Risk Factors Analysis for Post Inflammatory Hyperpigmentation Complication after Facial Pigmentation Treatment Using 1064 nm Nd:YAG Laser with Q-Switched

• Main Authors: PAN, JIAN-ZHI, 潘建智
• Other Authors: LEE, TSAIR-FWU
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/7dg558

碩士 === 國立高雄應用科技大學 === 電子工程系 === 106 === Purpose : To evaluate the risk factor analysis for post inflammatory hyperpigmentation complication after facial pigmentation treatment using 1064 nm Nd:YAG laser with Q-Switched. Materials and methods : This study collects data from 166 female patients receiving laser toning in facial between January 2014 and January 2016. After excluded 6 patients over the age of 70 and 4 patients who treated by hybrid laser mode remaining 156 patients. According to have occurred post inflammatory hyperpigmentation or haven't occurred post inflammatory hyperpigmentation can being classified. In predictors analysis, we uses physician to determine the predictors may cause PIH as all predictors, the clinical predictors of patient’s own characteristics and uses least absolute shrinkage and selection operator (LASSO) to select the risk predictors. Furthermore, the classification models of the selected risk predictors, all risk predictors and clinical risk predictors were established five kinds of algorithms by hierarchical clustering (HC), K-means clustering (K-means), naive Bayes (NB), C4.5 decision tree (C4.5) and logistic regression (LR). Finally, uses the area under the receiver operating characteristic curve (AUC), accuracy (ACC) and negative predictive value (NPV) was compared. Results : The area under the receiver operating characteristics curve (AUC) of HC model with three risk predictors was 0.45, 0.55 and 0.53, respectively. The AUC of K-means model with three risk predictors was 0.44, 0.55 and 0.55, respectively. The AUC of NB model was 0.74, 0.72 and 0.73, respectively. The AUC of C4.5 model was 0.72, 0.71 and 0.76, respectively. The AUC of LR model was 0.73, 0.76 and 0.71, respectively. Conclusions : The results of this study showed that the LASSO can select SPT, AHA Peels, IEL, Melasma as risk predictors easily influence by PIH. Therefore, this study suggests that physicians should pay special attention to the above four factors when using laser to treat facial pigmentation symptoms, in order to reduce the side effects of patients after receiving pure skin laser treatment and increase patient satisfaction.