Figure 1.
A transparency model sheet was designed in this study for wound area assessment, which included 340 pieces of tracing results, the date of experiment, observation times, the types of treatments, group numbers, animal/sample numbers, the time of wound closure as well as a set of standard area units/controls.
Figure 2.
Computer-aided wound area measurement with everyday-used software.
A. JPG image of the model sheet displayed in the window of Adobe Photoshop. Arrows indicate the wound regions outlined in turn in black, red and blue at Day-1, Day-3 and Day-5 and then at Day-7, Day-10 and Day-12 time points. Star mark: the day of wound healing. B. The locations of magnetic lasso tool button for defining the selected traced region (left) and “Histogram” button for calculating pixel value (right). C. Arrow indicates Magnetic lasso tool labeled R1 No. 2 wound region at Day-3. The pixel value of this region was calculated by pressing “Histogram” button.
Figure 3.
Shortcomings of traditional transparency-based wound assessment.
A. In traditional transparency-based wound healing assessment, a hundred pieces of transparency films had to be prepared for marking out the margins of individual wounds at different observation times. The outlined regions as well as the suitable standard area control(s) were cut off along the margin with an electronic cutter and weighed on an analytical balancer. B. The shapes and relative sizes of four standard area units isolated by electronic cutter (Traditional) or labeled with Magnetic lasso tool in Adobe Photoshop program (New). The insets are the images in original sizes.
Table 1.
Average wound areas of two experimental groups calculated with two methods on Day-10 for three times.
Figure 4.
Remarkable differences of measurement accuracy between transparency-based digital imaging and traditional transparency methods.
A. The wound areas of non-treated (N) and Reagent-1 treated (R1) groups calculated with transparency-based digital imaging (New) by three researchers and with transparency tracing method (Traditional) for three times on Day-10. Upper-right inset: the average sizes of the calculated regions with the two methods. B. Comparison of the accuracy of the calculated area data generated from transparency-based digital imaging and traditional transparency methods. The numbers on the top of each of the columns indicate the areas calculated according the pixel values or the transparency weights of the four standard units (1 mm2, 4 mm2, 25 mm2 and 100 mm2). *, P = 0.000<0.01; **, P = 0.000<0.01. #, P = 0.081>0.05; ##, P = 0.979>0.05. C. Reproducibility of the new and traditional wound assessment methods performed on Day-10 traced data. The pixel values and the transparency weights of 1 mm2 and 100 mm2 standard units were used as numerators, respectively. *, P = 0.857>0.05; **, P = 0.889>0.05. #, P = 0.000<0.01; ##, P = 0.000<0.01.
Figure 5.
Comparison of the areas calculated with a square (25 mm2) and a circle (24.62 mm2) standard units.
A. Area calculation performed on seven personal computer/PC-designed round standard controls in the known mathematic sizes. No statistical difference was found between the mathematic areas and the areas calculated with the PC-designed square or circle area unit (t-test; P>0.05). B. Area calculation performed on six transparency tracings marked out from a wound. In the middle histogram, the yellow part of the column indicates the merged areas calculated with the square (red) and the circle (green) standard unit. No distinct variation was found between the areas calculated with the scanned square and circle standard area units (t-test; P>0.05). C. Bland Altman Plots for assessing agreement between the areas measured with 25 mm2 square (S) and 24.62 mm2 circle (C) standard units. Mean±2SD represents 95% confidence interval of the Ratio (S/C). Coefficient of variation (CV): 0.31% for PC-designed and 2.52% for scanned objects. 95% limits of agreement: 0.98-1.01 for PC-designed and 0.97-1.03 for scanned objects.