start-ver=1.4 cd-journal=joma no-vol=30 cd-vols= no-issue=8 article-no= start-page=7412 end-page=7424 dt-received= dt-revised= dt-accepted= dt-pub-year=2023 dt-pub=20230804 dt-online= en-article= kn-article= en-subject= kn-subject= en-title= kn-title=Mean Heart Dose Prediction Using Parameters of Single-Slice Computed Tomography and Body Mass Index: Machine Learning Approach for Radiotherapy of Left-Sided Breast Cancer of Asian Patients en-subtitle= kn-subtitle= en-abstract= kn-abstract=Deep inspiration breath-hold (DIBH) is an excellent technique to reduce the incidental radiation received by the heart during radiotherapy in patients with breast cancer. However, DIBH is costly and time-consuming for patients and radiotherapy staff. In Asian countries, the use of DIBH is restricted due to the limited number of patients with a high mean heart dose (MHD) and the shortage of radiotherapy personnel and equipment compared to that in the USA. This study aimed to develop, evaluate, and compare the performance of ten machine learning algorithms for predicting MHD using a patient's body mass index and single-slice CT parameters to identify patients who may not require DIBH. Machine learning models were built and tested using a dataset containing 207 patients with left-sided breast cancer who were treated with field-in-field radiotherapy with free breathing. The average MHD was 251 cGy. Stratified repeated four-fold cross-validation was used to build models using 165 training data. The models were compared internally using their average performance metrics: F2 score, AUC, recall, accuracy, Cohen's kappa, and Matthews correlation coefficient. The final performance evaluation for each model was further externally analyzed using 42 unseen test data. The performance of each model was evaluated as a binary classifier by setting the cut-off value of MHD & GE; 300 cGy. The deep neural network (DNN) achieved the highest F2 score (78.9%). Most models successfully classified all patients with high MHD as true positive. This study indicates that the ten models, especially the DNN, might have the potential to identify patients who may not require DIBH. en-copyright= kn-copyright= en-aut-name=Al-HammadWlla E. en-aut-sei=Al-Hammad en-aut-mei=Wlla E. kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=1 ORCID= en-aut-name=KurodaMasahiro en-aut-sei=Kuroda en-aut-mei=Masahiro kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=2 ORCID= en-aut-name=KamizakiRyo en-aut-sei=Kamizaki en-aut-mei=Ryo kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=3 ORCID= en-aut-name=TekikiNouha en-aut-sei=Tekiki en-aut-mei=Nouha kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=4 ORCID= en-aut-name=IshizakaHinata en-aut-sei=Ishizaka en-aut-mei=Hinata kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=5 ORCID= en-aut-name=KurodaKazuhiro en-aut-sei=Kuroda en-aut-mei=Kazuhiro kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=6 ORCID= en-aut-name=SugimotoKohei en-aut-sei=Sugimoto en-aut-mei=Kohei kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=7 ORCID= en-aut-name=OitaMasataka en-aut-sei=Oita en-aut-mei=Masataka kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=8 ORCID= en-aut-name=TanabeYoshinori en-aut-sei=Tanabe en-aut-mei=Yoshinori kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=9 ORCID= en-aut-name=BarhamMajd en-aut-sei=Barham en-aut-mei=Majd kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=10 ORCID= en-aut-name=SugiantoIrfan en-aut-sei=Sugianto en-aut-mei=Irfan kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=11 ORCID= en-aut-name=ShimizuYudai en-aut-sei=Shimizu en-aut-mei=Yudai kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=12 ORCID= en-aut-name=NakamitsuYuki en-aut-sei=Nakamitsu en-aut-mei=Yuki kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=13 ORCID= en-aut-name=AsaumiJunichi en-aut-sei=Asaumi en-aut-mei=Junichi kn-aut-name= kn-aut-sei= kn-aut-mei= aut-affil-num=14 ORCID= affil-num=1 en-affil=Department of Oral and Maxillofacial Radiology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences kn-affil= affil-num=2 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=3 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=4 en-affil=Department of Oral and Maxillofacial Radiology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences kn-affil= affil-num=5 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=6 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=7 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=8 en-affil=Graduate School of Interdisciplinary Sciences and Engineering in Health Systems, Okayama University kn-affil= affil-num=9 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=10 en-affil=Department of Dentistry and Dental Surgery, College of Medicine and Health Sciences, An-Najah National University kn-affil= affil-num=11 en-affil=Department of Oral Radiology, Faculty of Dentistry, Hasanuddin University kn-affil= affil-num=12 en-affil=Department of Oral and Maxillofacial Radiology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences kn-affil= affil-num=13 en-affil=Radiological Technology, Graduate School of Health Sciences, Okayama University kn-affil= affil-num=14 en-affil=Department of Oral and Maxillofacial Radiology, Okayama University Graduate School of Medicine, Dentistry and Pharmaceutical Sciences kn-affil= en-keyword=breast cancer kn-keyword=breast cancer en-keyword=radiotherapy kn-keyword=radiotherapy en-keyword=heart dose kn-keyword=heart dose en-keyword=machine learning kn-keyword=machine learning en-keyword=deep neural network kn-keyword=deep neural network en-keyword=deep inspiration breath-hold technique kn-keyword=deep inspiration breath-hold technique en-keyword=computed tomography kn-keyword=computed tomography END