Geometric Deep Learning of the Human Connectome Project Multimodal Cortical Parcellation

Understanding the topographic heterogeneity of cortical organisation is an essential step towards precision modelling of neuropsychiatric disorders. While many cortical parcellation schemes have been proposed, few attempt to model inter-subject variability. For those that do, most have been proposed for high-resolution research quality data, without exploration of how well they generalise to clinical quality scans. In this paper, we benchmark and ensemble four different geometric deep learning models on the task of learning the Human Connectome Project (HCP) multimodal cortical parcellation. We employ Monte Carlo dropout to investigate model uncertainty with a view to propagate these labels to new datasets. Models achieved an overall Dice overlap ratio of >0.85 ± 0.02. Regions with the highest mean and lowest variance included V1 and areas within the parietal lobe, and regions with the lowest mean and highest variance included areas within the medial frontal lobe, lateral occipital pole and insula. Qualitatively, our results suggest that more work is needed before geometric deep learning methods are capable of fully capturing atypical cortical topographies such as those seen in area 55b. However, information about topographic variability between participants was encoded in vertex-wise uncertainty maps, suggesting a potential avenue for projection of this multimodal parcellation to new datasets with limited functional MRI, such as the UK Biobank.

Foundations of human problem solving

Despite great strides in both machine learning and neuroscience, we do not know how the human brain solves problems in the general sense. We approach this question by drawing on the framework of engineering control theory. We demonstrate a computational neural model with only localist learning laws that is able to find solutions to arbitrary problems. Using a combination of computational neural modeling, human fMRI, and representational similarity analysis, we show here that the roles of a number of brain regions can be reinterpreted as interacting mechanisms of a control theoretic system. The results suggest a new set of functional perspectives on the orbitofrontal cortex, hippocampus, basal ganglia, anterior temporal lobe, lateral prefrontal cortex, and visual cortex, as well as a new path toward artificial general intelligence.

Olfactory Projection Neurons From the Moth Antennal Lobe Lateral Cluster exhibit Diverse Morphological and Neurophysiological Characteristics

Olfactory projection neurons convey information from the insect antennal lobe (AL) to higher centers in the brain. Many studies on moths have reported excitatory projection neurons with cell bodies in the medial cell cluster (mcPNs) that predominantly send an axon from the AL to calyces of the mushroom body (CA) via the medial antennal lobe tract (mALT) and then to the lateral horn (LH) of the protocerebrum. These neurons tend to have dendritic arbors restricted to a single glomerulus (i.e. they are uniglomerular). In this study, we report on the physiological and morphological properties of a group of pheromone-responsive olfactory projection neurons with cell bodies in the moth AL lateral cell cluster (lcPNs) of two heliothine moth species. While mcPNs typically exhibit a narrow odor tuning range related to the restriction of their dendritic arbors within a single glomerulus, lcPNs exhibited an array of morphological and physiological configurations. Pheromone-responsive lcPNs varied in their associations with glomeruli (uniglomerular and multiglomerular), dendritic arborization structure and connections to higher brain centers with projections primarily through the lateral antennal lobe tract and to a lesser extent the mediolateral antennal lobe tract to a variety of protocerebral targets including ventrolateral and superior neuropils as well as LH. Physiological characterization of lcPNs also revealed a diversity of response profiles including those either enhanced by or reliant upon presentation of a pheromone blend. These responses manifested themselves as higher maximum firing rates and/or improved temporal resolution of pulsatile stimuli. lcPNs therefore participate in conveying a variety of olfactory information relating to qualitative and temporal facets of the pheromone stimulus to a more expansive number of protocerebral targets than their mcPN counterparts. The role of lcPNs in the overall scheme of olfactory processing is discussed.

Tail kinematics of the chub mackerel Scomber japonicus: testing the homocercal tail model of fish propulsion

Scombrid fishes possess a homocercal caudal fin with reduced intrinsic musculature and dorso-ventrally symmetrical external and internal morphology. Because of this symmetrical morphology, it has often been assumed that scombrid caudal fins function as predicted by the homocercal tail model. According to that model, the caudal fin moves in a dorso-ventrally symmetrical manner and produces no vertical lift during steady swimming. To test this hypothesis, we examined the tail kinematics of chub mackerel, Scomber japonicus (24.8+/−1.3 cm total length, L). Markers were placed on the caudal fin to identify specific regions of the tail, and swimming chub mackerel were videotaped from lateral and posterior views, allowing a three-dimensional analysis of tail motion. Analysis of tail kinematics suggests that, at a range of swimming speeds (1.2-3.0 L s(−)(1)), the dorsal lobe of the tail undergoes a 15 % greater lateral excursion than does the ventral lobe. Lateral excursion of the dorsal tail-tip also increases significantly by 32 % over this range of speeds, indicating a substantial increase in tail-beat amplitude with speed. In addition, if the tail were functioning in a dorso-ventrally symmetrical manner, the tail should subtend an angle of 90 degrees relative to the frontal (or xz) plane throughout the tail beat. Three-dimensional kinematic analyses reveal that the caudal fin actually reaches a minimum xz angle of 79.8 degrees. In addition, there is no difference between the angle subtended by the caudal peduncle (which is anterior to the intrinsic tail musculature) and that subtended by the posterior lobes of the tail. Thus, asymmetrical movements of the tail are apparently generated by the axial musculature and transmitted posteriorly to the caudal fin. These results suggest that the caudal fin of the chub mackerel is not functioning symmetrically according to the homocercal model and could produce upward lift during steady swimming.