GLCM is an image texture-analyzing algorithm based on the occurrence probability of gray level pairs. GLCM can be used to describe the orientation, amplitude, and period information of an image texture.^{19}^{,}^{20} First, the GLCMs $P(i,j,d,\psi )$ are calculated by the numbers of repetitive pixel pairs ($i,j$) in the p-SI $Ikl(y,z)$, which have an assigned distance $d$ at an assigned angle $\psi $ with intensities $i$ and $j$ where $i,j=1,2,3,\u2026,G$, $G$ is the maximum intensity, $d$ can be any integer smaller than the image size, and $\psi $ can be 0 deg, 45 deg, 90 deg, or 135 deg. Then, the normalized GLCMs $p(i,j,d,\psi )$ are obtained by normalizing $P(i,j,d,\psi )$ with the number of total pixel pairs. To study adjacent pixels, pixel pair distance $d$ was set to 1, and four normalized GLCMs $p(i,j,d,\psi )$ could be termed as $p(i,j,\psi )$ with $\psi =0\u2009\u2009deg$, 45 deg, 90 deg, and 135 deg. A total of 17 texture parameters (correlation COR, dissimilarity DIS, contrast CON, inverse difference moment IDM, entropy ENT, sum entropy SEN, difference entropy DEN, angular second moment ASM, variance VAR, sum variance SVA, difference variance DVA, mean MEA, sum average SAV, cluster shade CLS, cluster prominence CLP,^{23} maximum probability MAP, and minimum probability MIP) were extracted from each GLCM to characterize p-SI $Ikl(y,z)$, defined as follows: Display Formula
$CORkl,\psi =1\sigma x\sigma y\u2211i=0G\u22121\u2211j=0G\u22121(ij)p(i,j,\psi )\u2212\mu x(\psi )\mu y(\psi )DISkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121|i\u2212j|p(i,j,\psi )CONkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121(i\u2212j)2p(i,j,\psi )IDMkl,\psi =\u2211i=0G\u22121\u2211j=0G\u2212111+(i\u2212j)2p(i,j,\psi )ENTkl,\psi =\u2212\u2211i=0G\u22121\u2211j=0G\u22121p(i,j,\psi )\xb7log[p(i,j,\psi )]SENkl,\psi =\u2212\u2211h=02\u2009\u2009G\u22122px+y(h,\psi )\xb7log[px+y(h,\psi )]DENkl,\psi =\u2212\u2211h=0G\u22121px\u2212y(h,\psi )\xb7log[px\u2212y(h,\psi )]ASMkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121[p(i,j,\psi )]2VARkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121[i\u2212\mu (\psi )]2p(i,j,\psi )SVAkl,\psi =\u2211h=02\u2009\u2009G\u22122(h\u2212SENkl,\psi )2px+y(h,\psi )DVAkl,\psi =1G\u22121\u2211h=0G\u22121[px\u2212y(h,\psi )\u2212p\xafx\u2212y(\psi )]2MEAkl,\psi =\u2211i=0G\u22121i\u2211j=0G\u22121p(i,j,\psi )=\mu x(\psi )SAVkl,\psi =\u2211h=02\u2009\u2009G\u22122kpx+y(h,\psi )CLSkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121[i+j\u22122\mu (\psi )]3p(i,j,\psi )CLPkl,\psi =\u2211i=0G\u22121\u2211j=0G\u22121[i+j\u22122\mu (\psi )]4p(i,j,\psi )MIPkl,\psi =min[p(i,j,\psi )]MAPkl,\psi =max[p(i,j,\psi )],$(5)
where $k$ indicates the polarization of scattered light from cells, $l$ indicates the polarization of incident light, $p(i,j,\psi )$ are the normalized GLCMs of p-SI $Ikl(y,z)$, $G$ is the maximum gray level of p-SI $Ikl(y,z)$, $H$ and $W$ are the height and width of p-SI $Ikl(y,z)$, respectively, and Display Formula$px(i,\psi )=\u2211j=0G\u22121p(i,j,\psi )py(j,\psi )=\u2211i=0G\u22121p(i,j,\psi )px+y(h,\psi )=\u2211i=0G\u22121\u2211j=0G\u22121i+j=hp(i,j,\psi )px\u2212y(h,\psi )=\u2211i=0G\u22121\u2211j=0G\u22121|i\u2212j|=hp(i,j,\psi )\mu x(\psi )=\u2211i=0G\u22121ipx(i,\psi )\mu y(\psi )=\u2211j=0G\u22121jpy(j,\psi )\sigma x(\psi )2=\u2211i=0G\u22121[px(i,\psi )\u2212\mu x(\psi )]2\sigma y(\psi )2=\u2211j=0G\u22121[py(j,\psi )\u2212\mu y(\psi )]2\mu (\psi )=\mu x(\psi )=\mu y(\psi ).$(6)