Kim S C, Kim D W, Hong J P, Rah D K
Graduate Program in Biomedical Engineering, Yonsei University, Seoul, Korea.
Yonsei Med J. 2000 Jun;41(3):333-9. doi: 10.3349/ymj.2000.41.3.333.
The evaluation of pigmentary skin lesions by clinical doctors has been based on subjective and qualitative judgements. Observations have mostly relied on visual inspection, making the effects of treatment difficult to evaluate with any precision. For this reason there is a real need for an objective method to evaluate prognosis after treatment. Recent scientific measurements such as reflectance spectrophotometry and reflectance colorimetry have provided accurate quantitative color information about skin lesions, but these techniques are costly and difficult to apply in the clinical field. The purpose of this study was to develop a simple and cost-effective way of evaluating treatment results. We have developed a software program using the Lab* color coordinate system to quantify the effect of treatment and have successfully demonstrated its clinical usefulness. Our method compares the relative color difference between normal skin and skin lesions before and after treatment, instead of measuring the absolute color of skin lesions. The accuracy of our quantitative color analysis was confirmed by the simulated images of hemangioma and ota nevus. Clinical efficacy was also confirmed through a blind test involving 3 clinicians who were asked to grade the treatment effects of 13 cases of hemangioma and 7 cases of ota nevus. These subjective clinical grades correlated well with the treatment results obtained using the proposed color analysis system (Correlation coefficient = 0.84).
临床医生对色素性皮肤病变的评估一直基于主观和定性判断。观察大多依赖于目视检查,使得治疗效果难以精确评估。因此,确实需要一种客观方法来评估治疗后的预后。诸如反射分光光度法和反射比色法等最新科学测量方法已提供了有关皮肤病变的准确定量颜色信息,但这些技术成本高昂且难以应用于临床领域。本研究的目的是开发一种简单且经济高效的评估治疗效果的方法。我们利用Lab*颜色坐标系统开发了一个软件程序来量化治疗效果,并成功证明了其临床实用性。我们的方法比较治疗前后正常皮肤与皮肤病变之间的相对色差,而不是测量皮肤病变的绝对颜色。通过血管瘤和太田痣的模拟图像证实了我们定量颜色分析的准确性。通过一项盲法试验也证实了临床疗效,该试验让3名临床医生对13例血管瘤和7例太田痣的治疗效果进行评分。这些主观临床评分与使用所提出的颜色分析系统获得的治疗结果高度相关(相关系数 = 0.84)。