長期以來，中醫古書配方一直被用於治療患者。當出現外部傷口不幸被感染時，常伴隨著蜂窩性組織炎、組織化膿、傷口潰爛，近年許多文獻提到中醫外傷藥良好的治療效果，透過調控生理訊號還有中藥的活性成分，不僅能夠加速傷口癒合，也能減少疤痕的產生，所以我們採用明代最多方劑的醫書《普濟方》作為我們資料探勘來源，建立中藥方劑資料庫，用以資料探勘找出關聯性，為了方便驗證藥物預測的正確性，我們挑選方便實驗進行的外傷類藥物，建立藥物關聯性模組，接著我們整理金創門、杖瘡門、刺瘡門的資料，內容包含六卷四百九十方，使用Apriori Algorithm分析找出外傷藥的兩種配方，其中分為內服以及外敷用藥，進行細胞學實驗及爬行實驗驗證配方確實有助於成纖維母細胞生長速度。最後，我們以預測藥物混合水凝膠進行老鼠實驗，發現傷口較深的患者，以外敷用藥（74.36％）的恢復程度優於對照組（70.14％），且外敷藥物在6天內有顯著的復原情況。可知所預測的藥物配方均較於純水凝膠的實驗組有著較好的效果。最後以蘇木紫-伊紅染色結果，可以看出在外敷用藥的實驗下，傷口的表面形狀完整、復原良好，可預期我們以這樣的模組進行各大類方劑的配方預測，能夠找到更多新穎的核心藥方。;The Traditional Chinese Medicine (TCM) formula has been used to treat patients for a long time. We used the prescription of the Ming Dynasty medical book "Pu-Ji-Fang" to establish a database of Chinese medicine prescriptions for data exploration to find out the relevance, in order to facilitate the verification of drug predictions. We selected the traumatic drugs that are convenient for experiment, and established drug-related modules. Many literatures also mentioned that the traumatic drugs of traditional Chinese medicine can not only accelerate wound healing, but also reduce the scar production. Then we organize Jin-Chuang-Phylum, Zhang-Chuang- Phylum and Chih-Chuang- Phylum data, contain 6 volumes and 410 formula, using Apriori Algorithm to identify two formulations of traumatic drugs, which were divided into oral and external application. The cytology experiment and the wound healing test proved that the traumatic formulation really contributed to the growth rate of fibroblasts. Finally, we conducted rats experiment with a predictive drug-mixed hydrogel. The external wound medicine (74.36%) has better recovery degree than control group (70.14%) for deeper wounds, and the drugs have a significant recovery within 6 days. It can be seen that the predicted drug formula has better effects than the experimental group of pure hydrogel. In the future, we will use these modules to predict the formulation of various major prescriptions, and we can find more novel core prescriptions.
