麻醉的用藥劑量多寡,至今尚未有明確的介定,仍需從病人基本的生醫訊號取得,再依據麻醉醫師的經驗相對照配合而給予適當的用藥濃度,才能使病人在麻醉過程中得到安全舒適的感覺。 本論文能使其他不在現場的醫師不用進入手術房,也能了解病人當下的麻醉生理情況,是希望能減輕麻醉醫師的負擔,輔助特殊的即時協同診療的需求,使病人的基本生醫訊號借Intranet傳輸而即時存取,並模擬出病人的吸入性麻醉氣體Fiaa的給藥控制模型,來輔助麻醉醫師,能夠更輕易的了解病人的狀況與特性,讓麻醉醫生有更充裕的心力來增加麻醉過程的安全性。 本論文是將病人麻醉過程中的數據透過Intranet的解碼再編碼轉換方式存取至database,再取出分析,進而處理數據以得到較佳的效能,最後以20名病人的生醫資訊經由類神經網路來模擬訓練,使麻醉醫師在手術前能以病人的初始數據,模擬手術的狀況,進而更加了解病人的特性,增加手術安全性。 本論文中最後得到較佳預測的Fiaa之均方根誤差達0.0175,而輸出之相關係數也達到0.839的準確度。
The patients use posology of anaesthesia (i.e. unconsciousness)is not readily measurable.Therefore, we can only reply on the patients basic biomedical signals plus the anesthesiologists experience to determine a proper dose of medication, thus to ensure the safety and comfort of patients during the anesthesia posology. This thesis is to help those anesthesiologists who are not on the scene to immediately understand the patients’ physiological status during anaesthesia. Thus, the burden of anesthesiologists could be eased off. By using specific real-time synergistic diagnosis and treatment, the patients’ basic biomedical signals could be transmitted via Intranet for real-time access, so that the patients’ DOA data (Fiaa) could be simulated. The anesthesiologists could know the patients’ individual condition and characteristic better to be more capable of improving the safety of anaesthesia. In this thesis, the patients’ DOA data are accessed to database via the encoding and decoding in Intranet. After that, the data were taken out again for more analysis and further processing for better performance. Finally, the biomedical information of the 20 patients is used for simulated training through an artificial neural network. This can help the anesthesiologists to imitate the situation during the surgical operation beforehand with the patients’ initial data, thus to know the patients’ individual traits and improve the safety of the surgical operation. In this thesis, the final accuracy get to 0.0175(RMSE), and output 0.839(R).