No abstract

Objectives/Hypothesis: To develop a deep-learning-based computer-aided diagnosis system for distinguishing laryngeal neoplasms (benign, precancerous lesions, and cancer) and improve the clinician-based accuracy of diagnostic assessments of laryngoscopy findings. Study Design: Retrospective study. Methods: A total of 24,667 laryngoscopy images (normal, vocal nodule, polyps, leukoplakia and malignancy) were collected to develop and test a convolutional neural network (CNN)-based classifier. A comparison between the proposed CNNbased classifier and the clinical visual assessments (CVAs) by 12 otolaryngologists was conducted. Results: In the independent testing dataset, an overall accuracy of 96.24% was achieved; for leukoplakia, benign, malignancy, normal, and vocal nodule, the sensitivity and specificity were 92.8% vs. 98.9%, 97% vs. 99.7%, 89% vs. 99.3%, 99.0% vs. 99.4%, and 97.2% vs. 99.1%, respectively. Furthermore, when compared with CVAs on the randomly selected test dataset, the CNN-based classifier outperformed physicians for most laryngeal conditions, with striking improvements in the ability to distinguish nodules (98% vs. 45%, P < .001), polyps (91% vs. 86%, P < .001), leukoplakia (91% vs. 65%, P < .001), and malignancy (90% vs. 54%, P < .001). Conclusions: The CNN-based classifier can provide a valuable reference for the diagnosis of laryngeal neoplasms during laryngoscopy, especially for distinguishing benign, precancerous, and cancer lesions.

Path loss prediction is of great significance for the performance optimization of wireless networks. With the development and deployment of the fifth-generation (5G) mobile communication systems, new path loss prediction methods with high accuracy and low complexity should be proposed. In this paper, the principle and procedure of machine-learning-based path loss prediction are presented. Measured data are used to evaluate the performance of different models such as artificial neural network, support vector regression, and random forest. It is shown that these machine-learning-based models outperform the log-distance model. In view of the fact that the volume of measured data sometimes cannot meet the requirements of machine learning algorithms, we propose two mechanisms to expand the training dataset. On one hand, old measured data can be reused in new scenarios or at different frequencies. On the other hand, the classical model can also be utilized to generate a number of training samples based on the prior information obtained from measured results. Measured data are employed to verify the feasibility of these data expansion mechanisms. Finally, some issues for future research are discussed.

No abstract

Dimensionality reduction has been a long-standing research topic in academia and industry for two major reasons. First, in almost every domain, ranging from biology, social science, economics, to military data processing applications, the increasingly large volume of data is challenging the existing computing capability and raising the computing cost. Second, the notion of "intrinsic structure" allows us to remove some redundant dimensions from high-dimensional observations and reduce it into low-dimensional features without significant information loss. Autoencoder, as a powerful tool for dimensionality reduction has been intensively applied in image reconstruction, missing data recovery and classification. In this paper, we propose a new structure, folded autoencoder based on symmetric structure of conventional autoencoder, for dimensionality reduction. The new structure reduces the number of weights to be tuned and thus reduces the computational cost. Simulation results over MNIST data benchmark validate the effectiveness of this structure.