Optimizing Deep Learning Algorithms for Segmentation of Acute Infarcts on Noncontrast CT of the Brain Using Simulated Lesions

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. To develop a framework to generate synthetic noncontrast CT acute stroke lesions with exact labels and adjustable levels of attenuation to test the efficacy of lesion segmentation deep learning algorithms. The signal intensity on the NCCT images was depressed by 4 HU (a 13% drop) in the region of the diffusion-weighted lesion.