We investigate the effectiveness of a simple solution to the common problem of deep learning in medical image analysis with limited quantities of labeled training data. The underlying idea is to assign artificial labels to abundantly available unlabeled medical images and, through a process known as surrogate supervision, pre-train a deep neural network model for the target medical image analysis task lacking sufficient labeled training data. In particular, we employ 3 surrogate supervision schemes, namely rotation, reconstruction, and colorization, in 4 different medical imaging applications representing classification and segmentation for both 2D and 3D medical images. 3 key findings emerge from our research: 1) pre-training with surrogate supervision is effective for small training sets; 2) deep models trained from initial weights pre-trained through surrogate supervision outperform the same models when trained from scratch, suggesting that pretraining with surrogate supervision should be considered prior to training any deep 3D models; 3) pre-training models in the medical domain with surrogate supervision is more effective than transfer learning from an unrelated domain (e.g., natural images), indicating the practical value of abundant unlabeled medical image data.

No abstract

No abstract

Passive Optical Network (PON) is the promising technology in future broadband access networks. Protection is an important and imperative issue in PON system while protection of cascaded PON (with two-stage splitter or more) system has been ignored. This paper presents a comprehensive study on protection strategy for cascaded PON system. The reliability and protection issues of cascaded PON are studied using theoretical analysis and numerical simulations. A function called Reliability Cost Ratio (RCR) is presented to evaluate the performance of different protection strategies. Simulation shows the variation trend of RCR with different splitter schemes.