• Home
  • News & Blog
  • Quality control for 3D printing in the clinic: Machine Learning and Machine Vision pave the way for personalised orodispersible films

Written by Colm O’Reilly

Personalised ODF Quality Control

Orodispersible films are thin polymeric films that dissolve rapidly in the mouth. The dose can be adjusted by changing the film dimensions and the films are easy to swallow. These qualities make orodispersible films excellent candidates for personalised medicine but conventional orodispersible film manufacturing methods are not suitable for the clinical setting. 3D printing can manufacture personalised orodispersible films rapidly. However, 3D printing lacks quality control suitable for small, personalised batches. Near infrared spectroscopy can rapidly analyse samples in a non-destructive manner. The interpretation of these results would present a challenge in a hectic clinical environment. Machine learning models could be trained to interpret these results.

3D printing, Near Infrared Spectroscopy, Machine Learning, and Machine Vision

In this study, we demonstrated how 3D printing, near infrared spectroscopy and machine learning can be combined to rapidly manufacture orodispersible films and verify their drug and dose. Additionally, a machine vision tool analysed the in vitro disintegration of the orodispersible films. 3D printing manufactured complex shapes, with thicknesses below 100 µm. A supervised machine learning model, linear discriminant analysis, classified the drug present in the orodispersible film with 100% accuracy. The doses were verified using a regression model, partial least square, and the R2 values were all above 0.96.

Near spectroscopy, machine learning, and machine vision show potential to automate the research workflow and facilitate the transition of 3D printing into clinical settings.

Read our open access research article at: