Pediatric bone age assessment (BAA) is an extremely important clinical method to investigate endocrinology, genetic and growth disorders of adolescents. The current deep learning-based BAA scheme generally feeds the images into the training model, but ignores the local details of the skeleton images and does not exclude the noise around the images. Many methods train segmentation or detection networks to exploit local information, but requires a lot of manual annotation and additional costs. In this paper, we proposed an attentional region localization method for BAA to automatically localize the hand region and local regions without any additional annotations. First, an attentional hand location module (AHLM) was used to obtain a clearer hand region, which eliminates the interference noise of the original image. Then an attentional region generation module (ARGM) was used to extract the local attentional regions with high discriminant features, which can help to optimize the entire network framework. We integrate the entire network into an end-to-end structure by jointly optimizing the network through a shared backbone and fully connected layers. The effectiveness of the proposed attentional region localization method was evaluated on an open dataset Radiological Society of North America (RSNA) with an average absolute error (MAE) of 6.14, which performs better than most existing methods.
The biclustering algorithm based on greedy randomized adaptive search procedure (GRASP) has two main steps, firstly, bicluster seeds are generated by Kmeans clustering; secondly, the greedy randomized adaptive search procedure is used to expand these bicluster seeds so that to get some better biclusters. However, K-means clustering assumes that each gene belongs to only one cluster, which is not biologically valid, because some genes may not belong to only one category, and K-means clustering requires multiple runs to determine the number of "clusters," all of which will affect the identification of bicluster seeds. This paper optimizes GRASP-based bIclustering algorithm. In this paper, the single column vector clustering is used to generate biclusters seed, and then the bicluster seeds are added more rows and columns to generate the final biclusters. Experiment shows that this algorithm can generate a large number of biclusters with coexpression level, and the algorithm also finds more biclusters.
S i m i l a r i t y r e t r i e v a l from a pict o r i a l d a t a base can be made more e f f i c i e n t by encoding t h e o r i g i n a l d a t a base i n t o cert a i n convenient format. Generalized hypercube (GH) encoding is one such technique. To optimize GH encoding, a h e u r i s t i c approach h a s been formulated. Two o p t i m i z a t i o n problems have been considered here: (1) given t h e handle l e n g t h m, f i n d optimal GHm encoding;( 2 ) given t h e threshold d e n s i t y , f i n d optimal GH encoding such t h a t each hypercube has dens i t y no less than t h r e s h o l d d e n s i t y .
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.