Validation of computer morphs for grading contact lens complications

Area of Science:

• Ophthalmology
• Optometry
• Biomedical Imaging

Background:

• Assessing contact lens-induced ocular pathology severity traditionally relies on printed grading scales.
• Printed scales may have limitations in precision and application.
• Computer-generated grading morphs offer a potential alternative for standardized assessment.

Purpose of the Study:

• To validate the accuracy and reliability of computer-generated grading morphs compared to traditional printed scales for contact lens complications.
• To evaluate the utility of computer morphs in clinical practice and research.

Main Methods:

• Nine experienced optometrists graded 16 contact lens complications using both printed scales (Efron Grading Scales) and computer morphs (Efron Grading Morphs).
• Grading was performed to the nearest 0.1 increment.
• The entire procedure was repeated after approximately two weeks to assess test-retest reliability, generating 576 grading estimates.

Main Results:

• Computer morphs demonstrated good accuracy, with similar mean test and retest grading estimates compared to printed scales (Morphs: 2.6 +/- 0.8 vs. Printed: 2.8 +/- 0.7).
• No significant difference was found in median reliability between computer morphs (+/- 0.43) and printed scales (+/- 0.41).
• Statistical analysis showed no significant difference in accuracy (F1,8 = 2.7, p = 0.14) or reliability (Z = 0.1, p = 0.95) between the two methods.

Conclusions:

• Computer-generated grading morphs are validated as an accurate and reliable tool for assessing contact lens-induced ocular pathology.
• Computer morphs offer significant advantages for computer-based clinical record keeping, education, and research.
• This digital approach provides a robust alternative to printed grading scales in optometric practice.