Evaluation of different pre-processing methods of X-ray micro computed tomography images

Adequate images of real tablets provide the possibility to validate in silico software tools like F-CAD. The goal of this work was to evaluate the influence of different pre-processing methods of X-ray micro-computed tomography images of 12 different tablet batches. The images were processed with the open source software ImageJ. Different bleach- and lightning corrections were applied.

Highlights

Tablets of 12 binary mixtures with different contrast were analysed.

Desirability was chosen to compare the preprocessing pathways.

Suitable preprocessing pathways could be identified for each binary mixture.

No pathway was suitable for all binary mixtures.

Preprocessing and segmentation of μCT-images was performed with ImageJ

The influence of a filtering process and two different contrast enhancement methods were evaluated. The images were segmented with the k-means algorithm and F-CAD determined the recovery rates. All results were analysed in terms of their desirability. Suitable pathways were found for each binary mixture. The mean recovery rates for all batches were approximately 100%. However, it was not possible to find a suitable program procedure to cover all binary mixtures simultaneously. Nevertheless, it was possible to confirm the possibility of a suitable validation method for in silico tools.

Continue on evaluation of different pre-processing methods of X-ray micro computed tomography images

Keywords: X-μCT, In silico development, Image processing, Formulation development, Ethyl cellulose (EC, Aqualon N10, lactose monohydrate (Tablettose 80)

Sebastian Bollmann, Peter Kleinebudde, Evaluation of different pre-processing methods of X-ray micro computed tomography images, Powder Technology, Volume 381, 2021, Pages 539-550, ISSN 0032-5910,
https://doi.org/10.1016/j.powtec.2020.11.074.