Bilaloglu, S. | Search bioRxiv

bioRxiv2022

InfoMuNet: Information-theory-based Functional Muscle Network Tracks Sensorimotor Integration Post-stroke

Sensory information is critical for motor coordination. However, understanding sensorimotor integration is complicated, especially in individuals with nervous system impairment. This research presents a novel functional biomarker, based on a nonlinear network graph of muscle connectivity, called InfoMuNet, to quantify the role of sensory information in motor performance. Thirty-two individuals with post-stroke hemiparesis performed a grasp-and-lift task while muscle activities were measured using eight surface electromyography (sEMG) sensors. Subjects performed the task with their affected hand before and after exposure to the sensory stimulation elicited by performing the task with the less-affected hand. For the first time, this work shows that InfoMuNet robustly quantifies functional muscle connectivity improvements in the affected hand after exposure of the less-affected side to sensory information. >90% of the subjects conformed with the improvement resulting from this sensory exposure. InfoMuNet also shows high sensitivity to tactile, kinesthetic, and visual input alterations at the subject level, highlighting the potential use in precision rehabilitation interventions.

neuroscience↗

bioRxiv2019

Efficient pan-cancer whole-slide image classification and outlier detection using convolutional neural networks

Visual analysis of solid tissue mounted on glass slides is currently the primary method used by pathologists for determining the stage, type and subtypes of cancer. Although whole slide images are usually large (10s to 100s thousands pixels wide), an exhaustive though time-consuming assessment is necessary to reduce the risk of misdiagnosis. In an effort to address the many diagnostic challenges faced by trained experts, recent research has been focused on developing automatic prediction systems for this multi-class classification problem. Typically, complex convolutional neural network (CNN) architectures, such as Googles Inception, are used to tackle this problem. Here, we introduce a greatly simplified CNN architecture, PathCNN, which allows for more efficient use of computational resources and better classification performance. Using this improved architecture, we trained simultaneously on whole-slide images from multiple tumor sites and corresponding non-neoplastic tissue. Dimensionality reduction analysis of the weights of the last layer of the network capture groups of images that faithfully represent the different types of cancer, highlighting at the same time differences in staining and capturing outliers, artifacts and misclassification errors. Our code is available online at: https://github.com/sedab/PathCNN.

pathology↗