Abstract Scope |
An emerging platform for tissue engineering, extrusion-based (bio)printing, allows for the rapid fabrication of scaffolds with clinically relevant dimensions but suffers from limited resolution and print defect occurrence. Print defects risk the functionality of printed tissue engineering scaffolds, making it vital to negate defect occurrence. As such, the goal of this work is the defect-free biological additive manufacturing of tissue constructs (Bio-AM). In the context of this goal, the in-process detection of flaw formation is the critical first step toward the clinical scaling of Bio-AM processes. The objective of this work was to detect flaw formation in Bio-AM of bone tissue constructs as they develop using data from in-situ infrared thermocouple sensors. The data was analyzed using several machine learning approaches to ascertain critical quality metrics: print regime, strand width, strand height, and strand fusion severity. Defects were classified and predicted using this approach with statistical accuracy nearing 90%. |