Study on the methodology of striae gravidarum severity evaluation

Autor: Hongyan Dai, Yangyang Liu, Yan Zhu, Yun Yu, Lin Meng
Jazyk: angličtina
Rok vydání: 2021
Předmět:
Zdroj: BioMedical Engineering OnLine, Vol 20, Iss 1, Pp 1-12 (2021)
Druh dokumentu: article
ISSN: 1475-925X
DOI: 10.1186/s12938-021-00945-w
Popis: Abstract Background Striae gravidarum is a common occurrence in pregnancy and many women expect to prevent its development. At present, laser treatment has been used to improve the appearance of striae gravidarum, but the choice of laser type, treatment time, and frequency depend on the therapeutic effect. How to obtain an effective evaluation of striae gravidarum during and after treatment is very important. However, there is no unified evaluation parameter about striae gravidarum. In this paper, we studied the methodology evaluation of striae gravidarum severity. First, the laser therapeutic apparatus was selected as the experimental equipment and different striae gravidarum photos during treatment were obtained. Second, the subject evaluation parameters were chosen based on the literature research and the dermatologists’ guidance. Then, the striae gravidarum photos were divided into different groups by dermatologists based on these parameters. Finally, the objective detection parameters were designed based on the photos feature and subject evaluation parameters. Then, the objective detection parameters were used as the input of the support vector machine and the evaluation results were compared. Results Based on the subject evaluation parameters, the experimental data could be divided into mild, moderate and severe groups. The experiment results showed that the striae gravidarum severity of two randomly patients were improved before and after treatment, which verified the validity of the parameters. In addition, the chosen objective detection parameters were different among different groups. With all the objective parameters as the support vector machine input, we could achieve the best recognition rate (82.71%) in the striae gravidarum severity classification. The four parameters (color difference, average density, average width, distribution area) calculated from the photos as the input could achieve acceptable accuracy (81.69%). Conclusions The subject evaluation parameters and objective detection parameters proposed in this paper can be used to evaluate the striae gravidarum severity, which is of great significance for the construction of auxiliary diagnostic instrument for striae gravidarum treatment.
Databáze: Directory of Open Access Journals