PhD Scientific Days 2024

Poster Session B - Pharmaceutical Sciences and Health Technologies 1.

Real-time Particle Size Measurement During the Pellet Layering Process Using Artificial Intelligence-aided Endoscopic Imaging

Előadó neve

Ms. Orsolya, Péterfi

Neptun code

T78T70

Előadó munkahelye

Department of Organic Chemistry and Technology, Faculty of Chemical Technology and Biotechnology, Budapest University of Technology and Economics

Előadó telefonszáma

06707724734

Előadó e-mail címe

peterfiorsolya@yahoo.com

Az előadás címe

Real-time Particle Size Measurement During the Pellet Layering Process Using Artificial Intelligence-aided Endoscopic Imaging

Szerző(k) neve és munkahelye

Orsolya Péterfi¹, Kincső Demeter¹, Nikolett Kállai-Szabó², Ádám-Tibor Barna², István Antal², Zsombor Kristóf Nagy¹, Dorián László Galata¹

1: Department of Organic Chemistry and Technology, Faculty of Chemical Technology and Biotechnology, Budapest University of Technology and Economics
2: Department of Pharmaceutics, Semmelweis University

Bemutatás módja

Poszter

Szekció

Poster Session B - Pharmaceutical Sciences and Health Technologies 1.

Language of the presentation

English

Preferred session

Pharmaceutical Sciences and Health Technologies

Összefoglaló szövege

Introduction: Particle size must be monitored during pellet layering in order to ensure the quality of the final product. In-line monitoring of the pellet layering process provides real-time information about pellet size and layer uniformity, enabling timely intervention in the case of out-of-specification products.
Aims: The aim of our work was to develop an artificial intelligence-based image analysis tool to monitor particle size during the pellet layering process in real-time.
Method: The key components of the imaging system are a rigid endoscope, a light source and a high-speed camera. For image processing, we employed convolutional neural networks (CNN) to detect the pellets and determine their particle size. After training the AI-based model, the developed imaging system was implemented in-line during the pellet layering of microcrystalline cellulose (MCC) pellet cores. The binder solution contained ibuprofen and hydroxypropyl methylcellulose (HPMC), which were used as a model drug and binder, respectively. Reference samples collected at various stages of the drug layering process were analysed with off-line dynamic image analysis and laser diffraction.
Results: The convolutional neural network-based model proved capable of detecting the particles in focus despite the dense material flow. The in-line measurements and off-line reference methods showed similar trends. During pellet layering, undesired particle agglomeration might occur, which affects the quality of the final product. The trained AI-based model was also able to detect pellet agglomeration during in-line process monitoring.
Conclusion: The developed imaging system is promising; in-line measurements showed good correlation with the reference methods. The CNN-based system is highly feasible as a process analytical technology (PAT) tool for monitoring particle size in real-time during fluidized bed processes.
Funding: The research has been implemented with the support provided by the Agency for Credits and Study Grants coordinated by the Romanian Ministry of National Education. Further support was received the Doctoral Excellence Fellowship Programme (DCEP) funded by the National Research Development and Innovation Fund of the Ministry of Culture and Innovation of Hungary and the Budapest University of Technology and Economics.

University

Other, please specify in the next box

Other university, not listed above

Budapest University of Technology and Economics

Supervisor

Zsombor Kristóf Nagy

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

Kind

Szabad

Status

elfogadva

Accepted presentation method

poszter

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

8092

Start

15:10

End

15:13