A new class of chaotic attractors using different activation functions in neuron with multi dendrites.

This paper introduces a novel class of chaotic attractors by lever- aging different activation functions within neurons possessing multiple dendrites. We propose a comprehensive framework where the activation functions in neurons are varied, allowing for diverse behaviors such as amplification, fluctuation, and folding of scrolls within the resulting chaotic attractors. By employing wavelet functions and other model-specific activation functions, we demonstrate the capability to modify scroll characteristics, including size and direction.

Deep learning based neural network application for automatic ultrasonic computed tomographic bone image segmentation.

Deep-learning techniques have led to technological progress in the area of medical imaging segmentation especially in the ultrasound domain. In this paper, the main goal of this study is to optimize a deep-learning-based neural network architecture for automatic segmentation in Ultrasonic Computed Tomography (USCT) bone images in a short time process. The proposed method is based on an end to end neural network architecture.

Chaotic attractors with separated scrolls.

This paper proposes a new behavior of chaotic attractors with separated scrolls while combining Julia's process with Chua's attractor and Lorenz's attractor. The main motivation of this work is the ability to generate a set of separated scrolls with different behaviors, which in turn allows us to choose one or many scrolls combined with modulation (amplitude and frequency) for secure communication or synchronization. This set seems a new class of hyperchaos because each element of this set looks like a simple chaotic attractor with one positive Lyapunov exponent, so the cardinal of this set is greater than one.