Dokument: A new validation concept for in silico tools using X-ray micro computed tomography images of pharmaceutical formulations
| Titel: | A new validation concept for in silico tools using X-ray micro computed tomography images of pharmaceutical formulations | |||||||
| URL für Lesezeichen: | https://docserv.uni-duesseldorf.de/servlets/DocumentServlet?id=59807 | |||||||
| URN (NBN): | urn:nbn:de:hbz:061-20220620-105501-2 | |||||||
| Kollektion: | Dissertationen | |||||||
| Sprache: | Englisch | |||||||
| Dokumententyp: | Wissenschaftliche Abschlussarbeiten » Dissertation | |||||||
| Medientyp: | Text | |||||||
| Autor: | Bollmann, Sebastian [Autor] | |||||||
| Dateien: |
| |||||||
| Beitragende: | Prof. Dr. Dr. h.c. Kleinebudde, Peter [Gutachter] Prof. Dr. Breitkreutz, Jörg [Gutachter] | |||||||
| Dewey Dezimal-Klassifikation: | 600 Technik, Medizin, angewandte Wissenschaften » 610 Medizin und Gesundheit | |||||||
| Beschreibung: | This thesis addresses a new concept to validate in silico tools for full in silico formulation development and to investigate and improve either the performance of the in silico tool or the pharmaceutical dosage form. The basis of this concept is the non-destructive 3D imaging of pharmaceutical dosage forms using XμCT measurements. The image processing of the obtained 3D images was evaluated and presented in detail. Both the effort required for the application of this concept and some crucial conditions for the applicability of the methodology were depicted. A compromise between effort and specification of the image processing could be found, which led to high performances of the image processing. Quantitative segmentation was achieved and the required virtual matrices representing real pharmaceutical dosage forms were obtained by using the desirability to evaluate the pathway performances of image processing. However, the desirability calculation should be optimised as visual verification was mandatory due to high performing pathways that contained the correct amount of each compound with an incorrect distribution pattern. The coefficient of variation determined for grey values of images was an appropriate method to reduce the evaluation effort for image processing pathways within calibrated ranges. This could accelerate future image processing evaluations of similar samples. The concept identified incorrect calculations of the diffusion process in a dissolutionsimulation of different tablet batches of an example in silico tool. It was also shown that the virtual matrices designed by the software did not reflect real pharmaceutical dosage forms. In spite of these issues, the software was able to correctly predict the dissolution of tablets when the dissolution profile was based on the dissolution rate of the API. Feedback from these experiments has been and will be used to develop software updates to resolve identified issues. In addition to the ability to improve in silico tools or pharmaceutical formulations, these experiments demonstrate that the concept is capable of validating in silico tools. It was possible to benchmark the performance of the in silico tool by comparing the dissolution of virtual matrices representing real pharmaceutical dosage forms based on processed XμCT images with the experimental results of the imaged dosage form. Thus, a direct comparison of different in silico tools or the approval assessment by authorities would also be possible with this concept. In principle, the concept can be transferred from tablets to oral films. However, the images of the oral films were disturbed by artefacts. Air bubbles within the films indicated a Pickering emulsion-like system in the raw images. This could be disproved by confocal Raman microscopy. The processed images and the resulting virtual matrices reflected these artefacts. This emphasises the importance of checking the plausibility of the non-destructive 3D images before subsequent image processing. If the raw images are plausible and an appropriate image processing pathway has been applied, a virtual matrix representing the real pharmaceutical dosage form could be used to validate, improve or develop in silico tools or formulations to predict the performance of pharmaceutical dosage forms in quality control tests. However, this work is only a starting point and further investigations of the new concept are imperative to evaluate its applicability and limitations in more detail. | |||||||
| Lizenz: |  Urheberrechtsschutz | |||||||
| Fachbereich / Einrichtung: | Mathematisch- Naturwissenschaftliche Fakultät » WE Pharmazie » Pharmazeutische Technologie und Biopharmazie | |||||||
| Dokument erstellt am: | 20.06.2022 | |||||||
| Dateien geändert am: | 20.06.2022 | |||||||
| Promotionsantrag am: | 16.03.2022 | |||||||
| Datum der Promotion: | 24.05.2022 |
