Noninterchangeability of Vessel Density Measurements Across Next-Generation SS-OCTA Systems

Comparing swept-source optical coherence tomography angiography (OCTA) vessel density in the superficial and deep capillary plexuses across leading OCTA systems.

OCTAs were acquired from healthy participants using two OCTA systems (BMizar 400 kHz and Yalkaid 100 kHz, TowardPi, Beijing, China) with scan sizes of 3 × 3, 6 × 6, 12 × 12, and 15 × 15 mm2. The vessel density was quantified using the manufacturer's software.

Across 30 subjects comprising 60 eyes and a total of 225 OCTA scans, significant interdevice differences were observed at the Early Treatment Diabetic Retinopathy Study sector level in all eight scan size × plexus comparisons (three-way analysis of variance; P < 1.16 × 10e-4). The mean BMizar vessel density values for the superficial and deep plexuses were 43.4% ± 1.6% and 44.9% ± 1.5%, 44.7% ± 2.7% and 47.9% ± 1.1%, 44.2% ± 1.3% and 45.7% ± 1.1%, and 42.4 ± 1.6 % and 44.3 ± 1.3% for the 3 × 3, 6 × 6, 12 × 12, and 15 × 15 mm² scans; the Yalkaid values were lower. BMizar showed greater vessel density than Yalkaid, particularly in 3 × 3 mm² deep slab scans, with differences of approximately 5% to 10%. Across scans larger than 3 × 3 mm², BMizar showed greater central vessel density; the peripheral patterns varied, with reversals favoring Yalkaid and greater border effects in larger fields of view.

Despite interdevice differences, both devices showed consistent OCTA acquisition.

Device differences may affect clinical interpretation, longitudinal follow-up, and multicenter studies.