首页|Patent Issued for Machine learning generation of low-noise and high structural c onspicuity images (USPTO 12141900)
Patent Issued for Machine learning generation of low-noise and high structural c onspicuity images (USPTO 12141900)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
From the background information supplied by the inventors, news correspondents o btained the followingquote: “A computed tomography (CT) image is often captured or generated via CT imaging equipmentthat is set at a given energy and dose le vel and reconstructed with a desired frequency kernel setting.The choice of fre quency kernel is decided based on the clinical application and the structure of diagnosticinterest. While high frequency kernels have better structural conspic uity and sharpness, they suffer fromhigh noise and loss of contrast in soft tis sues. On the other hand, low frequency kernels are well-suitedfor soft tissue a nalysis and provide low noise and good contrast, but such low frequency kernels result inlow resolution. So, in clinical practice, the CT image is often subjec ted to postprocessing to address suchhigh noise and/or low resolution.
NETL
NSTL
