Indirect frontocingulate structural connectivity predicts clinical response to accelerated rTMS in major depressive disorder

Background: Repetitive transcranial magnetic stimulation (rTMS) is an established treatment for major depressive disorder (MDD), but its clinical efficacy remains rather modest. One reason for this could be that the propagation of rTMS effects via structural connections from the stimulated area to deeper brain structures (such as the cingulate cortices) is suboptimal. Methods: We investigated whether structural connectivity ­ derived from diffusion MRI data ­ could serve as a biomarker to predict treatment response. We hypothesized that stronger structural connections between the patient-specific stimulation position in the left dorsolateral prefrontal cortex (dlPFC) and the cingulate cortices would predict better clinical outcomes. We applied accelerated intermittent theta burst stimulation (aiTBS) to the left dlPFC in 40 patients with MDD. We correlated baseline structural connectivity, quantified using various metrics (fractional anisotropy, mean diffusivity, tract density, tract volume and number of tracts), with changes in depression severity scores after aiTBS. Results: Exploratory results (p < 0.05) showed that structural connectivity between the patient-specific stimulation site and the caudal and posterior parts of the cingulate cortex had predictive potential for clinical response to aiTBS. Limitations: We used the diffusion tensor to perform tractography. A main limitation was that multiple fibre directions within voxels could not be resolved, which might have led to missing connections in some patients. Conclusion: Stronger structural frontocingular connections may be of essence to optimally benefit from left dlPFC rTMS treatment in MDD. Even though the results are promising, further investigation with larger numbers of patients, more advanced tractography algorithms and classic daily rTMS treatment paradigms is warranted. Clinical trial registration: http://clinicaltrials.gov/show/NCT01832805

Graph convolutional networks for automated intracranial artery labeling.

Purpose: Unruptured intracranial aneurysms (UIAs) can cause aneurysmal subarachnoid hemorrhage, a severe and often lethal type of stroke. Automated labeling of intracranial arteries can facilitate the identification of risk factors associated with UIAs. This study aims to improve intracranial artery labeling using atlas-based features in graph convolutional networks. Approach: We included three-dimensional time-of-flight magnetic resonance angiography scans from 150 individuals. Two widely used graph convolutional operators, GCNConv and GraphConv, were employed in models trained to classify 12 bifurcations of interest. Cross-validation was applied to explore the effectiveness of atlas-based features in node classification. The results were tested for statistically significant differences using a Wilcoxon signed-rank test. Model repeatability and calibration were assessed on the test set for both operators. In addition, we evaluated model interpretability and node feature contribution using explainable artificial intelligence. Results: Atlas-based features led to statistically significant improvements in node classification (p<0.05). The results showed that the best discrimination and calibration performances were obtained using the GraphConv operator, which yielded a mean recall of 0.87, precision of 0.90, and expected calibration error of 0.02. Conclusions: The addition of atlas-based features improved node classification results. The GraphConv operator, which incorporates higher-order structural information during training, is recommended over the GCNConv operator based on the accuracy and calibration of predicted outcomes.

Geometric Deep Learning Using Vascular Surface Meshes for Modality-Independent Unruptured Intracranial Aneurysm Detection.

Early detection of unruptured intracranial aneurysms (UIAs) enables better rupture risk and preventative treatment assessment. UIAs are usually diagnosed on Time-of-Flight Magnetic Resonance Angiographs (TOF-MRA) or contrast-enhanced Computed Tomography Angiographs (CTA). Various automatic voxel-based deep learning UIA detection methods have been developed, but these are limited to a single modality. We propose a modality-independent UIA detection method using a geometric deep learning model with high resolution surface meshes of brain vessels. A mesh convolutional neural network with ResU-Net style architecture was used. UIA detection performance was investigated with different input and pooling mesh resolutions, and including additional edge input features (shape index and curvedness). Both a higher resolution mesh (15,000 edges) and additional curvature edge features improved performance (average sensitivity: 65.6%, false positive count/image (FPC/image): 1.61). UIAs were detected in an independent TOF-MRA test set and a CTA test set with average sensitivity of 52.0% and 48.3% and average FPC/image of 1.04 and 1.05 respectively. We provide modality-independent UIA detection using a deep-learning vascular surface mesh model with comparable performance to state-of-the-art UIA detection methods.