The purpose of this annex is to explain key IVOCT FOR PROCESSING parameters, describe the relationship between IVOCT FOR PROCESSING and FOR PRESENTATION images. It also explains Intravascular Longitudinal Reconstruction.
When an OCT image is acquired, the path length difference between the reference and sample arms may vary, resulting in a shift along the axial direction of the image, known as the Z Offset. With FOR PROCESSING images, in order to convert the image in Cartesian coordinates and make measurements, this Z Offset should be corrected, typically on a per-frame or per-image basis. Z Offset is corrected by shifting Polar data rows (A-lines) + OCT Z Offset Correction (0052,0030) pixels along the axial dimension of the image.
Z Offset correction may be either a positive or negative value. Positive values mean that the A-lines are shifted further away from the catheter optics. Negative values mean that the A-lines are shifted closer to the catheter optics. Figure EEE.2-1 illustrates a negative Z Offset Correction.
Figure EEE.2-1 - Z Offset Correction
The axial distances in an OCT image are dependent on the refractive index of the material that that IVOCT light passes through. As a result, in order to accurately make measurements in images derived from FOR PROCESSING data, the axial dimension of the pixels should be globally corrected by dividing the A-line Pixel Spacing (0052,0014) value (in air) by the Effective Refractive Index (0052,0004) and setting the Refractive Index Applied (0052,003A) to YES. Although not recommended, if A-line Pixel Spacing (0052,0014) is reported in air (i.e. not corrected by dividing by Effective Refractive Index) then the Refractive Index Applied value shall be set to NO.
FOR PROCESSING Polar data is specified such that each column represents a subsequent axial (z) location and each row an angular (θ) coordinate. Following Z Offset and Refractive Index Correction, Polar data can be converted to Cartesian data by first orienting the seam line position so that it is at the correct row location. This can be accomplished by shifting the rows Seam Line Index (0052,0036) pixels so that its Seam Line Location (0052,0033) is located at row “A-lines Per Frame * Seam Line Location / 360”. Once the seam line is positioned correctly, the Cartesian data can be obtained by remapping the Polar (z, θ) data into Cartesian (x, y) space, where the leftmost column of the Polar image corresponds to the center of the Cartesian image. Figure EEE.2-2 illustrates the Polar to Cartesian conversion. The scan-converted frames are constructed using the Catheter Direction of Rotation (0052,0031) attribute to determine the order in which the A-lines are acquired. Scan-converted frames are constructed using A-lines that contain actual data (I.e. not padded A-lines). Padded A-lines are added at the end of the frame and are contiguous. Figure EEE.2-2 is an example of Polar to Cartesian conversion.
Figure EEE.2-2 – Polar to Cartesian Conversion
An Intravascular Longitudinal Image (L-Mode) is a constrained three-dimensional reconstruction of an IVUS or IVOCT multi-frame image. The Longitudinal Image can be reconstructed from either FOR PROCESSING or FOR PRESENTATION Images. Figure EEE.3-1 is an example of an IVUS cross-sectional image (on the left) with a reconstructed longitudinal view (on the right).
Figure EEE.3-1 - IVUS Image with Vertical Longitudinal View
The Longitudinal reconstruction is comprised of a series of perpendicular cut planes, typically consisting of up to 360 slices spaced in degree increments. The cut planes are perpendicular to the cross-sectional plane, and rotate around the catheter axis (I.e. center of the catheter) to provide a full 360 degrees of rotation. A longitudinal slice indicator is used to select the cut plane to display, and is normally displayed in the associated cross-sectional image (e.g. blue arrow cursor in figure EEE.3-1). A current frame marker (e.g. yellow cursor located in the longitudinal view) is used to indicate the position of the corresponding cross-sectional image, within the longitudinal slice.
When pullback rate information is provided, distance measurements are possible along the catheter axis. The Intravascular Longitudinal Distance (0052,0028) or IVUS Pullback Rate (0018,3101) attributes are used along with the Frame Acquisition DateTime (0018,9074) attribute to facilitate measurement calculations. This allows for lesion, calcium, stent and stent gap length measurements. Figure EEE.3-2 is an example of an IVOCT cross-sectional image (on the top), with a horizontal longitudinal view on the bottom. The following example also illustrates how the tint specified by the Palette Color LUT is applied to the OCT image.
Figure EEE.3-2 - IVOCT Image with Horizontal Longitudinal View
Figure EEE.3-3 - Longitudinal Reconstruction
Figure EEE.3-3 illustrates how the 2D cross-sectional frames are stacked along the catheter longitudinal axis. True geometric representation of the vessel morphology cannot be rendered, since only the Z position information is known. Position (X and Y) and rotation (X, Y and Z) information of the acquired cross-sectional frames is unknown.