Intraimage reproducibility of measurements in the macular area using a computerized system

Area of Science:

• Ophthalmology
• Medical Imaging
• Retinal Diagnostics

Background:

• Accurate analysis of posterior pole morphological changes is vital for diagnosing and treating retinal pathologies.
• Image digitalization techniques offer potential for precise retinal measurements.

Purpose of the Study:

• To assess the intraobserver and interobserver reproducibility of the Image-net system 100 for retinal measurements.
• To evaluate the reliability of linear and areal measurements of the retina using this system.

Main Methods:

• The 'Linear/Areal Measurement functions' of the Image-net system 100 were tested.
• Retinal images from 12 patients were analyzed for intraobserver and interobserver reproducibility.
• Measurements included distances between points, and perimeters and areas of retinal zones.

Main Results:

• Intraobserver reproducibility for linear measurements ranged from 0.32% to 7.38%.
• Interobserver reproducibility for linear measurements ranged from 0.46% to 5.22%.
• Repeatability and reproducibility for perimeter and area measurements showed similar ranges, generally below 10%.

Conclusions:

• The Image-net system 100 exhibits good reproducibility for retinal measurements.
• Factors such as fundus image quality, number of observers, and anatomical complexity can influence measurement variability.