CNN Feature Extractor Research Articles

MO-DE-207B-06: Computer-Aided Diagnosis of Breast Ultrasound Images Using Transfer Learning From Deep Convolutional Neural Networks

Purpose: To assess the performance of using transferred features from pre-trained deep convolutional networks (CNNs) in the task of classifying cancer in breast ultrasound images, and to compare this method of transfer learning with previous methods involving human-designed features. Methods: A breast ultrasound dataset consisting of 1125 cases and 2393 regions of interest (ROIs) was used. Each ROI was labeled as cystic, benign, or malignant. Features were extracted from each ROI using pre-trained CNNs and used to train support vector machine (SVM) classifiers in the tasks of distinguishing non-malignant (benign+cystic) vs malignant lesions and benign vs malignant lesions. For a baseline comparison, classifiers were also trained on prior analytically-extracted tumor features. Five-fold cross-validation (by case) was conducted with the area under the receiver operating characteristic curve (AUC) as the performance metric. Results: Classifiers trained on CNN-extracted features were comparable to classifiers trained on human-designed features. In the non-malignant vs malignant task, both the SVM trained on CNN-extracted features and the SVM trained on human-designed features obtained an AUC of 0.90. In the task of determining benign vs malignant, the SVM trained on CNN-extracted features obtained an AUC of 0.88, compared to the AUC of 0.85 obtained by the SVM trained on human-designed features. Conclusion: We obtained strong results using transfer learning to characterize ultrasound breast cancer images. This method allows us to directly classify a small dataset of lesions in a computationally inexpensive fashion without any manual input. Modern deep learning methods in computer vision are contingent on large datasets and vast computational resources, which are often inaccessible for clinical applications. Consequently, we believe transfer learning methods will be important for computer-aided diagnosis schemes in order to utilize advancements in deep learning and computer vision without the associated costs. This work was partially funded by NIH grant U01 CA195564 and the University of Chicago Metcalf program. M.L.G. is a stockholder in R2/Hologic, co-founder and equity holder in Quantitative Insights, and receives royalties from Hologic, GE Medical Systems, MEDIAN Technologies, Riverain Medical, Mitsubishi, and Toshiba. K.D. received royalties from Hologic.

People counting based on head detection combining Adaboost and CNN in crowded surveillance environment

People counting is one of the key techniques in video surveillance. This task usually encounters many challenges in crowded environment, such as heavy occlusion, low resolution, imaging viewpoint variability, etc. Motivated by the success of R-CNN (Girshick et al., 2014) [1] on object detection, in this paper we propose a head detection based people counting method combining the Adaboost algorithm and the CNN. Unlike the R-CNN which uses the general object proposals as the inputs of CNN, our method uses the cascade Adaboost algorithm to obtain the head region proposals for CNN, which can greatly reduce the following classification time. Resorting to the strong ability of feature learning of the CNN, it is used as a feature extractor in this paper, instead of as a classifier as its commonly-used strategy. The final classification is done by a linear SVM classifier trained on the features extracted using the CNN feature extractor. Finally, the prior knowledge can be applied to post-process the detection results to increase the precision of head detection and the people count is obtained by counting the head detection results. A real classroom surveillance dataset is used to evaluate the proposed method and experimental results show that this method has good performance and outperforms the baseline methods, including deformable part model and cascade Adaboost methods.