In this paper, we consider densely connected convolutional networks and their applicability to the problem of assessment of knee osteoarthritis (OA) severity in the five-point Kellgren-Lawrence scale. First, we use trained from scratch Single Shot Detector (SSD) to localize knee joint areas in radiographs. Then, we apply DenseNets to quantify OA stages in the images of detected knee joints. We consider networks of different depths, trained both from scratch and pre-trained on the ImageNet dataset and fine-tuned in the images from Osteoarthritis Initiative dataset (OAI). Also, different loss functions are examined to understand which one gives the best training results. In the knee joint localization task, we obtain an accuracy of 94.03% under the Jaccard index threshold of 0.75. Also, our classifier outperforms the current state-of-the-art with accuracy of 71% in the classification task.
Detection of objects of interest is a crucial step in the automatic analysis of the medical X-ray images. However, medical X-rays are often characterized by the low contrast as well as great variability in range of colours, which makes it more difficult to be analysed by the common methods based on the regions homogeneity principles. In our paper, we present an alternative approach to the contours detection problem that does not require the homogeneity criteria to be satisfied. Our method is based on the identification of edge fragments and elimination of discontinuities between them. Moreover, we describe a numeric criterion for quality evaluation of contours detection. The obtained results can used for diagnosis of abnormalities and diseases, and also as an intermediate step for more sophisticated methods of image analysis.
This paper presents a novel approach for measuring the transverse velocity of large-scale objects based on stereo vision. The suggested approach uses a high-speed stereo camera located perpendicular to the traffic lane, and consists in matching frames from the left and right cameras. Compared to methods based on monocular vision, this approach solves the problem of dependence of object speed on distance. The proposed algorithm is part of a system for non-contact measurement of large-sized objects and has a calculated measurement error that does not exceed 1.5% of the measured speed up to 30 km/h while having low computational complexity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.