@inproceedings{MTMT:34567948, title = {In vitro Dissolution Prediction using Deep Convolutional Neural Networks}, url = {https://m2.mtmt.hu/api/publication/34567948}, author = {Knyihár, Gábor and Csorba, Kristóf}, booktitle = {Proceedings of the Automation and Applied Computer Science Workshop 2023 (AACS'23)}, unique-id = {34567948}, year = {2023}, pages = {27-38} } @article{MTMT:33604474, title = {Evaluating the Accuracy of a Linear Regression Model in Predicting the Dissolution of Tablets based on Raman Maps}, url = {https://m2.mtmt.hu/api/publication/33604474}, author = {Knyihár, Gábor and Csorba, Kristóf and Charaf, Hassan}, doi = {10.5121/sipij.2023.14102}, journal-iso = {SIPIJ}, journal = {SIGNAL & IMAGE PROCESSING: AN INTERNATIONAL JOURNAL}, volume = {14}, unique-id = {33604474}, issn = {2229-3922}, abstract = {Investigation of the dissolution of tablets is an important area of pharmaceutical research. Such research aims to predict the dissolution process as accurately as possible without destroying the tablets. Several methods have been published that can estimate dissolution with approximate accuracy, but they are primarily complex learning algorithms that are time-consuming and require many samples to train. This article seeks to answer whether these complex models are necessary or whether a similar result can be achieved with the help of more straightforward methods. Therefore, during this work, a simpler linear regression model was created and analysed its effectiveness in estimating the dissolution curves. The investigation concluded that the results are not as accurate as in the case of more complex methods, but the model is more robust and can be used in the case of fewer samples. Thus, by further developing these and combining the methods, we can achieve better results in the future.}, keywords = {Linear regression; Principal component analysis (PCA); RAMAN spectroscopy; Dissolution curve}, year = {2023}, eissn = {0976-710X}, pages = {23-33} } @inproceedings{MTMT:33604424, title = {Tabletták kioldódásának előrejelzése Raman térképek alapján lineáris regressziós modell segítségével}, url = {https://m2.mtmt.hu/api/publication/33604424}, author = {Knyihár, Gábor and Csorba, Kristóf and Charaf, Hassan}, booktitle = {Hungarian Association for Image Analysis and Pattern Recognition - 14th Conference}, unique-id = {33604424}, abstract = {A tabletták oldódásának vizsgálata a gyógyszerészeti kutatások egyik fontos területe. Az ilyen kutatások célja az oldódás folyamatának előrejelzése a lehető legpontosabban a tabletták roncsolása nélkül. Számos olyan módszert publikáltak már, ami közelítő pontossággal meg tudja becsülni a kioldódást, de ezek többnyire olyan összetett tanuló algoritmusok, amelyek időigényesek és betanításukhoz nagy számú mintára van szükség. Ez a cikk arra a kérdésre keresi a választ, hogy szükségeseke ezek az összetett modellek, vagy egyszerűbb módszerek segítségével is elérhető hasonló eredmény. A munka során ezért egy egyszerűbb lineáris regressziós modellt készítettünk, és ennek hatékonyságát mértük a kioldódási görbék becslésében. A vizsgálattal arra a következtetésre jutottunk, hogy az eredmények ugyan nem annyira pontosak, mint a komplexebb módszerek esetében, ellenben a modell robosztusabb és kevesebb minta esetében is használható. Így a jövőben ezek továbbfejlesztésével és a módszerek kombinálásával akár jobb eredményeket is elérhetünk.}, year = {2023} } @inproceedings{MTMT:33543170, title = {Predicting the Dissolution of Tablets based on Raman Maps using a Linear Regression Model}, url = {https://m2.mtmt.hu/api/publication/33543170}, author = {Knyihár, Gábor and Csorba, Kristóf and Charaf, Hassan}, booktitle = {Computer Science and Machine Learning Trends 2023}, doi = {10.5121/csit.2023.130107}, unique-id = {33543170}, abstract = {Investigation of the dissolution of tablets is an important area of pharmaceutical research. Such research aims to predict the dissolution process as accurately as possible without destroying the tablets. Several methods have been published that can estimate dissolution with approximate accuracy, but they are mostly complex and time-consuming. This article seeks to answer whether these complex models are necessary or whether a similar result can be achieved with the help of more straightforward methods. Therefore, during this work, a simpler linear regression model was created and analysed its effectiveness in estimating the dissolution curves. The investigation concluded that the results are not as accurate as in the case of more complex methods, but they are not far behind. Thus, even similar results may be achieved by fine-tuning and possibly developing these methods.