English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Finding automatically multiple lesions in large images is a common problem inmedical image analysis. Solving this problem can be challenging if, duringoptimization, the automated method cannot access information about the locationof the lesions nor is given single examples of the lesions. We propose a newweakly supervised detection method using neural networks, that computesattention maps revealing the locations of brain lesions. These attention mapsare computed using the last feature maps of a segmentation network optimizedonly with global image-level labels. The proposed method can generate attentionmaps at full input resolution without need for interpolation duringpreprocessing, which allows small lesions to appear in attention maps. Forcomparison, we modify state-of-the-art methods to compute attention maps forweakly supervised object detection, by using a global regression objectiveinstead of the more conventional classification objective. This regressionobjective optimizes the number of occurrences of the target object in an image,e.g. the number of brain lesions in a scan, or the number of digits in animage. We study the behavior of the proposed method in MNIST-based detectiondatasets, and evaluate it for the challenging detection of enlargedperivascular spaces - a type of brain lesion - in a dataset of 2202 3D scanswith point-wise annotations in the center of all lesions in four brain regions.In the brain dataset, the weakly supervised detection methods come close to thehuman intrarater agreement in each region. The proposed method reaches the bestarea under the curve in two out of four regions, and has the lowest number offalse positive detections in all regions, while its average sensitivity overall regions is similar to that of the other best methods. The proposed methodcan facilitate epidemiological and clinical studies of enlarged perivascularspaces.

Weakly Supervised Object Detection with 2D and 3D Regression Neural Networks