US 12,343,118 B2
Neurovascular age prediction system based on white matter hyperintensity signal and method thereof
Chu-Chung Huang, Hsinchu (TW); and Ching-Po Lin, Hsinchu (TW)
Assigned to MiTech.ai Co., Ltd., Taipei (TW)
Filed by NATIONAL YANG MING CHIAO TUNG UNIVERSITY, Hsinchu (TW)
Filed on Aug. 5, 2022, as Appl. No. 17/882,513.
Claims priority of application No. 110133826 (TW), filed on Sep. 10, 2021.
Prior Publication US 2023/0080821 A1, Mar. 16, 2023
Int. Cl. G06K 9/00 (2022.01); A61B 5/02 (2006.01); A61B 5/055 (2006.01); G06T 7/00 (2017.01); G06T 7/10 (2017.01)
CPC A61B 5/02007 (2013.01) [A61B 5/055 (2013.01); G06T 7/0012 (2013.01); G06T 7/10 (2017.01); G06T 2207/10088 (2013.01); G06T 2207/30016 (2013.01); G06T 2207/30096 (2013.01); G06T 2207/30101 (2013.01)] 4 Claims
OG exemplary drawing
 
1. A neurovascular age prediction method based on white matter, comprising:
performing a brain scan on a subject to generate magnetic resonance images, by a magnetic resonance device, wherein the magnetic resonance images comprise a T1 weighted image and a T2 fluid attenuated inversion recovery image;
receiving the magnetic resonance images from the magnetic resonance device, by an analysis device;
transforming the individual coordinates of the T1 weighted image into international standard coordinates through a nonlinear space counterpoint technology, to generate a T1 weighted coordinate transformation image, by the analysis device;
generating a ventricle mask based on a MNI152 international standard brain template, and extending the ventricle mask outwardly by a preset distance to cover the T1 weighted coordinate transformation image to generate a periventricular white matter atlas and a deep white matter atlas, by the analysis device;
reversely transforming the individual coordinate into a transition matrix of international standard coordinate, and applying the transition matrix of international standard coordinate to the periventricular white matter atlas and the deep white matter atlas to generate an individual space periventricular area mask and an individual space deep white matter mask, respectively, by the analysis device;
performing a white matter hyperintensity signal image processing on the T1 weighted image and the T2 fluid attenuated inversion recovery image to generate a T1 weighted white matter hyperintensity signal image and a T2 fluid attenuated inversion recovery white matter hyperintensity signal image, by the analysis device;
using the individual space periventricular area mask and the individual space deep white matter mask to perform calculation on the T1 weighted white matter hyperintensity signal image and the T2 fluid attenuated inversion recovery white matter hyperintensity signal image, to obtain a periventricular white matter hyperintensity volume and a deep white matter hyperintensity volume, respectively, by the analysis device;
performing logarithm conversion on the periventricular white matter hyperintensity volume and the deep white matter hyperintensity volume to generate a logarithm of the periventricular white matter hyperintensity volume and a logarithm of the deep white matter hyperintensity volume value, by the analysis device; and
substituting the logarithm of periventricular white matter hyperintensity volume and the logarithm of the deep white matter hyperintensity volume into a neurovascular age prediction model to obtain a neurovascular age prediction result, by the analysis device.