Arthroscopic image distortion-part I: the effect of lens and viewing angles in a 2-dimensional in vitro model

Arthroscopic images are subject to distortion, which may increase when using arthroscope lenses with greater reflecting angles and/or viewing structures at oblique angles. The purpose of this study was to determine the magnitude of image distortion experienced when using arthroscopes with different lens angles and when the line-of-sight (i.e., viewing angle) is not directly perpendicular to the target.

A dot calibration target was captured through 0°, 30°, and 70° arthroscopes from straight (i.e., directly perpendicular) and 30° oblique viewing angles. Distortions in horizontal and vertical distances in deep (located at 87.5 % length of arthroscopic image diameter) or shallow (12.5 % diameter length) regions were calculated, from which a deformity ratio (horizontal/vertical distance) was determined.

From the straight viewing angle (0°), both horizontal and vertical distances were artificially reduced (i.e., <100 % magnification) in the shallow and deep regions. The deformity ratio was ~100 % in the central region, declining to ~80 % peripherally. From the oblique viewing angle (30°), magnification was below 100 % in the deep area but exceeded 100 % in the shallow area, with increasing distortion associated with increasing lens angle (0° < 30° < 70°). For all lens angles, the deformity ratio was ~50 % in the deep area but neared 100 % in the shallow region.

Arthroscopic image distortion in peripheral regions should be considered when using angled-lens arthroscopes, especially when the viewing angle is not straight. As viewing the femoral ACL footprint through the anterolateral portal involves using an oblique viewing angle, visualization through the anteromedial portal is recommended.