Semi-automated measurement from X-ray diffraction of connective tissues with an application to the uterine cervix during pregnancy
X-ray diffraction from soft tissues can give information about collagen contained within the tissue matrix. Semi-automated analysis software was developed for diffraction pattern analysis. Diffraction patterns are often very different for different tissues and can be relatively indistinct and have more than tow diffractions maxima positions. A marked centre and the first two maxima positions provide distance values. Three different materials containing collagen were used and the variation of distances used to obtain a measure of accuracy. The peaks could be found on the images used to an accuracy of about 4% relating to a precision measured on the originals film of about 4mm. Accuracy of the methods varies according to the diffraction pattern and can vary from 2% to 7% accuracy. A preliminary study was carried out to assess the uterine cervix at various stages during pregnancy using a rat model. Resulting diffraction patterns were scanned to resolution of 105mm per pixel, to create a digital image. Data extracted from the image around a circle passing through the first two peaks located has a Gaussian functional fit applied. There were statistically significant changes towards a more vertical (circumferential to the cervix canal) orientation between pairs of measurements both in separation and in bisector angle measurements. The separation between orientations corresponding to the two diffraction maxima appear to decrease during pregnancy with two pairs of values giving significant changes. Fitted Gaussian curves with the larger sigma value demonstrated a marked increase in width at about the time of parturition. Conclusions from the results on testing the software and using the methods on a real practical study indicate that the software methods developed were easy to use, reasonably accurate and obtain information fairly rapidly from individual images of diffuse peaks.