2018 · 39 citations · 6 references
ComputingComputer ScienceRadiologyComputer VisionDiagnostic ImagingMachine LearningComputer-aided DiagnosisDeep LearningImage ClassificationMedical Image AnalysisMedical ImagingPulmonary TuberculosisComparative TestingConvolutional Neural Network DetectionDeep Image FeaturesDeep Learning In RadiologyDeep Learning-based ClassificationDeep Learning-based DetectionDeep Learning-based DiagnosisPulmonary Tuberculosis Diagnosis
Because the tuberculosis is increasingly threatening the health and life of patients, this paper does the research on classification of pulmonary tuberculosis based on CT images. We proposed a new method called AECNN, which combines the feature extraction of Conv and the unsupervised features of AutoEncoder to be AE-CNN block, and constructs the deep network model of AECNN based on deep learning, processes and analyzes the whole ROI image to realize abnormal classification of tuberculosis. After comparative testing, our method is significantly superior to other machine learning algorithms and improved the accuracy of tuberculosis classification.
6
Gradient-based learning applied to document recognition
Yann LeCun, Léon Bottou, Yoshua Bengio et al. · Proceedings of the IEEE · 1998 · 58.2K citations · Full text
Corinna Cortes, Vladimir Vapnik · Machine Learning · 1995 · 40.5K citations · Full text