Answers are grounded in this paper and cite their sources.
Analysis reveals how much information dictates prescriptions in clinical refraction, indicating system performance criteria.
Clinical refraction may be viewed as a procedure for generating messages—prescriptions—each containing a certain definite amount of information acquired from the patient. As part of the development of an automatic refraction system, this information flow has been analyzed, using concepts and metrics introduced by Shannon. Results from a sample block of 650 consecutive patients entering a University optometric clinic indicate that the general use prescription for both eyes contains about 23 bits of information. Total information acquired on sex, age, visual acuity, static skiametry and subjective refraction was about 60 bits, indicating 62% redundancy. Detailed analysis showed that subjective refraction could alone supply 75% of the required information, and that the inclusion of either age or visual acuity data increased that to 85%. The missing 15% of the prescription may either be genuine information derived from unrecorded or multiple sources (“clinical judgment”) or spurious information (error or “noise”). Hence minimum criteria are established for the informational performance of an automatic refraction system.
No takes yet. Share an insight, caveat, or question.
Crossman et al. (1971) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: