Artifact Removal procedure distorts multifocal electroretinogram

Purpose: To study whether the Artifact Removal procedure available for elim inating artifacts in multifocal electroretinograms (mERG) works correctly o r not. Methods: A test response was made using a photo-diode circuit. mERGs were r ecorded from 3 well-trained normal subjects using the Veris III system, and were then analyzed by the procedure that is included in the Veris Science (Artifact Removal) software program. The stimuli consisted of densely arran ged arrays of 103 or 37 hexagonal elements. It took a total of 8 minutes to obtain one mERG record, and 16 sessions were required to complete this rec ord. The first-order as well as the second-order kernel response components were extracted by Veris Science software, and the Artifact Removal procedu re was used for both components. Results: The Artifact Removal procedure influenced both the test response o n the center element as well as the neighboring traces just around the test response. After the repetitions of the Artifact Removal procedure, the sha pe of the test response changed considerably. Some of the traces of the sec ond-order kernel response components elicited from a normal subject changed irregularly when the Artifact Removal procedure was repeatedly used. The n oise increased at the first iteration of the Artifact Removal procedure. Conclusion: This procedure has been considered useful for eliminating artif act distortion in mERG, but should be carefully checked by well-established testing methods before clinical use. (C) 2000 Japanese Ophthalmological So ciety.