Identifying central and peripheral nerve fibres with an artificial intelligence approach

Abstract

Distinguishing axons from central or peripheral nervous systems (CNS or PNS, respectively) is often a complicated task. The main objective of this work was to facilitate and support the process of automatically distinguishing the different types of nerve fibres by analysing their morphological characteristics. Our approach was based on a multi-level hierarchical classifier architecture that can handle the complexity of directly identifying nerve-fibre groups belonging to either the CNS or the PNS. The approach adopted comprises supervised methods (multilayer perceptron and decision trees), which are responsible for distinguishing the origin of the axons (CNS or PNS), whereas the unsupervised method (K-Means clustering) performs nerve fibre clustering based on similar characteristics for both the CNS and PNS. Our experiments produced results with an accuracy higher than 88%. Our findings suggest that the development and implementation of a multi-level system improves automation capabilities and increases accuracy in the classification of nerves. Furthermore, our architecture allows for generalisation and flexibility, which can subsequently be extended to other biological control systems.

Integrated Evolving Spiking Neural Network and Feature Extraction Methods for Scoliosis Classification

“…The L and M were adjusted between 0.1 and 1, with one hidden neural net layer. On the other hand, the number of H was adjusted between 1 and 25 [50,51]. From this experiment, the highest accuracy can be listed with the parameter value of H, L, and M. The results of the experiments are shown in Tab.…”

“…This process significantly impacts classification performance. Normalization enhances data quality where greater numeric feature values are unable to dominate smaller numeric values [50]. This experiment implements normalization with a range between 0 to 1, as shown in Eq.…”

“…To build the MLP model, three hyperparameters are fine-tuned in this research: The Learning rate (L), Momentum rate (M), and Number of hidden layers (H). These hyperparameters are modified to achieve an optimal solution for the model [50]. The number of hidden layers represents a trade-off between performance and the risk of overfitting.…”

A Machine Learning and Integration Based Architecture for Cognitive Disorder Detection Used for Early Autism Screening

“…Our goal has also been to learn from and disseminate research and best practices to help address this broader goal. Specifically, the authors previously used machine learning methods in the health area to diagnose and model urological dysfunctions [7][8][9], predict seminal quality factors [10,11], and classify central and peripheral nerve fibers [12].…”