectivity changes in a clinical cohort of patients at risk for dementia.

Objective: Progressive conditions characterized by cognitive decline, including mild cognitive impairment (MCI) and subjective cognitive decline (SCD) are clinical conditions representing a major risk factor to develop dementia, however, the diagnosis of these pre-dementia conditions remains a challenge given the heterogeneity in clinical trajectories. Earlier diagnosis requires data-driven approaches for improved and targeted treatment modalities. Methods: Neuropsychological tests, baseline anatomical T1 magnetic resonance imaging (MRI), resting-state functional MRI (rsfMRI), and diffusion weighted scans were obtained from 35 patients with SCD, 19 with MCI, and 36 age-matched healthy controls (HC). A recently developed machine learning technique, Hollow Tree Super (HoTS) was utilized to classify subjects into diagnostic categories based on their FC, and derive network and parcel-based FC features contributing to each model. The same approach was used to identify features associated with performance in a range of neuropsychological tests. We concluded our analysis by looking at changes in PageRank centrality (a measure of node hubness) between the diagnostic groups. Results: Subjects were classified into diagnostic categories with a high area under the receiver operating characteristic curve (AUC-ROC), ranging from 0.73 to 0.84. The language networks were most notably associated with classification. Several central networks and sensory brain regions were predictors of poor performance in neuropsychological tests, suggesting maladaptive compensation. PageRank analysis highlighted that basal and limbic deep brain region, along with the frontal operculum demonstrated a reduction in centrality in both SCD and MCI patients compared to controls. Conclusion: Our methods highlight the potential to explore the underlying neural networks contributing to the cognitive changes and neuroplastic responses in prodromal dementia.

Connectomic disturbances underlying insomnia disorder and predictors of treatment response.

Objective: Despite its prevalence, insomnia disorder (ID) remains poorly understood. In this study, we used machine learning to analyze the functional connectivity (FC) disturbances underlying ID, and identify potential predictors of treatment response through recurrent transcranial magnetic stimulation (rTMS) and pharmacotherapy. Materials and methods: 51 adult patients with chronic insomnia and 42 healthy age and education matched controls underwent baseline anatomical T1 magnetic resonance imaging (MRI), resting-stage functional MRI (rsfMRI), and diffusion weighted imaging (DWI). Imaging was repeated for 24 ID patients following four weeks of treatment with pharmacotherapy, with or without rTMS. A recently developed machine learning technique, Hollow Tree Super (HoTS) was used to classify subjects into ID and control groups based on their FC, and derive network and parcel-based FC features contributing to each model. The number of FC anomalies within each network was also compared between responders and non-responders using median absolute deviation at baseline and follow-up. Results: Subjects were classified into ID and control with an area under the receiver operating characteristic curve (AUC-ROC) of 0.828. Baseline FC anomaly counts were higher in responders than non-responders. Response as measured by the Insomnia Severity Index (ISI) was associated with a decrease in anomaly counts across all networks, while all networks showed an increase in anomaly counts when response was measured using the Pittsburgh Sleep Quality Index. Overall, responders also showed greater change in all networks, with the Default Mode Network demonstrating the greatest change. Conclusion: Machine learning analysis into the functional connectome in ID may provide useful insight into diagnostic and therapeutic targets.

A connectivity model of the anatomic substrates underlying Gerstmann syndrome.

