Affiliations: [a] Hubei Key Laboratory of Intelligent Convey Technology and Device (Hubei Polytechnic University), Huangshi, Hubei, China | [b] School of Mechanical and Electrical Engineering, Hubei Polytechnical University, Huangshi, Hubei, China
Correspondence:
[*]
Corresponding author: Shu Wang, School of Mechanical and Electrical Engineering, Hubei Polytechnical University, Huangshi, Hubei 435003, China. E-mail: [email protected].
Abstract: Based on the development of the 3D vascular printer, the forming process of ink from the nozzle to the rotating rod was studied. In this study, to online detect the ink flow from the nozzle during 3D bioprinting of tubular tissue, we established a geometric model according to the region of interest (ROI) of the ink flow picture of 3D printing of tubular tissue, selected description features of the ink contour, and studied how to select mathematical expressions of the features. Principal component analysis (PCA) was used to simplify the image features into 15 features. We used a back propagation (BP) neural network to predict the printing ink flow. The results show that the error between the actual ink flow rate and the flow rate based on the BP neural network is within 5%. The BP neural network can be used to monitor the quality status of the printing target in real time, evaluate the 3D bioprinting quality online, and predict the printing ink flow for the subsequent improvement of the 3D bioprinting accuracy of tubular tissue.
Keywords: Tubular tissue, 3D bioprinting, Ink flow, Back propagation neural network, Prediction algorithm
