Related Experiment Video
Updated: Aug 6, 2026
11:04
Control of Cell Geometry through Infrared Laser Assisted Micropatterning
Published on: July 10, 2021
Machine Learning-Guided Near-Infrared Laser Processing for Precise Geometry Control in Microneedles with Downstream
Ismail Eş1,2,3, Burak M Gormus4, Pinar Ayas5
1Institute of Biomedical Engineering, Old Road Campus Research Building, University of Oxford, Oxford, UK.
Small (Weinheim an Der Bergstrasse, Germany)
|July 16, 2026
Summary
Machine learning precisely fabricates microneedles (MNs) for transdermal applications. This data-driven approach ensures accurate MN geometry for drug delivery and sensing, reducing trial-and-error.
Area of Science:
- Biomedical Engineering
- Materials Science
- Laser Processing
Background:
- Transdermal drug delivery and sensing offer non-invasive methods for localized treatment and biomarker monitoring.
- Microneedles (MNs) provide pain-free subdermal access, but precise control over their geometry is crucial due to skin variations.
- Current MN fabrication often involves empirical trial-and-error, limiting efficiency and customization.
Purpose of the Study:
- To develop a machine learning-guided predictive fabrication method for microneedles (MNs) using near-infrared (NIR) laser processing.
- To establish a data-driven pipeline for designing application-specific MN geometries with high accuracy.
- To demonstrate the translation of predictive metal masters into functional polymer MN devices.
Main Methods:
- Systematic mapping of laser parameters (power, repetition) and computer-aided design (CAD) dimensions for duralumin processing.
- Training a data-driven model to predict MN height and tip radius based on fabrication inputs.
- Fabrication of polymer MNs from predictive metal masters and integration with a micro-light-emitting diode (LED) platform.
Main Results:
- High accuracy achieved in predicting MN height and tip radius, with a 1.24% deviation for a target height of 524 µm on the female chin.
- Reliable replication and geometric fidelity demonstrated in polymer MNs fabricated from predictive metal masters.
- Successful integration of polyvinyl alcohol (PVA) MNs with NIR-responsive dye and a flexible micro-LED for localized photothermal heating.
Conclusions:
- An end-to-end, data-driven pipeline for laser-based MN fabrication is established.
- The approach enables scalable, application-specific MN design for transdermal drug delivery, sensing, and bioelectronic interfaces.
- This method reduces empirical fabrication steps, enhancing efficiency and precision for advanced microneedle devices.
