Title: 咽喉胃酸逆流之嚴重程度診斷系統的設計
The Design of Diagnosis System for the Degree of Laryngopharyngeal Reflux
Authors: 羅國彰
Kuo-Chang Lo
Sheng-Fuu Lin
Keywords: 診斷系統;咽喉胃酸逆流;可變形模板;色彩分析;類神經網路;Diagnosis system;Laryngopharyngeal reflux;Deformable template;Color analysis;Neural network
Issue Date: 2007
Abstract: 咽喉胃酸逆流疾病引起的主因是胃酸在逆流的過程中對咽喉產生損傷,且同時在食道或是胃也有相關併發症發生。目前臨床上主要透過內視鏡對器官組織進行檢查,但容易因為醫師個人主觀的因素,使得臨床診斷準確率不高,而標準咽喉胃酸逆流疾病的診斷方式需住院24小時,且為一種侵入性的檢查,不利於門診施行。而在可以把咽喉胃酸逆流發病情況診斷出來的情形下,其同時在食道及胃也有不同程度損傷之併發症,系統根據醫師臨床給藥標準將併發症種類予以分級,其結果可輔助醫師診斷咽喉胃酸逆流的嚴重程度,且具有臨床價值。本論文的研究目的在於發展一套可以處理喉部影像的分析技術,可以自動分離出診斷上的特徵區域,並且大量收集正常無症狀及合併有各種情形併發症之咽喉胃酸逆流案例進行統計分析與比較,藉由電腦以更客觀的方法輔助醫師做出更精準的診斷結果。 本論文的主要貢獻有三,第一,利用特徵追蹤的技巧,解決以人工方式從內視鏡影像序列中,挑選出合適的喉部影像再進行分析的耗時缺點。第二,利用數位影像處理的技巧,將各個部位上的特徵以數值表示,以利於尋找正確的數據以診斷病症。第三,完成一個咽喉胃酸逆流嚴重程度診斷系統,實驗結果顯示本系統在診斷的正確度上能夠達到不錯的成效。
Laryngopharyngeal reflux disease is caused by the acid in the refluxing process to make the larynx injured, and in esophagus or stomach also have others complications. Recently, in the clinical observation, the organs are inspected mainly through the endoscopy. As a result of the physician’s subjective, the clinical diagnosis accuracy is easy to be low. However, the standard way to diagnosis laryngopharyngeal reflux needs to be in hospital for 24-hours, and is not a physical examination, so it is not convenient in the clinic. In a situation of the laryngopharyngeal reflux can be diagnosed, the esophagus and the stomach also suffered the same lesions simultaneously. According to the standards of clinical diagnoses, system can help the physicians to diagnosis the degree of laryngopharyngeal reflux. The purpose of this thesis is to design a system which can analyze the image of fauces and separate the image with different region of fauces automatically. By collecting a lot of normal cases and symptom cases with others complications for statistic analysis and comparison to finding an objective way to assist the physician make a precise diagnosis with computer. The contributions of this thesis may be summarized as follows. First, use feature-based tracking technique to solve the disadvantage of time consuming when the suitable image of fauces is picked out from the endoscope image sequences by hand. Second, use the technique of the digital image processing, transforming the features of image to different values, in order to find the correct way for diagnosis. Third, complete a diagnosis system for the degree of laryngopharyngeal reflux. The experimental results show the system is satisfactory.
Appears in Collections:Thesis