Automated calibration of 3D-printed microfluidic devices based on computer vision

Area of Science:

• * Materials Science and Engineering
• * Biomedical Engineering
• * Additive Manufacturing

Background:

• * 3D printing is increasingly valuable for microfluidic/Lab-on-a-Chip (LoC) fabrication.
• * Achieving high-quality prints often requires extensive calibration, even with advanced printers.
• * Current calibration methods are time-consuming and hinder the adoption of 3D printing in microfluidics.

Purpose of the Study:

• * To develop a rapid and precise computer vision (CV)-based method for 3D printing calibration.
• * To enhance the efficiency of calibration for average-priced 3D printers.
• * To promote the use of 3D printing technology within the microfluidic research community.

Main Methods:

• * An automated, five-step computer vision (CV) method utilizing a camera and convolutional neural network recognition.
• * Calibration demonstrated on cylindrical features (0.2–2.4 mm diameter) and rectangular features (0.2–1.0 mm width) using a stereolithography-based 3D printer.
• * Open-source software and calibration ruler design files provided for customization and broad applicability.

Main Results:

• * The CV-based method achieved accurate calibrated dimensions in a single print of the calibration ruler.
• * Higher photo resolution directly correlated with increased calibration precision.
• * Only one additional print was required for the target microfluidic/LoC device after calibration.

Conclusions:

• * The developed CV method offers a quick and precise calibration tool for 3D printing microfluidic/LoC devices.
• * This approach lowers the barrier to entry for researchers using affordable 3D printers.
• * The open-source nature of the software and design files allows for adaptable calibration across various 3D printers and custom needs.