National and Kapodistrian University of Athens
Design and preparation of solid dosage forms of bioactive compounds and application of Artificial Intelligence/Machine Learning on the prediction of dissolution profile
Abstract
dc:descriptionThe development and optimization of novel pharmaceutical dosage forms for personalized drug delivery constitutes a significant challenge in modern pharmaceutics. Traditional “one-size-fits-all” approaches are increasingly being replaced by patient-centric strategies that aim to tailor therapies according to individual needs. However, conventional manufacturing approaches often lack the flexibility to produce personalized drug products with specific characteristics, such as dose adjustment, geometry, and route of administration, limiting their applicability in personalized medicine. In this context, additive manufacturing (3D printing), has emerged as a promising and versatile platform for the fabrication of customized pharmaceutical systems. The application of these technologies enables the production of drug-loaded formulations with controlled doses and structure. This thesis investigates the integration of advanced pharmaceutical manufacturing technologies and data-driven methodologies for the development of personalized drug delivery systems. The work focuses on the application of 3D printing for the fabrication of personalized oral dosage forms, including tablets and buccal films, with particular emphasis on improving accessibility, handling, identification and personalization for specific patient populations such as visually impaired and paediatric patients. The initial approach to producing personalized tablets involved the use of a direct extrusion 3D printer. In this method, hot-melt extrusion (HME) was first employed to fabricate drug-loaded polymer filaments with adequate mechanical strength, thermal stability, and suitable physicochemical properties, enabling their subsequent use in extrusion-based 3D printing. A range of pharmaceutical polymers and active pharmaceutical ingredients were investigated, and formulation compositions were systematically optimized to achieve desirable printability and performance characteristics. Processing parameters for both HME and 3D printing were carefully controlled and evaluated to ensure reproducibility, dimensional accuracy, and structural integrity of the fabricated dosage forms.The second approach involved the use of a liquid crystal display (LCD) 3D printer, which operates on a photopolymerisation process. In this technique, a UV light source is selectively projected through an LCD screen to mask each layer’s pattern, curing and solidifying the liquid resin in a layer-by-layer manner until the final object is formed. This technique enables the fabrication of highly detailed and precise objects, offering superior resolution compared to extrusion-based printing methods. This technique enables the production of highly detailed and precise structures, compared to extrusion-based 3D printing methods. Considering these advantages, this thesis employed LCD 3D printing for the development of personalized tablets with Braille imprints for visually impaired patients, as well as for the fabrication of buccal films. A third approach involved direct micro-extrusion 3D printing using a BIO X 3D printer. In this method, tablets with personalized doses and sizes for paediatric patients were printed directly from powder blends, eliminating the need for a prior HME step. This reduced the number of manufacturing steps and simplified the fabrication process.Comprehensive physicochemical and structural characterization was conducted to elucidate the relationship between material properties, processing conditions, and final product performance. Techniques such as scanning electron microscopy (SEM), X-ray powder diffraction (XRPD), differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), and X-ray micro-computed tomography (microCT) were employed to assess morphology, crystallinity, thermal behaviour, and internal architecture. Nuclear Magnetic Resonance (NMR) analysis was performed on the formulations produced by LCD 3D printing both before and after printing to confirm the complete polymerization of the toxic monomers. These analyses revealed that variations in formulation and manufacturing parameters significantly influence porosity, drug dispersion, and the formation of amorphous or crystalline phases, which in turn affect drug release mechanisms.The drug release behaviour of the developed systems was evaluated through in vitro dissolution studies, demonstrating the capability to tailor dissolution profiles by modifying formulation composition, geometrical design, and processing parameters. Also, ex vivo experiments and adhesion tests were conducted on the buccal films to evaluate drug permeability and the mucoadhesive strength of the printed formulations. Both immediate and sustained-release profiles were achieved through the printing of tablets and buccal films, highlighting the versatility of additive manufacturing in producing personalized dosage forms. Furthermore, kinetic modelling and mathematical analysis were applied to describe drug release mechanisms and to support the interpretation of dissolution data. Another aspect of the present thesis involved the application of artificial intelligence (AI) to predict dissolution profiles, an essential test for every formulation used to evaluate drug release. Direct compression (DC) remains a widely used manufacturing technology for producing solid dosage forms; however, formulation optimisation is a labour-intensive, costly, and time-consuming process. The aim of this study was to investigate whether machine learning (ML) can be used to accelerate development by predicting drug release profiles under dynamic dissolution conditions based on formulation composition. A total of 377 formulations were produced in-house, and their release profiles were measured under dynamic dissolution conditions from 0 to 480 minutes across multiple time points.A ML analysis was then conducted to predict the complete drug release profiles. In addition, an alternative modelling strategy was explored in which ML was used to predict kinetic parameters from established release models, which were then used to reconstruct the dissolution profiles. The findings demonstrate that ML can be effectively applied to predict entire drug release profiles under dynamic conditions while also providing access to underlying kinetic information. Overall, this thesis demonstrates the synergistic potential of combining additive manufacturing, materials engineering, and AI in advancing pharmaceutical science and personalized medicine. The findings contribute to the development of flexible, patient-centred pharmaceutical systems and highlight the role of emerging technologies in transforming conventional drug manufacturing into a more adaptable, efficient, and data-driven process.
Degree
thesis:*- Grantor dc:publisher
- National and Kapodistrian University of Athens
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Protopapa, Chrystalla
- Πρωτοπαπά, Χρυστάλλα
Subjects
dc:subject × 14- Φαρμακευτική τεχνολογία
- Τρισδιάστατη εκτύπωση (3D printing)
- Ανάπτυξη φαρμάκων
- Τεχνιτή νοημοσύνη
- Pharmaceutical technology
- 3d printing
- Formulation
- Artificial Intelligence (AI)
- Επιστήμες Υγείας
- Ιατρική και Επιστήμες Υγείας
- Επιστήμες υγείας, άλλοι τομείς
- Health Sciences
- Medical and Health Sciences
- Health sciences, miscellaneous
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/63138