Figure 65

Two applications of vessel segmentation: mosaicing and spatial referencing. Panels A through E show some of the images taken from the same eye on a single visit. Panel F shows the mosaic (or spatial map) built from these and other images [23] using traces of the vasculature as spatial landmarks, and with a desired treatment area outlined. Panel G simulates an image acquired during a surgical procedure in which the aiming point in G is referenced against the spatial map in real-time to determine the precise location of the laser and check if it is within the desired treatment area before allowing the laser to be turned on [95].

Figure 6.5 illustrates spatial mapping. Panel G simulates an image frame captured by a camera during a surgical procedure. This image is registered onto the spatial map, effectively estimating the location of a laser beam (illustrated by the cross-hairs on panel G) relative to the spatial map. The arrow illustrates this mapping. The vascular traces from the image frame (panel G) are superimposed on the mosaic to show correct registration. Spatial referencing can, for example, determine whether or not the surgical laser is aimed within a desired treatment region. While performing laser retinal surgery, the images that the surgeon is viewing can be processed for vascular landmarks in much the same fashion as they were extracted for registration. These landmarks can then be compared against the landmarks in the mosaic image so as to identify whether the surgical tool is in the correct region. If so, the surgeon can then apply the laser to the desired treatment area, while the spatial referencing algorithms enable a recording of the optical dosage administered to each point. If not, the spatial referencing processor can perform a protective cut-off of the laser.

Thus, the extraction of vessels is vital in providing the landmarks necessary to perform such spatial tasks in computer-assisted surgical instrumentation.

0 0

Post a comment