数码照相结合ImageJ医学图像分析软件法与无菌薄膜勾边法测量糖尿病皮肤溃疡面积的可靠性比较

A COMPARISON OF DIGITAL PLANIMETRY AND TRANSPARENCY TRACING BASED METHODS FOR MEASURING DIABETIC CUTANEOUS ULCER SURFACE AREA

目的探讨一种可靠、无污染、操作简便、适用于临床糖尿病皮肤溃疡表面积(diabetic cutaneous ulcersurface area,DCUSA)测量新方法。方法纳入2005年9月-2007年4月住院治疗的糖尿病皮肤溃疡患者35例。采用数码相机拍摄法(A法)收集患者不同时间点(平均每3天或1周1次)共305个溃疡图像资料,同步采用无菌薄膜勾边法(B法)收集305个溃疡面资料作为对照。将数码照片导入计算机结合Image J医学图像分析软件测得DCUSA,而B法所勾勒溃疡面轮廓通过自带坐标网格计算出DCUSA,由此获得305对面积数据。两测量者之间的一致性采用组内相关系数(intraclass correlation coefficients,ICCs)检验,并应用多元线性回归分析确定两种测量方法之间的相关关系。结果A法测得DCUSA为(4.84±7.73)cm2,B法测得DCUSA为(5.03±7.89)cm2,两种方法所测DCUSA差异无统计学意义(P>0.05)。应用A法测量DCUSA,两测量者间的ICCs为0.965,95%可信区间为(0.940,0.980);应用B法,两测量者间ICCs为0.949,95%可信区间为(0.913,0.971),不同测量者采用同样方法测得的溃疡面积之间的一致性很好。A法与B法测量DCUSA有很强的相关性(r=0.957,P<0.05)。结论数码照相结合Image J医学图像分析软件能较准确地测量糖尿病皮肤溃疡面积,且是一种无污染、易操作、更适用于临床的溃疡面积测量方法。

Objective To assess the reliability of diabetic cutaneous ulcer surface area （DCUSA） measurement using digital planimetry method （A） and transparency tracing method （B）. Methods Images of diabetic cutaneous ulcers from 35 inpatients with diabetic skin ulcers from September 2005 to April 2007 were taken by a digital camera once a week or twice a week over a period of 12 weeks, resulting in 305 photographs; the ulcers were traced on a grid with acetate wound tracings, simultaneously. A total of 305 pairs of DCUSA which were calculated respectively throughout digital camera combined with Image J medical imaging software and transparency tracing with grid sheet by two independent observers sequentially were obtained. The intraclass correlation coefficients （ICCs, one-way random effect model） was used as an indicator of chancecorrected agreement to estimate the relative reliability for the interobserver data. Multiple linear regression analysis was also used to measure the relationship of these two methods. Results DCUSA obtained from method A and obtained from method B was （4.84 ± 7.73） cm^2 and （5.03 ± 7.89） cm^2, respectively; no significant difference was found （P 〉 0.05）. ICCs was high （ICCs=0.949 for method B and 0.965 for method A）, indicating that the relative reliability for the interobserver was excellent. The method A were highly correlated with measurements obtained from method B （r = 0.957, P 〈 0.05）. Condusion The digital planimetry method described in this study represents a simple, practical, without any wound damage and contamination, and inexpensive technique to accurately evaluate the areas of diabetic cutaneous ulcers. The photographic technique combined with Image J medical imaging software should be considered for wound measurement.