计算机工程与应用2024，Vol.60Issue(19)：18-31,14.DOI:10.3778/j.issn.1002-8331.2402-0070
深度学习检测肺结节难点问题综述
Review of Difficult Problems of Deep Learning to Detect Lung Nodules
摘要
Abstract
Lung cancer is one of the deadliest cancers worldwide,and pulmonary nodules are an early manifestation of lung cancer.Detection models based on deep learning have gradually become an effective method to assist doctors in de-tecting pulmonary nodules due to their high detection accuracy and efficiency.However,current deep learning-based pul-monary nodule detection models still have shortcomings,and there are several challenging issues that need to be resolved.First,it addresses the issue of insufficient pulmonary nodule data and class imbalance during model training through trans-fer learning,GAN networks,semi-supervised learning,and unsupervised learning.Second,it enhances the model's fea-ture extraction capabilities to improve the sensitivity and accuracy of pulmonary nodule detection.Third,it improves the model's capability to screen out false-positive pulmonary nodules,thereby reducing the false-positive rate.Fourth,it strengthens the model's interpretability in detecting pulmonary nodules.Fifth,it solves the above four challenging issues using large model technology.Finally,it introduces the datasets and evaluation metrics required for the training and test-ing of detection models and discusses future optimization directions for pulmonary nodule detection.关键词
肺癌/肺结节检测/深度学习/计算机辅助检测系统Key words
lung cancer/lung nodule detection/deep learning/computer-aided detection system分类
信息技术与安全科学引用本文复制引用
包强强,唐思源,谷宇..深度学习检测肺结节难点问题综述[J].计算机工程与应用,2024,60(19):18-31,14.基金项目
内蒙古自然科学基金(2021MS06026,2024LHMS06006) （2021MS06026,2024LHMS06006）
内蒙古自治区卫生健康委员会项目(202201395) （202201395）
包头市卫生健康科技计划项目(wsjkkj2022120) （wsjkkj2022120）
国家自然科学基金(62001255) （62001255）
中央引导地方科技发展资金项目(2021ZY0004) （2021ZY0004）
内蒙古自治区高等学校青年科技英才计划项目(NJYT23057) （NJYT23057）
内蒙古科技大学基本科研业务费专项资金(019) （019）
内蒙古自治区大学生创新创业训练计划项目(s202310130004,s202410130004). （s202310130004,s202410130004）
