The study of wave propagation in biomimetic porous scaffold requires the inclusion of some complex physics such as the interaction of the ultrasonic wave with pore fluid, solid phase, and porous material. Also, due to viscous interactions between the pore fluid and skeletal frame, the dynamic tortuosity as a fractional function of frequency in the clinically relevant ultrasound frequency range is considered. The bone scaffold here is simulated using a porous slab whose two dimensions are infinite. The Biot-JKD theory used for wave propagation in porous media is conditioned with many physical parameters. Solving such governing equations for complex multi-physics problems is computationally expensive. Therefore, developing efficient tools and numerical methods to address multi-physics problems is appealing. Artificial Neural Network (ANN) can efficiently solve convoluted-parametric problems. The purpose of this research is to propose a physics-aware ANN to simulate wave propagation in bone scaffold filled with a viscous fluid. A set of data including porosity, viscosity, tortuosity, viscous characteristics length, Poisson’s ratio, and elastic modulus which are sensitive to the transmission and reflection signals are applied to the ANN as inputs and the reflection and transmission signals are obtained as outputs. The reflected and transmitted waves for different porosities are considered and the results show an excellent agreement with the proposed analytical theory and experimental data found in the literature.