Improving Automated Hemorrhage Detection in Sparse-view Computed Tomography via Deep Convolutional Neural Network based Artifact Reduction

Johannes ThalhammerManuel SchultheissTina DorostiTobias LasserFranz PfeifferDaniela PfeifferFlorian Schaff

Johannes ThalhammerManuel SchultheissTina Dorosti ...+3 Florian Schaff

Mar 2023

摘要

Intracranial hemorrhage poses a serious health problem requiring rapid andoften intensive medical treatment. For diagnosis, a Cranial Computed Tomography(CCT) scan is usually performed. However, the increased health risk caused byradiation is a concern. The most important strategy to reduce this potentialrisk is to keep the radiation dose as low as possible and consistent with thediagnostic task. Sparse-view CT can be an effective strategy to reduce dose byreducing the total number of views acquired, albeit at the expense of imagequality. In this work, we use a U-Net architecture to reduce artifacts fromsparse-view CCTs, predicting fully sampled reconstructions from sparse-viewones. We evaluate the hemorrhage detectability in the predicted CCTs with ahemorrhage classification convolutional neural network, trained on fullysampled CCTs to detect and classify different sub-types of hemorrhages. Ourresults suggest that the automated classification and detection accuracy ofhemorrhages in sparse-view CCTs can be improved substantially by the U-Net.This demonstrates the feasibility of rapid automated hemorrhage detection onlow-dose CT data to assist radiologists in routine clinical practice.

展开全部

图表提取

论文十问

参考文献

被引用

提问与回答

暂无人提供速读十问回答

被引用

笔记

问答

领域