Natalia Turosz, Kamila Chęcińska, Maciej Chęciński, Karolina Lubecka, Filip Bliźniak, Maciej Sikora
{"title":"人工智能(AI)评估儿童牙科全景X光片(DPR):临床研究。","authors":"Natalia Turosz, Kamila Chęcińska, Maciej Chęciński, Karolina Lubecka, Filip Bliźniak, Maciej Sikora","doi":"10.3390/pediatric16030067","DOIUrl":null,"url":null,"abstract":"<p><p>This clinical study aimed to evaluate the sensitivity, specificity, accuracy, and precision of artificial intelligence (AI) in assessing permanent teeth in pediatric patients. Over one thousand consecutive DPRs taken in Kielce, Poland, with the Carestream CS9600 device were screened. In the study material, 35 dental panoramic radiographs (DPRs) of patients of developmental age were identified and included. They were automatically evaluated with an AI algorithm. The DPRs were then analyzed by researchers. The status of the following dichotomous variables was assessed: (1) decay, (2) missing tooth, (3) filled tooth, (4) root canal filling, and (5) endodontic lesion. The results showed high specificity and accuracy (all above 85%) in detecting caries, dental fillings, and missing teeth but low precision. This study provided a detailed assessment of AI performance in a previously neglected age group. In conclusion, the overall accuracy of AI algorithms for evaluating permanent dentition in dental panoramic radiographs is lower for pediatric patients than adults or the entire population. Hence, identifying primary teeth should be implemented in AI-driven software, at least so as to ignore them when assessing mixed dentition (ClinicalTrials.gov registration number: NCT06258798).</p>","PeriodicalId":45251,"journal":{"name":"Pediatric Reports","volume":"16 3","pages":"794-805"},"PeriodicalIF":1.4000,"publicationDate":"2024-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11417896/pdf/","citationCount":"0","resultStr":"{\"title\":\"Artificial Intelligence (AI) Assessment of Pediatric Dental Panoramic Radiographs (DPRs): A Clinical Study.\",\"authors\":\"Natalia Turosz, Kamila Chęcińska, Maciej Chęciński, Karolina Lubecka, Filip Bliźniak, Maciej Sikora\",\"doi\":\"10.3390/pediatric16030067\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>This clinical study aimed to evaluate the sensitivity, specificity, accuracy, and precision of artificial intelligence (AI) in assessing permanent teeth in pediatric patients. Over one thousand consecutive DPRs taken in Kielce, Poland, with the Carestream CS9600 device were screened. In the study material, 35 dental panoramic radiographs (DPRs) of patients of developmental age were identified and included. They were automatically evaluated with an AI algorithm. The DPRs were then analyzed by researchers. The status of the following dichotomous variables was assessed: (1) decay, (2) missing tooth, (3) filled tooth, (4) root canal filling, and (5) endodontic lesion. The results showed high specificity and accuracy (all above 85%) in detecting caries, dental fillings, and missing teeth but low precision. This study provided a detailed assessment of AI performance in a previously neglected age group. In conclusion, the overall accuracy of AI algorithms for evaluating permanent dentition in dental panoramic radiographs is lower for pediatric patients than adults or the entire population. Hence, identifying primary teeth should be implemented in AI-driven software, at least so as to ignore them when assessing mixed dentition (ClinicalTrials.gov registration number: NCT06258798).</p>\",\"PeriodicalId\":45251,\"journal\":{\"name\":\"Pediatric Reports\",\"volume\":\"16 3\",\"pages\":\"794-805\"},\"PeriodicalIF\":1.4000,\"publicationDate\":\"2024-09-11\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC11417896/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Pediatric Reports\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.3390/pediatric16030067\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"PEDIATRICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Pediatric Reports","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3390/pediatric16030067","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"PEDIATRICS","Score":null,"Total":0}
Artificial Intelligence (AI) Assessment of Pediatric Dental Panoramic Radiographs (DPRs): A Clinical Study.
This clinical study aimed to evaluate the sensitivity, specificity, accuracy, and precision of artificial intelligence (AI) in assessing permanent teeth in pediatric patients. Over one thousand consecutive DPRs taken in Kielce, Poland, with the Carestream CS9600 device were screened. In the study material, 35 dental panoramic radiographs (DPRs) of patients of developmental age were identified and included. They were automatically evaluated with an AI algorithm. The DPRs were then analyzed by researchers. The status of the following dichotomous variables was assessed: (1) decay, (2) missing tooth, (3) filled tooth, (4) root canal filling, and (5) endodontic lesion. The results showed high specificity and accuracy (all above 85%) in detecting caries, dental fillings, and missing teeth but low precision. This study provided a detailed assessment of AI performance in a previously neglected age group. In conclusion, the overall accuracy of AI algorithms for evaluating permanent dentition in dental panoramic radiographs is lower for pediatric patients than adults or the entire population. Hence, identifying primary teeth should be implemented in AI-driven software, at least so as to ignore them when assessing mixed dentition (ClinicalTrials.gov registration number: NCT06258798).