Citation: Tchórz, A.; Korona, K.; Krzak, I.; Bitka, A.; Ksi ˛ a˙ zek, M.; Ja´ skowiec, K.; Malysza, M.; Glowacki, M.; Wilk-Kolodziejczyk, D. Development of a CT Image Analysis Model for Cast Iron Products Based on Artificial Intelligence Methods. Materials 2022, 15, 8254. https:// doi.org/10.3390/ma15228254 Academic Editor: Francesco Iacoviello Received: 12 October 2022 Accepted: 10 November 2022 Published: 21 November 2022 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations. Copyright: © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). materials Article Development of a CT Image Analysis Model for Cast Iron Products Based on Artificial Intelligence Methods Adam Tchórz 1, *, Krzysztof Korona 2 , Izabela Krzak 1 , Adam Bitka 1 , Marzanna Ksi ˛ a˙ zek 3 , Krzysztof Ja´ skowiec 1 , Marcin Malysza 1 , Miroslaw Glowacki 2,4 and Dorota Wilk-Kolodziejczyk 1,2 1 Lukasiewicz Research Network–Krakow Institute of Technology, Zakopia´ nska 73, 30-418 Krakow, Poland 2 Faculty of Metals Engineering and Industrial Computer Science, AGH University of Science and Technology, Al. Mickiewicza 30, 30-059 Krakow, Poland 3 Faculty of Non-Ferrous Metals, AGH University of Science and Technology, Al. Mickiewicza 30, 30-059 Krakow, Poland 4 Faculty of Natural Sciences, Jan Kochanowski University of Humanities and Sciences in Kielce, Ul. ˙ Zeromskiego 5, 25-369 Kielce, Poland * Correspondence: adam.tchorz@kit.lukasiewicz.gov.pl Abstract: This paper presents an assessment of the possibility of using digital image classifiers for tomographic images concerning ductile iron castings. The results of this work can help the development of an efficient system suggestion allowing for decision making regarding the qualitative assessment of the casting process parameters. Special attention should be focused on the fact that automatic classification in the case of ductile iron castings is difficult to perform. The biggest problem in this aspect is the high similarity of the void image, which may be a sign of a defect, and the nodular graphite image. Depending on the parameters, the tests on different photos may look similar. Presented in this article are test scenarios of the module analyzing two-dimensional tomographic images focused on the comprehensive assessment by convolutional neural network models, which are designed to classify the provided image. For the purposes of the tests, three such models were created, different from each other in terms of architecture and the number of hyperparameters and trainable parameters. The described study is a part of the decision-making system, supporting the process of qualitative analysis of the obtained cast iron castings. Keywords: cast iron; 3D tomography for cast metal; recommendation system; neural networks; defect analysis 1. Introduction The analysis of the image obtained on the basis of an examination carried out with the use of a computer tomograph (CT) allows to estimate the probability of occurrence of defects in the tested castings. The CT is a non-destructive method that allows you to evaluate the product without damaging it (in some cases, because sometimes the sample needs to be cut out, it depends on the thickness of the casting walls and the studied material). The interpretation of the picture is performed by experts. The obtained image is not unambiguous. Its analysis requires a lot of experience, especially in the case of iron castings. A CT tool to take such images and evaluate the information obtained from their manual interpretation was used in this study. The “manual” interpretation of images is performed on a daily basis. In our article we attempted to develop a model using the methods of artificial intelligence, which allows to assess whether the visible area on the photo is a void, crack, or some other detail. It should be noted that the methods of artificial intelligence were used to predict properties as well as evaluate and classify the results of research on metal products. A significant problem when using this type of solution is the amount of data. The use of artificial intelligence or machine learning as well as data analysis to develop predictive models to determine the mechanical Materials 2022, 15, 8254. https://doi.org/10.3390/ma15228254 https://www.mdpi.com/journal/materials