[Velocity model for Heidelberg retinal flowmetry]

An application of scanning-laser ophthalmoscopic flowmetry, Heidelberg Retina Flowmetry (HRF) is reported to faithfully quantify retinal perfusion. Indeed, without convincing regard for the fact that the method produces numerical values for such so-called "VELOCITY" (and subsequently "calculates" "FLOW", and "VOLUME") with no physical units or proven real-life meaning, current clinical drug studies and patient care are being based on this method. To test the physical "fidelity" of the HRF method, the author hence devised a simple, reliable mechanical model to generate known velocities of movement of a test surface.

Movement of the circumferential surface of an 89 mm (3 1/2-inch) Ø cylinder, belt-driven by a brass spindle with segments of increasing diameter, was "analyzed" with the HRF method. The true velocities (mm/sec) with which the surface passed the HRF's focal point were then determined using a stopwatch. A 10 degrees x 2.5 degrees measuring field and a 20 x 20-pixel analysis window were used. Measurements were made for horizontal, diagonal (45 degrees and 135 degrees), and vertical motion at all velocity settings of the model.

The relationship between real velocities of cylinder-surface motion and the corresponding HRF-"VELOCITY" values was nonlinear, an effect which increased significantly for diagonal and vertical motion.

Considering the dependency on orientation of motion and the non-linearity of the relationship between HRF results and true velocity, as well as several other weaknesses discussed herein, the question arises whether the validity of the HRF method should be reconsidered.