近年來，隨著質譜儀在各醫療機構的普遍應用，在大型的醫療機構都已經累積了數量可觀的資料，但是目前能使用質譜儀資料的軟體皆來自於質譜儀的廠商，除了高額的軟體授權費之外，也只能使用廠商提供的方式來使用資料。因此，在本研究中，我們與長庚醫療財團法人合作，建立了一個資料視覺化的工具，能讀入現有的大量歷史資料或是新產生的資料，並產生視覺化的資料圖表，也能與不同的分類模型結合，訓練分類模型來預測抗藥性結果，並將結果資料整合顯示於圖表之中，使臨床或是研究人員能更容易觀察大量的資料。;In recent years, as matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) widely used in the field of biomedical, most of the medical care institutions has accumulated quite amount of data, but for now, the only way to access data of MALDI-TOF is through the software packaged by MALDI-TOF company. Except the high license fee charged, we can only use the data in limited way that the company implement in the software. Based on the fact we mention above, we cooperate with Chang Gung Memorial Hospital, and develop a program that can present visualized resistance data from MALDI-TOF and is able to combine with different feature selection algorithms to predict resistance result, let our users easily compare between different data and different features. With our own designed program, we can adjust our user interface, set parameters or extend functions to fit our own user. And as our closely cooperation with our users, we can handle our user’s feedback more efficiently than the original software.