The Gerstmann syndrome is a constellation of neurological deficits that include agraphia, acalculia, left-right discrimination and finger agnosia. Despite a growing interest in this clinical phenomenon, there remains controversy regarding the specific neuroanatomic substrates involved. Advancements in data-driven, computational modelling provides an opportunity to create a unified cortical model with greater anatomic precision based on underlying structural and functional connectivity across complex cognitive domains. A literature search was conducted for healthy task-based functional MRI and PET studies for the four cognitive domains underlying Gerstmann's tetrad using the electronic databases PubMed, Medline, and BrainMap Sleuth (2.4). Coordinate-based, meta-analytic software was utilized to gather relevant regions of interest from included studies to create an activation likelihood estimation (ALE) map for each cognitive domain. Machine-learning was used to match activated regions of the ALE to the corresponding parcel from the cortical parcellation scheme previously published under the Human Connectome Project (HCP). Diffusion spectrum imaging-based tractography was performed to determine the structural connectivity between relevant parcels in each domain on 51 healthy subjects from the HCP database. Ultimately 102 functional MRI studies met our inclusion criteria. A frontoparietal network was found to be involved in the four cognitive domains: calculation, writing, finger gnosis, and left-right orientation. There were three parcels in the left hemisphere, where the ALE of at least three cognitive domains were found to be overlapping, specifically the anterior intraparietal area, area 7 postcentral (7PC) and the medial intraparietal sulcus. These parcels surround the anteromedial portion of the intraparietal sulcus. Area 7PC was found to be involved in all four domains. These regions were extensively connected in the intraparietal sulcus, as well as with a number of surrounding large-scale brain networks involved in higher-order functions. We present a tractographic model of the four neural networks involved in the functions which are impaired in Gerstmann syndrome. We identified a 'Gerstmann Core' of extensively connected functional regions where at least three of the four networks overlap. These results provide clinically actionable and precise anatomic information which may help guide clinical translation in this region, such as during resective brain surgery in or near the intraparietal sulcus, and provides an empiric basis for future study.

Using a ResNet-18 Network to Detect Features of Alzheimer's Disease on Functional Magnetic Resonance Imaging: A Failed Replication. Comment on Odusami et al. Analysis of Features of Alzheimer's Disease: Detection of Early Stage from Functional Brain Changes in Magnetic Resonance Images Using a Finetuned ResNet18 Network. Diagnostics 2021, 11, 1071.

There is considerable interest in developing effective tools to detect Alzheimer's Disease (AD) early in its course, prior to clinical progression [...].

12 Plagues of AI in Healthcare: A Practical Guide to Current Issues With Using Machine Learning in a Medical Context.

The healthcare field has long been promised a number of exciting and powerful applications of Artificial Intelligence (AI) to improve the quality and delivery of health care services. AI techniques, such as machine learning (ML), have proven the ability to model enormous amounts of complex data and biological phenomena in ways only imaginable with human abilities alone. As such, medical professionals, data scientists, and Big Tech companies alike have all invested substantial time, effort, and funding into these technologies with hopes that AI systems will provide rigorous and systematic interpretations of large amounts of data that can be leveraged to augment clinical judgments in real time. However, despite not being newly introduced, AI-based medical devices have more than often been limited in their true clinical impact that was originally promised or that which is likely capable, such as during the current COVID-19 pandemic. There are several common pitfalls for these technologies that if not prospectively managed or adjusted in real-time, will continue to hinder their performance in high stakes environments outside of the lab in which they were created. To address these concerns, we outline and discuss many of the problems that future developers will likely face that contribute to these failures. Specifically, we examine the field under four lenses: approach, data, method and operation. If we continue to prospectively address and manage these concerns with reliable solutions and appropriate system processes in place, then we as a field may further optimize the clinical applicability and adoption of medical based AI technology moving forward.

Machine Learning Decomposition of the Anatomy of Neuropsychological Deficit in Alzheimer's Disease and Mild Cognitive Impairment.

