AIM: To gather existing radiographic classifications of mandibular canals branching, considering the criteria on which these were based.METHODS: The search for studies on mandibular canals based on imaging exams inclu...AIM: To gather existing radiographic classifications of mandibular canals branching, considering the criteria on which these were based.METHODS: The search for studies on mandibular canals based on imaging exams included literature reviews, epidemiological studies of prevalence, descriptive studies, or case reports. An electronic search in the MEDLINE(Ovid SP), Pub Med, EMBASE(Ovid SP), Web of Science(Thompson Reuters), and Scopus(Elsevier) databases was performed, as well as a manual evaluation of the references of the selected articles. Combinations of key words were placed in each database. No restrictions were imposed regarding the year of publication or language. References collected in duplicate were removed by the authors. A table was drawn up, containing the included studies and respective interest data.RESULTS: Six classifications of mandibular canals branching were selected for the present literature review. Four were based on two-dimensional radiographic exams, and two were performed based on threedimensional tomographic exams. Three-dimensional classifications were determined based on the analysis found in the least number of exams, comparatively to two-dimensional studies. The prevalence of mandibular canal branching varied from 0% to 38.75% in the works based on two-dimensional exams, while those found inthree-dimensional exams ranged from 15.6% to 65%. The studies were mostly referred to branches that began in the mandibular ramus. Just one classification considered the branches that began in the mandibular body region.CONCLUSION: Three-dimensional exams appear to be the best method to view mandibular canal branching. Further studies are warranted to determine its true prevalence and questions concerning to associations.展开更多
目的构建轻量型单步深度学习网络利用曲面断层片自动检测下颌智齿牙根与下颌管位置关系。方法将1570例2543颗同时拍摄曲面断层片和CBCT的成人下颌智齿病例,随机分成训练组(80%)、验证组(10%)和测试组(10%)。分别使用曲面断层片和CBCT评...目的构建轻量型单步深度学习网络利用曲面断层片自动检测下颌智齿牙根与下颌管位置关系。方法将1570例2543颗同时拍摄曲面断层片和CBCT的成人下颌智齿病例,随机分成训练组(80%)、验证组(10%)和测试组(10%)。分别使用曲面断层片和CBCT评估下颌智齿牙根与下颌管的关系,分为非接触和接触。构建基于YOLO(You only look once)模型改良的轻量型单步深度学习网络算法模型(IAC⁃MTMnet),以曲面断层片作为输入端,以配对CBCT诊断作为金标准,评估下颌智齿牙根与下颌管的关系。诊断性能使用正确率、灵敏度、特异度、阳性预测值以及受试者工作曲线及曲线下面积表示。结果经CBCT诊断,下颌智齿牙根与下颌管接触的发生率为31.38%。IAC⁃MTMnet的诊断正确率为0.885,灵敏度为0.747,特异度为0.956,阳性预测值为0.899,受试者工作曲线下面积为0.95,测试运行时间为0.059 s。结论IAC⁃MTMnet模型提升了曲面断层片诊断下颌智齿牙根与下颌管关系的性能。展开更多
文摘AIM: To gather existing radiographic classifications of mandibular canals branching, considering the criteria on which these were based.METHODS: The search for studies on mandibular canals based on imaging exams included literature reviews, epidemiological studies of prevalence, descriptive studies, or case reports. An electronic search in the MEDLINE(Ovid SP), Pub Med, EMBASE(Ovid SP), Web of Science(Thompson Reuters), and Scopus(Elsevier) databases was performed, as well as a manual evaluation of the references of the selected articles. Combinations of key words were placed in each database. No restrictions were imposed regarding the year of publication or language. References collected in duplicate were removed by the authors. A table was drawn up, containing the included studies and respective interest data.RESULTS: Six classifications of mandibular canals branching were selected for the present literature review. Four were based on two-dimensional radiographic exams, and two were performed based on threedimensional tomographic exams. Three-dimensional classifications were determined based on the analysis found in the least number of exams, comparatively to two-dimensional studies. The prevalence of mandibular canal branching varied from 0% to 38.75% in the works based on two-dimensional exams, while those found inthree-dimensional exams ranged from 15.6% to 65%. The studies were mostly referred to branches that began in the mandibular ramus. Just one classification considered the branches that began in the mandibular body region.CONCLUSION: Three-dimensional exams appear to be the best method to view mandibular canal branching. Further studies are warranted to determine its true prevalence and questions concerning to associations.
文摘目的构建轻量型单步深度学习网络利用曲面断层片自动检测下颌智齿牙根与下颌管位置关系。方法将1570例2543颗同时拍摄曲面断层片和CBCT的成人下颌智齿病例,随机分成训练组(80%)、验证组(10%)和测试组(10%)。分别使用曲面断层片和CBCT评估下颌智齿牙根与下颌管的关系,分为非接触和接触。构建基于YOLO(You only look once)模型改良的轻量型单步深度学习网络算法模型(IAC⁃MTMnet),以曲面断层片作为输入端,以配对CBCT诊断作为金标准,评估下颌智齿牙根与下颌管的关系。诊断性能使用正确率、灵敏度、特异度、阳性预测值以及受试者工作曲线及曲线下面积表示。结果经CBCT诊断,下颌智齿牙根与下颌管接触的发生率为31.38%。IAC⁃MTMnet的诊断正确率为0.885,灵敏度为0.747,特异度为0.956,阳性预测值为0.899,受试者工作曲线下面积为0.95,测试运行时间为0.059 s。结论IAC⁃MTMnet模型提升了曲面断层片诊断下颌智齿牙根与下颌管关系的性能。
