Deep convolutional neural networks (CNNs) based transfer learning is an effective tool to reduce the dependence on hand-crafted features for handling medical classification problems, which may mitigate the problem of the insufficient training caused by the limited sample size. In this study, we investigated the discrimination power of the features at different CNN levels for the task of classifying epithelial and stromal regions on digitized pathologic slides which are prepared from breast cancer tissue. We extracted the low level and high level features from four different deep CNN architectures namely, AlexNet, Places365-AlexNet, VGG, and GoogLeNet. These features are used as input to train and optimize different classifiers including support vector machine (SVM), random forest (RF), and k-nearest neighborhood (KNN). A number of 15000 regions of interest (ROIs) acquired from the public database are employed to conduct this study. The result was observed that the low-level features of AlexNet, Places365-AlexNet and VGG outperformed the high-level ones, but the situation is in the opposite direction when the GoogLeNet is applied. Moreover, the best accuracy was achieved as 89.7% by the relatively deep layer of max pool 4 of GoogLeNet. In summary, our extensive empirical evaluation may suggest that it is viable to extend the use of transfer learning to the development of high-performance detection and diagnosis systems for medical imaging tasks.
The objective of this study is to investigate the performance of global and local features to better estimate the characteristics of highly heterogeneous metastatic tumours, for accurately predicting the treatment effectiveness of the advanced stage ovarian cancer patients. In order to achieve this , a quantitative image analysis scheme was developed to estimate a total of 103 features from three different groups including shape and density, Wavelet, and Gray Level Difference Method (GLDM) features. Shape and density features are global features, which are directly applied on the entire target image; wavelet and GLDM features are local features, which are applied on the divided blocks of the target image. To assess the performance, the new scheme was applied on a retrospective dataset containing 120 recurrent and high grade ovary cancer patients. The results indicate that the three best performed features are skewness, root-mean-square (rms) and mean of local GLDM texture, indicating the importance of integrating local features. In addition, the averaged predicting performance are comparable among the three different categories. This investigation concluded that the local features contains at least as copious tumour heterogeneity information as the global features, which may be meaningful on improving the predicting performance of the quantitative image markers for the diagnosis and prognosis of ovary cancer patients.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.