Objective: Alzheimer's Disease (AD) is a progressive condition characterized by cognitive decline. AD is often preceded by mild cognitive impairment (MCI), though the diagnosis of both conditions remains a challenge. Early diagnosis of AD, and prediction of MCI progression require data-driven approaches to improve patient selection for treatment. We used a machine learning tool to predict performance in neuropsychological tests in AD and MCI based on functional connectivity using a whole-brain connectome, in an attempt to identify network substrates of cognitive deficits in AD. Methods: Neuropsychological tests, baseline anatomical T1 magnetic resonance imaging (MRI), resting-state functional MRI, and diffusion weighted imaging scans were obtained from 149 MCI, and 85 AD patients; and 140 cognitively unimpaired geriatric participants. A novel machine learning tool, Hollow Tree Super (HoTS) was utilized to extract feature importance from each machine learning model to identify brain regions that were associated with deficit and absence of deficit for 11 neuropsychological tests. Results: 11 models attained an area under the receiver operating curve (AUC-ROC) greater than 0.65, while five models had an AUC-ROC ≥ 0.7. 20 parcels of the Human Connectome Project Multimodal Parcelation Atlas matched to poor performance in at least two neuropsychological tests, while 14 parcels were associated with good performance in at least two tests. At a network level, most parcels predictive of both presence and absence of deficit were affiliated with the Central Executive Network, Default Mode Network, and the Sensorimotor Networks. Segregating predictors by the cognitive domain associated with each test revealed areas of coherent overlap between cognitive domains, with the parcels providing possible markers to screen for cognitive impairment. Conclusion: Approaches such as ours which incorporate whole-brain functional connectivity and harness feature importance in machine learning models may aid in identifying diagnostic and therapeutic targets in AD.

Eigenvector PageRank difference as a measure to reveal topological characteristics of the brain connectome for neurosurgery.

PURPOSE: Applying graph theory to the human brain has the potential to help prognosticate the impacts of intracerebral surgery. Eigenvector (EC) and PageRank (PR) centrality are two related, but uniquely different measures of nodal centrality which may be utilized together to reveal varying neuroanatomical characteristics of the brain connectome. METHODS: We obtained diffusion neuroimaging data from a healthy cohort (UCLA consortium for neuropsychiatric phenomics) and applied a personalized parcellation scheme to them. We ranked parcels based on weighted EC and PR, and then calculated the difference (EP difference) and correlation between the two metrics. We also compared the difference between the two metrics to the clustering coefficient. RESULTS: While EC and PR were consistent for top and bottom ranking parcels, they differed for mid-ranking parcels. Parcels with a high EC centrality but low PR tended to be in the medial temporal and temporooccipital regions, whereas PR conferred greater importance to multi-modal association areas in the frontal, parietal and insular cortices. The EP difference showed a weak correlation with clustering coefficient, though there was significant individual variation. CONCLUSIONS: The relationship between PageRank and eigenvector centrality can identify distinct topological characteristics of the brain connectome such as the presence of unimodal or multimodal association cortices. These results highlight how different graph theory metrics can be used alone or in combination to reveal unique neuroanatomical features for further clinical study.

Connectivity-based parcellation of normal and anatomically distorted human cerebral cortex.

For over a century, neuroscientists have been working toward parcellating the human cortex into distinct neurobiological regions. Modern technologies offer many parcellation methods for healthy cortices acquired through magnetic resonance imaging. However, these methods are suboptimal for personalized neurosurgical application given that pathology and resection distort the cerebrum. We sought to overcome this problem by developing a novel connectivity-based parcellation approach that can be applied at the single-subject level. Utilizing normative diffusion data, we first developed a machine-learning (ML) classifier to learn the typical structural connectivity patterns of healthy subjects. Specifically, the Glasser HCP atlas was utilized as a prior to calculate the streamline connectivity between each voxel and each parcel of the atlas. Using the resultant feature vector, we determined the parcel identity of each voxel in neurosurgical patients (n = 40) and thereby iteratively adjusted the prior. This approach enabled us to create patient-specific maps independent of brain shape and pathological distortion. The supervised ML classifier re-parcellated an average of 2.65% of cortical voxels across a healthy dataset (n = 178) and an average of 5.5% in neurosurgical patients. Our patient dataset consisted of subjects with supratentorial infiltrating gliomas operated on by the senior author who then assessed the validity and practical utility of the re-parcellated diffusion data. We demonstrate a rapid and effective ML parcellation approach to parcellation of the human cortex during anatomical distortion. Our approach overcomes limitations of indiscriminately applying atlas-based registration from healthy subjects by employing a voxel-wise connectivity approach based on individual data.

