Abstract
Abstract: :
Purpose: To optimize and standardize the image acquisition protocol (since no accepted one currently exists) for the Heidelberg Retinal Flowmeter to 1) produce a methodology that gives a high number of analyzed valid pixels, 2) to allow different investigators to compare their data and results by measuring equivalent retinal capillary regions, and 3) to compare two current protocols to determine the yield of valid pixels and to see if the one that includes more measured rim area is advantageous. Methods: Two protocols were compared. In both, each of the 4 images were obtained using an acquisition window of 2.7mm X 0.7 mm (16,384 pixels) over 2 seconds which were analyzed with the Automatic Full Field Perfusion Image Analyzer software (AFFPIA-SLDF v. 3.3). The number of pixels that the program determined were valid (analyzed %), was compared between the two protocols. Protocol A made measurements focusing on the retina 1) superotemporal and 2) inferotemporal to the optic nerve, and across the 3) superior one-half of the entire nerve rim including nasal retina and 4) inferior one-half of the entire nerve rim including nasal retina. Protocol B made measurements focusing on the retina 1) superotemporal and 2) inferotemporal to the optic nerve, 3) across the entire middle portion of the nerve rim with nasal retina and 4) across the nasal rim of the nerve with nasal retina. 80 scans from Protocol A were compared with 88 scans of Protocol B and the number of valid pixels in the superotemporal retina (STR), inferotemporal retina (ITR), optic nerve rim (ONR), and nasal retina (NR) were analyzed. Results: Mean values for Protocol A and Protocol B were: STR: A 9,900 2440 , B 11,813 1343 (p=.003); ITR: A 10,115 1367, B 11,358 1673 (p=.01); ONR: A 2984 1879, B 2984 1118 (p=1.0); NR: A 12,098 3236, B 10,259 2648 (p=.05). Both protocols were equally repeatable (p≷0.3). Conclusion: STR and ITR imaging yielded more valid pixels with Protocol B; ONR yielded an equal number of valid pixels in both protocols even though more rim was measured with Protocol A; NR yielded more valid pixels in Protocol A but this is probably the least important area. Protocol A's greater rim measurement did not yield more valid pixels than Protocol B's because it is difficult to focus on the optic nerve rim and the retina simultaneously. We have adopted Protocol B as our standardized protocol and believe this should be universally used.
Keywords: 331 blood supply • 430 imaging/image analysis: clinical • 432 imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)