}, keywords = {Linear regression; Principal component analysis (PCA); RAMAN spectroscopy; Dissolution curve}, year = {2023}, pages = {79-85} } @inproceedings{MTMT:34476285, title = {Droplet Based Prediction of Viscosity of Water-PVP Solutions Using Convolutional Neural Networks}, url = {https://m2.mtmt.hu/api/publication/34476285}, author = {Mrad, Mohamed Azouz and Csorba, Kristóf and Galata, Dorián László and Nagy, Zsombor Kristóf and Nagy, Brigitta}, booktitle = {Proceedings of the Automation and Applied Computer Science Workshop 2023 (AACS'23)}, unique-id = {34476285}, abstract = {The viscosity of a liquid is the property that measures the liquid internal resistance to flow. Monitoring viscosity is a vital component of quality control in several industrial fields including chemical, pharmaceutical, food, and energy-related industries. The most commonly used instrument for measuring viscosity is capillary viscometers, but their cost and complexity pose challenges for industries where accurate and real-time viscosity information is vital. In this work, we prepared thirteen solutions with different water and PVP (Polyvinylpyrrolidone) ratios, measured their different viscosity values, and produced videos of their droplets. We extracted the images of the fully developed droplets from the videos and we used the images to train a Convolutional neural network model to estimate the viscosity values of the WaterPVP solutions. The proposed model was able to predict the viscosity values of the samples using images of their droplets with a high accuracy on the test dataset.}, year = {2023}, pages = {1-15}, orcid-numbers = {Mrad, Mohamed Azouz/0000-0001-5223-3680} } @article{MTMT:34045465, title = {Droplet Based Estimation of Viscosity of Water–PVP Solutions Using Convolutional Neural Networks}, url = {https://m2.mtmt.hu/api/publication/34045465}, author = {Mrad, Mohamed Azouz and Csorba, Kristóf and Galata, Dorián László and Nagy, Zsombor Kristóf and Charaf, Hassan}, doi = {10.3390/pr11071917}, journal-iso = {PROCESSES}, journal = {PROCESSES}, volume = {11}, unique-id = {34045465}, issn = {2227-9717}, abstract = {The viscosity of a liquid is the property that measures the liquid’s internal resistance to flow. Monitoring viscosity is a vital component of quality control in several industrial fields, including chemical, pharmaceutical, food, and energy-related industries. In many industries, the most commonly used instrument for measuring viscosity is capillary viscometers, but their cost and complexity pose challenges for these industries where accurate and real-time viscosity information is vital. In this work, we prepared fourteen solutions with different water and PVP (Polyvinylpyrrolidone) ratios, measured their different viscosity values, and produced videos of their droplets. We extracted the images of the fully developed droplets from the videos and we used the images to train a convolutional neural network model to estimate the viscosity values of the water–PVP solutions. The proposed model was able to accurately estimate the viscosity values of samples of unseen chemical formulations with the same composition with a low MSE score of 0.0243 and R2 score of 0.9576. The proposed method has potential applications in scenarios where real-time monitoring of liquid viscosity is required.}, year = {2023}, eissn = {2227-9717}, orcid-numbers = {Mrad, Mohamed Azouz/0000-0001-5223-3680} } @inproceedings{MTMT:33605280, title = {Viscosity Estimation of Water-PVP Solutions from Droplets using Artificial Neural Networks and Image Processing}, url = {https://m2.mtmt.hu/api/publication/33605280}, author = {Mrad, Mohamed Azouz and Csorba, Kristóf and Galata, Dorián László and Nagy, Zsombor Kristóf and Nagy, Brigitta}, booktitle = {Artificial Intelligence and Soft Computing}, doi = {10.1007/978-3-031-42505-9_14}, unique-id = {33605280}, year = {2023}, pages = {157-166}, orcid-numbers = {Mrad, Mohamed Azouz/0000-0001-5223-3680} } @inproceedings{MTMT:33595245, title = {Spectroscopy-Based Prediction of In Vitro Dissolution Profile Using Random Decision Forests}, url = {https://m2.mtmt.hu/api/publication/33595245}, author = {Mrad, Mohamed Azouz and Csorba, Kristóf and Galata, Dorián László and Nagy, Zsombor Kristóf and Nagy, Brigitta}, booktitle = {Artificial Intelligence and Soft Computing}, doi = {10.1007/978-3-031-23492-7_35}, unique-id = {33595245}, year = {2023}, pages = {411-422}, orcid-numbers = {Mrad, Mohamed Azouz/0000-0001-5223-3680} } @article{MTMT:32793973, title = {Automated Classification of Left Ventricular Hypertrophy on Cardiac MRI}, url = {https://m2.mtmt.hu/api/publication/32793973}, author = {Budai, Ádám and Suhai, Ferenc Imre and Csorba, Kristóf and Dohy, Zsófia and Szabó, Liliána and Merkely, Béla Péter and Vágó, Hajnalka}, doi = {10.3390/app12094151}, journal-iso = {APPL SCI-BASEL}, journal = {APPLIED SCIENCES-BASEL}, volume = {12}, unique-id = {32793973}, abstract = {Left ventricular hypertrophy is an independent predictor of coronary artery disease, stroke, and heart failure. Our aim was to detect LVH cardiac magnetic resonance (CMR) scans with automatic methods. We developed an ensemble model based on a three-dimensional version of ResNet. The input of the network included short-axis and long-axis images. We also introduced a standardization methodology to unify the input images for noise reduction. The output of the network is the decision whether the patient has hypertrophy or not. We included 428 patients (mean age: 49 ± 18 years, 262 males) with LVH (346 hypertrophic cardiomyopathy, 45 cardiac amyloidosis, 11 Anderson–Fabry disease, 16 endomyocardial fibrosis, 10 aortic stenosis). Our control group consisted of 234 healthy subjects (mean age: 35 ± 15 years; 126 males) without any known cardiovascular diseases. The developed machine-learning-based model achieved a 92% F1-score and 97% recall on the hold-out dataset, which is comparable to the medical experts. Experiments showed that the standardization method was able to significantly boost the performance of the algorithm. The algorithm could improve the diagnostic accuracy, and it could open a new door to AI applications in CMR.}, year = {2022}, eissn = {2076-3417}, orcid-numbers = {Budai, Ádám/0000-0003-0454-5158; Suhai, Ferenc Imre/0000-0002-5268-3873; Dohy, Zsófia/0000-0002-0706-5179; Szabó, Liliána/0000-0002-4699-3648; Merkely, Béla Péter/0000-0001-6514-0723; Vágó, Hajnalka/0000-0002-3568-3572} } @article{MTMT:32601462, title = {Real-Time Monitoring of Continuous Pharmaceutical Mixed Suspension Mixed Product Removal Crystallization Using Image Analysis}, url = {https://m2.mtmt.hu/api/publication/32601462}, author = {Domokos, András and Madarász, Lajos and Stoffán, György Nimród and Tacsi, Kornélia and Galata, Dorián László and Csorba, Kristóf and Vass, Panna and Nagy, Zsombor Kristóf and Pataki, Hajnalka}, doi = {10.1021/acs.oprd.1c00372}, journal-iso = {ORG PROCESS RES DEV}, journal = {ORGANIC PROCESS RESEARCH & DEVELOPMENT}, volume = {26}, unique-id = {32601462}, issn = {1083-6160}, abstract = {In this work, we developed an in-line image analysis system for the monitoring of the continuous crystallization of an active pharmaceutical ingredient. Acetylsalicylic acid was crystallized in a mixed suspension mixed product removal crystallizer, which was equipped with overflow tubing as an outlet. A steep glass plate was placed under the outlet onto which the slurry dripped on its surface. The glass plate spread and guided the droplets toward the product collection filter. A high-speed process camera was mounted above the glass plate to capture images of the crystals. Several light sources were tested in various positions to find the appropriate experimental setup for the optimal image quality. Samples were taken during continuous operation for off-line particle size analysis in order to compare to the crystal size distributions calculated from the images. The results were in good agreement, and the trends of the process could be followed well using the images. As a next step, image analysis was operated throughout the entire continuous crystallization experiment, and a huge quantity of information was collected from the process. The crystal size distribution of the product was calculated every 30 s, which provided a thorough and detailed insight into the crystallization process.}, year = {2022}, eissn = {1520-586X}, pages = {149-158}, orcid-numbers = {Domokos, András/0000-0003-1968-4679; Vass, Panna/0000-0003-2206-565X} }