Changes in the Brain Connectome Following Repetitive Transcranial Magnetic Stimulation for Stroke Rehabilitation.

Repetitive transcranial magnetic stimulation (rTMS) is a promising approach for post-stroke rehabilitation but there lacks a rationale strategy to plan, execute, and monitor treatment. We present a case of targeted rTMS using the Omniscient Infinitome software to devise targets for treatment in a post-stroke patient and describe the functional connectomic changes after treatment. A 19-year-old female with no medical history presented 19 months after suffering a left middle cerebral artery (MCA) superior division ischemic stroke, resulting in language impairment and diminished right upper extremity motor function. She underwent a resting-state MRI (rsMRI) with tractography and images were processed using the Omniscient Infinitome software. Analysis using the anomaly detection within the software enabled us to identify three targets for rTMS (left area 1, left area 45, and right area SFL). These areas were treated with 25 sessions of intermittent Theta Burst Stimulation (iTBS) over five days at 80% of motor threshold concomitantly with targeted physical therapy and speech therapy. At five months follow-up, her language and right upper extremity functions significantly improved. Her connectomic analysis revealed substantial neural changes, including normalization of the sensorimotor network, substantially thicker callosal fiber bundle connecting the two hemispheres, and increased cortical recruitment in her language network. We present the first description of robust connectomic alterations in a post-stroke patient following targeted rTMS treatment. Further studies on the use of rTMS with an emphasis on functional connectomics are warranted.

Hollow-tree super: A directional and scalable approach for feature importance in boosted tree models.

PURPOSE: Current limitations in methodologies used throughout machine-learning to investigate feature importance in boosted tree modelling prevent the effective scaling to datasets with a large number of features, particularly when one is investigating both the magnitude and directionality of various features on the classification into a positive or negative class. This manuscript presents a novel methodology, "Hollow-tree Super" (HOTS), designed to resolve and visualize feature importance in boosted tree models involving a large number of features. Further, this methodology allows for accurate investigation of the directionality and magnitude various features have on classification and incorporates cross-validation to improve the accuracy and validity of the determined features of importance. METHODS: Using the Iris dataset, we first highlight the characteristics of HOTS by comparing it to other commonly used techniques for feature importance, including Gini Importance, Partial Dependence Plots, and Permutation Importance, and explain how HOTS resolves the weaknesses present in these three strategies for investigating feature importance. We then demonstrate how HOTS can be utilized in high dimensional spaces such as neuroscientific setting, by taking 60 Schizophrenic subjects from the publicly available SchizConnect database and applying the method to determine which regions of the brain were most important for the positive and negative classification of schizophrenia as determined by the positive and negative syndrome scale (PANSS). RESULTS: HOTS effectively replicated and supported the findings of feature importance for classification of the Iris dataset when compared to Gini importance, Partial Dependence Plots and Permutation importance, determining 'petal length' as the most important feature for positive and negative classification. When applied to the Schizconnect dataset, HOTS was able to resolve from 379 independent features, the top 10 most important features for classification, as well as their directionality for classification and magnitude compared to other features. Cross-validation supported that these same 10 features were consistently used in the decision-making process across multiple trees, and these features were localised primarily to the occipital and parietal cortices, commonly disturbed brain regions in those afflicted with Schizophrenia. CONCLUSION: HOTS effectively overcomes previous challenges of identifying feature importance at scale, and can be utilized across a swathe of disciplines. As computational power and data quantity continues to expand, it is imperative that a methodology is developed that is able to handle the demands of working with large datasets that contain a large number of features. This approach represents a unique way to investigate both the directionality and magnitude of feature importance when working at scale within a boosted tree model that can be easily visualized within commonly used software.