Fig 1.
Haralick texture features are calculated from the gray-level co-occurrence matrix (GLCM).
Example of how the GLCM is calculated for a given 4x4 pixel image (a) with the corresponding numerical gray-level pixel intensities (b) is shown here. GLCM is computed by going in horizontal direction with one pixel separation and recording the number of occurrences in which a pixel intensity of 2 is horizontally next to a 1 and allocated in the co-occurrence matrix (c). Figure adapted from ref. [20].
Table 1.
The 13 Haralick texture features [14] that were calculated using the Python-based in-house developed software, pyOsirix, which was integrated into the DICOM viewer, Osirix.
Fig 2.
Demographics of the imaging cohort included 25 cases determined to be normal and 19 cases with mild central nervous system (CNS) ventriculomegaly based on MR findings.
The gestational age ranged from 23 to 36 weeks.
Table 2.
MRI parameters of the standard T2W, T1W, and DWI.
Fig 3.
Left to Right: Representative T2W, T1W, DWI (b = 800 s/mm2), and ADC map of a 24-week (top row) and a 34-week (bottom row) gestation with their respective placental ROIs (green).
For DW images SNR calculation, the image background noise ROI (yellow) is drawn outside of the tissue. The 24-week fetus was identified to have mild CNS ventriculomegaly; the 34-week fetus had normal MR findings.
Fig 4.
Placental volume of the imaging cohort showed the expected correlation between placental volume and gestational age (ρ = 0.62, p <0.0001).
No significant correlation was observed between mean ADC value and gestational age (ρ = -0.06, p = 0.69).
Table 3.
Spearman correlation coefficient between the Haralick texture features that were found to be significant from placental MR acquisitions and gestational age.
The corresponding q values (FDR adjusted p-values) are provided in brackets. Correlation coefficient values that were statistically significant are highlighted.