吸毒成癮一直是社會關注的焦點。它不但在社會上引起很多問題，而且影響國家和社會安全。吸毒者的年輕化顯示與毒品有關的犯罪正變得越來越嚴重。因此，如何通過治療減少吸毒者對毒品的渴望是臨床上的挑戰。雖然對海洛因成癮者有美沙酮或丁丙諾啡等藥物的減害治療方法，但對有甲基安非他命使用障礙（MUD）的成癮者並沒有藥物治療。暴露療法結合生物反饋是毒癮治療的新方法。為了誘發毒癮患者對毒癮的渴望來成功實現暴露療法所需要的環境刺激，虛擬現實（VR）扮演著重要角色。我們的研究是開發一種帶有嗅覺模擬的VR系統以誘發MUD患者對毒品渴望來進行治療。藉由結合多種傳感器的技術：例如腦電圖（ECG）、心電圖（ECG）、皮膚電反應（GSR）和眼球跟?，用以記錄每個MUD患者在虛擬環境中的各種生理和反應行為。通過統計和機器學習的方法來評估誘發渴望的強度。根據臨床實驗，在VR刺激前和VR刺激後的統計分析結果發現，MUD患者和健康受試者的生理特徵和神經行為方面存在顯著差異。此外，應用了多種機器學習方法的結果表明，MUD患者在VR刺激前和VR刺激後之間的分類準確性高於健康受試者。綜上所述，所提出的VR系統經過驗證，可以有效地誘導MUD患者的毒品渴望。;Drug addiction has always been the focus of social attention. It not only causes a lot of problems in society, but also affects national and social security. The younger population of drug addict shows that the current situation of drug-related crimes is becoming more and more serious. Therefore, how to reduce the craving for drug addict via treatment is a challenge in clinics. While there is methadone or buprenorphine harm-reduction treatment for heroin addicts, there is no drug treatment for addicts with methamphetamine use disorder (MUD). Exposure therapy integrating with biofeedback is new method of drug addiction treatment. In order to induce the craving of drug addicts to successfully achieve the environmental stimuli required by exposure therapy, Virtual Reality (VR) plays an important role. Our research is to develop a VR system with flavor simulation to induce the drug cravings of drug addicts for MUD patients in therapy. Combined with multiple sensor technologies, such as Electroencephalography (EEG), Electrocardiography (ECG), Galvanic Skin Response (GSR), and eye tracking, to record the various physiological and reactive behaviors of each MUD patient in the virtual environment. Through statistics and machine learning method to evaluate the intensity of craving induced. According to clinical experiment, the results of statistical analysis found that there are significant differences in the physiological characteristics and neuro-behavior of MUD patients and healthy subjects between pre-VR stimulation and post-VR stimulation. Further, several machine learning methods were applied and showed that the classification accuracy between pre-VR stimulation and post-VR stimulation on MUD patients was higher than on healthy subjects. In conclusion, the proposed VR system was validated to effectively induce the drug craving in MUD patients.
