<p>The study of wave propagation in cancellous bone-like porous material has necessitated addressing some complexities such as the interaction of the ultrasonic wave with fluid, solid phase, and porous material. In addition, due to viscous exchange between pore fluid and skeletal frame, the dynamic tortuosity as a function of fractional frequency in the clinically relevant ultrasound frequency range is considered. &nbsp;The Biot-JKD theory can simulate ultrasonic propagation through a cancellous bone-like porous material and address the complexity of bone structure but this theory is conditioned with many parameters which it takes too computationally costly to solve parametric problems. Therefore, developing efficient tools and numerical methods to solve parametric problems efficiently could be so applicable. Artificial Neural Network, ANN, recognized as a simulation of the human brain is able to solve complicated-parametric problems. Although training the neural networks is tedious, it virtually instantaneously produces a realistic solution of the simulated cancellous bone-like-porous material. The purpose of this research is to propose an ANN to simulate wave propagation in cancellous bone like-porous material filled with a viscous fluid such as bone marrow. The ANNs used in this study are a feed-forward back-propagation structure with one input layer including 6 nodes, one hidden layer, and one output layer. The validation of the proposed ANN is evaluated with the analytical expression and experimental data by pertinent literature. A set of data such as porosity, viscosity, tortuosity, viscous characteristics length, Poisson’s ratio, and elastic modulus which are so sensitive to the transmission and reflection signals are applied to the ANN as inputs and the reflection and transmission signals are obtained as outputs. These parameters have been obtained previously in the literature by employing solely longitudinal ultrasonic waves probably due to their easier practical application. It should be noted that the input data applied to the network contain a number of 1,250,000 representing 7 features which six of them are relevant to the inputs and one of them is related to the output. The set of input data is selected randomly and applied to the ANN and the transmitted and reflected waves according to those input data obtained from the theoretical model and applied to the ANN as the output. Subsequently, the network will be learned after many simulations between these inputs and output data. The reflected and transmitted waves for different porosities are considered and the results show an excellent agreement with analytical solution and experimental data from the pertinent literature.</p>

Simulation of Wave Propagation in Biomimetic Porous Scaffold Using Artificial Neural Network

Paper Proceedings: https://asmedigitalcollection.asme.org/IMECE/IMECE2021/volume/85543

ASME 2021 International Mechanical Engineering Congress and Exposition (IMECE2021)

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.