An Image Model for Quantal Response Analysis in Perimetry
针对青光眼等眼病诊断中的视野检查，开发了一个马尔可夫随机场模型，用于高效估计阈值并分类正常或缺陷点，相比现有方法均方误差降低10-30%。
Measurement of the patient's seeing at different locations in the visual field is an important diagnostic tool for glaucoma and other eye diseases. A Markov random field model is developed and used for efficient estimation of the thresholds and for classification of points as normal or defective. The model allows for non-homogeneous spatial dependence and non-symmetric marginal distributions and has physically interpretable parameters. ICM threshold estimation resulted in 10-30% (depending on the patient population) reduction of mean square error as compared to currently used procedures and in a fair agreement between true and estimated defect status. Marginal posterior mean estimates had the same efficiency, but required more computation time. Non-standard features of the problem are: (i) a non-homogeneous directional dependence, (ii) thresholds are only measured indirectly, by binary responses to questions, where the probability of response depends on the threshold and the stimulus level.