| DC 欄位 |
值 |
語言 |
| DC.contributor | 資訊工程學系 | zh_TW |
| DC.creator | 江明謙 | zh_TW |
| DC.creator | Ming-Chien Chiang | en_US |
| dc.date.accessioned | 2020-7-28T07:39:07Z | |
| dc.date.available | 2020-7-28T07:39:07Z | |
| dc.date.issued | 2020 | |
| dc.identifier.uri | http://ir.lib.ncu.edu.tw:444/thesis/view_etd.asp?URN=107522045 | |
| dc.contributor.department | 資訊工程學系 | zh_TW |
| DC.description | 國立中央大學 | zh_TW |
| DC.description | National Central University | en_US |
| dc.description.abstract | 克雷伯氏肺炎桿菌（Klebsiella pneumoniae）是一種革蘭氏陰性菌，感染了這種病原體的患者可能會有肺炎、尿路感染和腹腔內感染伴隨嚴重的症狀，因此快速知道可有效治療的藥物是很重要的。近年來，基質輔助雷射脫附電離飛行時間質譜技術（matrix-assisted laser desorption ionization-time of flight mass spectrometry, MALDI-TOF MS）為一種新興的分析微生物的質譜方法，使用此方法可得到其相對應的質譜用來辨識其物種，也有研究用來辨識抗藥性，然而目前仍少有使用大量克雷伯氏肺炎桿菌質譜分析抗藥性的研究。本研究使用了多年來大量在長庚醫院的克雷伯氏肺炎桿菌質譜資料，並針對三種抗生素:環丙沙星（Ciprofloxacin, CIP），頭孢呋辛（Cefuroxime, CXM），頭孢曲松（Ceftriaxone, CRO），以及同時對這三種藥有效和無效的菌株資料集建立機器學習預測其抗藥性。在特徵選取之後，只使用少量的46個特徵峰值在多重抗藥的類別中，得到獨立測試準確率0.7858，其中敏感性和特異性分別為0.7298和0.8127，當中特徵峰值3657、4341、4519、4709、5070、5409、5921、5939和6516 m/z 為辨識有多重抗藥性克雷伯氏肺炎桿菌重要特徵，這些特徵峰值在多重抗藥菌株中皆佔有較高比例的特徵峰值。期望本研究當中的辨識抗藥性模型可提供協助臨床醫生第一時間判斷用藥的輔助參考，也提出相關的重要質譜峰值可供未來進一步實驗探討多重抗藥機制的原因。 | zh_TW |
| dc.description.abstract | Klebsiella pneumoniae (K. pneumoniae) is a kind of gram-negative bacteria. Patients infected with this pathogen might suffer pneumonia, urinary tract infections, and intra-abdominal infections with serious symptoms, such as toxic presentation with sudden onset, high fever, and hemoptysis. Recently, matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is an emerging technology for microbial identification. However, there have been less studies using large mass spectra dataset of K. pneumoniae to analyze multiple antibiotic resistance. Thus, we collected lots of mass spectra of K. pneumoniae and considered three antibiotics Ciprofloxacin (CIP), Cefuroxime (CXM), Ceftriaxone (CRO), and multiple antibiotic resistance or susceptibility to the three antibiotics above. After feature selection in prediction models for strains resistant or susceptible to three antibiotics above, the accuracy of independent testing can achieve 0.7858 with sensitivity 0.7289 and specificity 0.8127 using 46 features in combined dataset. The informative peaks 3657, 4341, 4519, 4709, 5070, 5409, 5921, 5939 and 6516 m/z might be the potential features for multiple antibiotic resistant K. pneumoniae and all of these peaks account for higher ratio in the resistant K. pneumoniae than in susceptible K. pneumoniae. We hope that the models for antibiotic resistance can assist doctors to evaluate the use of antibiotic in clinical. The association between resistant mechanism and informative mass spectra also needs to be further studied in the future experiment. | en_US |
| DC.subject | 基質輔助雷射脫附電離飛行時間質譜法 | zh_TW |
| DC.subject | 克雷伯氏肺炎桿菌 | zh_TW |
| DC.subject | 機器學習 | zh_TW |
| DC.subject | 多重抗藥性 | zh_TW |
| DC.subject | MALDI-TOF MS | en_US |
| DC.subject | Klebsiella pneumoniae | en_US |
| DC.subject | machine learning | en_US |
| DC.subject | multiple antibiotic resistance | en_US |
| DC.title | 基於質譜儀資料使用機器學習辨識克雷伯氏肺炎桿菌之多重抗藥性 | zh_TW |
| dc.language.iso | zh-TW | zh-TW |
| DC.title | Identification of Multiple Antibiotic Resistance of Klebsiella pneumoniae Based on MALDI-TOF MS by Using Machine Learning | en_US |
| DC.type | 博碩士論文 | zh_TW |
| DC.type | thesis | en_US |
| DC.publisher | National Central University | en_US |