Transverse Scan

Visual Acuity
t? tmmttt

Figure 1-6, Image measurement using Optical Coherence Tomography. Tomographic images are constructed by performing rapid, successive axial (longitudinal) measurements at different transverse points. Each axial measurement represents optical reflection and backscatter from microstructures which are intercepted by the optical beam. By scanning the beam transversely while performing axial measurements, a two dimensional set of data may be measured which is a cross-sectional map of the reflection and back-scatter within the tissue.

Aqueous

Figure 1*7. Grey scale Optical Coherence Tomography image of the anterior chamber of a human eye obtained in vivo. The image contains 200 pixels (horizontal) by 500 pixels (vertical) and spans 21 mm across by 6 mm deep. The image is displayed using a logarithmic mapping of the measured optical signal to brightness. The maximum signal is approximately -50 dB of the incident signal, while the minimum detectable signal is approximately -95 dB.

Figure 1-8. False color representation of the same Optical Coherence Tomography anterior eye image shown in Figure 1-7, The image is displayed using a logarithmic mapping of the measured optical signal to brightness- The maximum optical reflection and backscattering (-50 dB of the incident light) are represented by red-white colors, while the minimum detectable signals (-95 dB) are represented by blue-black colors* As in other medical imaging modalities, the use of false color does not necessarily correspond to microstructural features, but can aid in differentiating different structures in an image.

Figure 1-8. False color representation of the same Optical Coherence Tomography anterior eye image shown in Figure 1-7, The image is displayed using a logarithmic mapping of the measured optical signal to brightness- The maximum optical reflection and backscattering (-50 dB of the incident light) are represented by red-white colors, while the minimum detectable signals (-95 dB) are represented by blue-black colors* As in other medical imaging modalities, the use of false color does not necessarily correspond to microstructural features, but can aid in differentiating different structures in an image.

different optical properties and not necessarily different tissue morphology Care must be taken to avoid interpreting images analogously to conventional his topa thology. OCT imaging is similar to ultrasound where different degrees of ultrasound backscattering correspond to different color or grey levels except that in OCT images, different optical reflection and scattering properties are measured-

Image Resolution

The image resolution of OCT in the axial (or longitudinal) versus transverse directions is determined by different mechanisms, A more detailed description of the factors governing image resolution is presented in the Appendix, The quantitative information presented here is intended only as a rough guide to image resolution since different OCT systems can have significantly different performance depending upon their design and intended application.

The resolution of the image in the axial (longitudinal) direction is determined by the resolution of the optical ranging measurement. This is determined by the physical properties of the light source which is used for the measurement. If a short pulse laser source is used, the axial resolution is determined by the pulse duration. Conversely, if a continuous, low-coherence light source is used, the axial resolution is determined by the '"coherence length" of the light source. For typical systems, the axial resolution is approximately 10 to 20 microns. This number represents the apparent width on a linear scale of an isolated reflection and is a measure of the smallest feature that can be differentiated in the image. It is important to note that the measurement of distance or tissue thickness can, in practice, be performed with significantly higher resolution than this limit.

The transverse resolution of the image is determined by the size of the focussed optical beam. This is a function of the optics used to project the beam onto the eye and is determined by factors such as whether imaging is performed over a large depth, such as in the anterior eye, or whether the focussing angle is restricted, as in imaging the retina. Typically, the transverse resolution is in the range of 20 to 50 microns and depends upon whether imaging is performed in the retina or the anterior eye, respectively.

The image resolution is also a function of the size of the tomogram that is desired. If a larger area is to be scanned, the image maybe constructed with a larger number of axial (longitudinal) data sets which correspond to more transverse points (or pixel elements) in the image. For most intraocular structures it is not necessary to image with the full resolution possible in the transverse dimension to permit clinical diagnosis. Since the image acquisition time depends on the number of transverse pixel elements as well as the depth of the image being acquired, faster image acquisition may be obtained if very high resolution imaging is not required.

Computer Image Processing and the Correction of Eye Motion

Since the resolution of OCT is extremely high, it is essential to compensate for motion of the eye during image acquisition, since eye motion can cause image blurring. Movements of the eye can be caused by a variety of processes including fluctuations in intraocular pressure produced by pulse, microsaccades and tremor, and changes in the patient's fixation point, Since OCT measures the absolute distance or range of the tissue specimen, it is essential to correct for motion of the eye. This problem is addressed by powerful yet simple computer image processing techniques which can be used to dramatically enhance imaging performance by virtually eliminating image blurring from involuntary patient eye motion.

Figure 1 -9 shows an optical coherence tomogram of the fovea showing the raw image data without image processing and the image achieved after processing to correct for eye motion. The dominant motion which blurs the image occurs in the axial (longitudinal) direction because retinal structure in this direction has dimensions on the micron scale. These changes in the longitudinal position of the eye may be corrected because the optical ranging measurement itself determines the longitudinal position of the retina. The tomographic image is constructed by performing sequential axial (longitudinal) ranging measurements at different transverse positions within the eye. However, if the eye moves on a micron scale in the longitudinal direction, between the time that successive axial range data sets are acquired, then the positions of different features will be observed at different axial (longitudinal) ranges. Thus, the motion of the patient's eye in the axial (longitudinal) direction may be measured, by correlating adjacent axial data sets. Figure 1-10 shows an example of this measurement.

Once the axial position of the patient's eye is known, the scans in the optical tomographic image may be displaced in the axial direction so that microstructural features will be aligned. Changes in transverse eye position are not eliminated using image processing, Thus, drifts in patient fixation are not compensated for. However, small changes in transverse position do not produce significant degradation in image quality because the typical features in the trans-

Was this article helpful?

0 0

Post a comment