Inverted papilloma and nasal polyp classification using a deep convolutional network integrated with an attention mechanism.

BACKGROUND: Inverted papilloma (IP) is a common sinus neoplasm with a probability of malignant transformation. Nasal polyps (NP) are the most frequent masses in the sinus. The classification of IP and NP using computed tomography (CT) is highly significant for preoperative recognition, treatment, and clinical examination. Few visible differences exist between IP and NP in CT, making it a challenge for otolaryngologists to distinguish between them. This study intended to classify IP and NP using a neural network and analyze its ability to discriminate the differences. METHODS: IP and NP in CT were classified using a deep convolutional neural network (CNN) with an attention mechanism, which combines a densely connected convolutional network (DenseNet) and squeeze-and-excitation network (SENet). Using SENet's channel attention, the specific channel weights in the feature maps are improved, which can enhance feature discriminativeness. To discuss the interpretability of SE-DenseNet, we analyzed the heatmap of the final convolutional layer. RESULTS: We evaluated the classification performance of SE-DenseNet on a clinical dataset containing 3382 slices for 136 patients. The experimental results and a heatmap show that SE-DenseNet can effectively locate sinonasal lesions in patients and distinguish IP from NP with an average Acc of 88.4% and AUC of 0.87. CONCLUSION: Otolaryngologists can use the proposed model to diagnose IP and NP in CT because of its accuracy and efficiency. Moreover, the visualized heatmaps produced by the convolutional layers show that the method is reliable.

On Detection, Data Association and Segmentation for Multi-Target Tracking.

In this work, we propose a tracker that differs from most existing multi-target trackers in two major ways. First, our tracker does not rely on a pre-trained object detector to get the initial object hypotheses. Second, our tracker's final output is the fine contours of the targets rather than traditional bounding boxes. Therefore, our tracker simultaneously solves three main problems: detection, data association and segmentation. This is especially important because the output of each of those three problems are highly correlated and the solution of one can greatly help improve the others. The proposed algorithm consists of two main components: structured learning and Lagrange dual decomposition. Our structured learning based tracker learns a model for each target and infers the best locations of all targets simultaneously in a video clip. The inference of our structured learning is achieved through a new Target Identity-aware Network Flow (TINF), where each node in the network encodes the probability of each target identity belonging to that node. The probabilities are obtained by training target specific models using a global structured learning technique. This is followed by proposed Lagrangian relaxation optimization to find the high quality solution to the network. This forms the first component of our tracker. The second component is Lagrange dual decomposition, which combines the structured learning tracker with a segmentation algorithm. For segmentation, multi-label Conditional Random Field (CRF) is applied to a superpixel based spatio-temporal graph in a segment of video, in order to assign background or target labels to every superpixel. We show how the multi-label CRF is combined with the structured learning tracker through our dual decomposition formulation. This leads to more accurate segmentation results and also helps better resolve typical difficulties in multiple target tracking, such as occlusion handling, ID-switch and track drifting. The experiments on diverse and challenging sequences show that our method achieves superior results compared to competitive approaches for detection, multiple target tracking as well as segmentation.