Reconstruction of optical coherent tomography (OCT) images of morphological features

摘要

Imaging devices and processing techniques based on imaging information along the depth direction in an optical coherent tomography (OCT) image are disclosed to enhance observed morphological features. The methods and systems obtain different OCT images of the target object under different reference path lengths, process the different OCT images to obtain a derivative with respect to the signal path length of image information of the different images, and processing the derivative to extract improved image information of the target object. The derivatives may be in a form related to a gradient value of the normalized OCT image intensity, or an attenuation coefficient.

主权项

1. A method for extracting image information from images of a target object in an optical coherence tomography (OCT) device obtained via scanning a reference path length relative to a signal path length of a signal path in the interferometer of the OCT device where the target object is located at one end of the signal path, comprising:
obtaining different OCT images of the target object under different reference path lengths; processing the different OCT images to obtain a derivative with respect to the signal path length of image information of the different images; and processing the derivative to extract improved image information of the target object to reduce an effect of a signal decay due to scattering of light in the target object, wherein obtaining different OCT images of the target object under different reference path lengths comprises:
obtaining different OCT images of the target object at different locations shifted in position represented by x and y coordinates along a direction substantially perpendicular to a direction of incidence of light into the target object represented by z coordinate; andobtaining different OCT images of the target object at different reference path lengths at each of the different locations to obtain 3-dimensional image information of the target object at the different locations and at different depths with different z coordinate below a surface of the target object, and wherein processing the different OCT images to obtain a derivative with respect to the signal path length of image information of the different images comprises:
averaging a first number of consecutive B-scan images among the different OCT images to obtain an averaged B-scan image;obtaining a normalized OCT image intensity of a position in the averaged B-scan image by dividing a signal intensity of the position in the averaged B-scan image by a signal intensity at a surface of the target object;calculating a gradient value of the normalized OCT image intensity in the signal path length direction by subtracting a normalized OCT image intensity in a next depth to the normalized OCT image intensity; andobtaining the derivative with respect to the signal path length by dividing the gradient value of the normalized OCT image intensity by double actual distance of one pixel.