MDPI Acta Medica Okayama 2218-273X 11 6 2021 Multi-Task Deep Learning Model for Classification of Dental Implant Brand and Treatment Stage Using Dental Panoramic Radiograph Images 815 EN Shintaro Sukegawa Dentistry and Pharmaceutical Sciences, Department of Oral Pathology and Medicine, Okayama University Graduate School of Medicine Kazumasa Yoshii Department of Intelligence Science and Engineering, Graduate School of Natural Science and Technology, Gifu University Takeshi Hara Department of Intelligence Science and Engineering, Graduate School of Natural Science and Technology, Gifu University Tamamo Matsuyama Department of Oral and Maxillofacial Surgery, Kagawa Prefectural Central Hospital Katsusuke Yamashita Polytechnic Center Kagawa Keisuke Nakano Dentistry and Pharmaceutical Sciences, Department of Oral Pathology and Medicine, Okayama University Graduate School of Medicine Kiyofumi Takabatake Dentistry and Pharmaceutical Sciences, Department of Oral Pathology and Medicine, Okayama University Graduate School of Medicine Hotaka Kawai Dentistry and Pharmaceutical Sciences, Department of Oral Pathology and Medicine, Okayama University Graduate School of Medicine Hitoshi Nagatsuka Dentistry and Pharmaceutical Sciences, Department of Oral Pathology and Medicine, Okayama University Graduate School of Medicine Yoshihiko Furuki Department of Oral and Maxillofacial Surgery, Kagawa Prefectural Central Hospital It is necessary to accurately identify dental implant brands and the stage of treatment to ensure efficient care. Thus, the purpose of this study was to use multi-task deep learning to investigate a classifier that categorizes implant brands and treatment stages from dental panoramic radiographic images. For objective labeling, 9767 dental implant images of 12 implant brands and treatment stages were obtained from the digital panoramic radiographs of patients who underwent procedures at Kagawa Prefectural Central Hospital, Japan, between 2005 and 2020. Five deep convolutional neural network (CNN) models (ResNet18, 34, 50, 101 and 152) were evaluated. The accuracy, precision, recall, specificity, F1 score, and area under the curve score were calculated for each CNN. We also compared the multi-task and single-task accuracies of brand classification and implant treatment stage classification. Our analysis revealed that the larger the number of parameters and the deeper the network, the better the performance for both classifications. Multi-tasking significantly improved brand classification on all performance indicators, except recall, and significantly improved all metrics in treatment phase classification. Using CNNs conferred high validity in the classification of dental implant brands and treatment stages. Furthermore, multi-task learning facilitated analysis accuracy. No potential conflict of interest relevant to this article was reported.