基於支援向量機之新生兒篩檢系統

Abstract

新生兒篩檢系統是用血液去檢驗新生兒新陳代謝相關的先天異常疾病，因為這些先天異常的疾病若能及早診斷，並給予適當的治療或預防，使之不至於發病造成身心永久性的傷害，例如智能不足等嚴重之後果。所以這些疾病可以利用新生兒篩檢系統的功能來達到預防及早治療的目的，則病患就有正常成長的機會或將此疾病的後遺症降至最低。

在本論文，我們提出一個基於支援向量機的新生兒篩檢系統。我們利用支援向量機優異的資料分類功能，對於串聯質譜儀分析出的新陳代謝物濃度數據做預測，評估此新生兒是否罹患先天性新陳代謝異常的疾病。在服務導向架構SOA下，我們將支援向量機的功能架構在分散式環境,，利用中介軟體Web Services 整合異質的平台、服務與資料庫。利用此系統來預測新生兒是否罹患甲基丙二酸血症，它的敏感度Sensitivity 可於由傳統Cut-Off Value方法的76%提升到超過99%。

The clinical symptoms of metabolic disorders during neonatal period are often not apparent, if not treated early irreversible damages such as mental retardation may occur, even death. Therefore, practicing newborn screening is very important to prevent neonatal from these damages.

In this thesis, we propose the newborn screening system that using support vector machines (SVM) classification technique to evaluate the metabolic substances concentration raw data obtained from tandem mass spectrometry (MS/MS) and determine whether the newborn has some kind of metabolic disorder diseases. With concept of Service-Oriented Architecture, we design the system using Web Services technique which is suitable for integrating heterogeneous platforms, protocols and applications. In this system, the predicting accuracy (sensitivity) of MMA could be improved from 76% (cut-off value approach) to over 99%