從頸部斷層掃描影像自動偵測出異常淋巴及腺體之研究

Abstract

在醫學影像中，電腦斷層掃描(Computerized Tomography, CT) 技術應用於放射性檢查，而頸部斷層掃描檢查常常應用於研究頸部的腫瘤或腫塊，以及尋找頸部淋巴結腫大；然而，頸部CT影像中組織複雜，醫師於診斷時，時常基於此種因素，不斷猜測病灶的正確位置及大小，陷入膠著的狀況。 本研究的目的為利用影像處理與資訊技術的方式針對頸部CT影像進行電腦自動化醫學分析，並將結果顯示於電腦畫面或醫學儀器上，藉由此種方式，讓醫師更快了解病人影像中的狀況，進而減輕醫師負擔外，更使得效率提升；而此項技術未來更可應用於協助醫師進行診斷，且使醫生誤判率下降，以增進診斷過程的精確度。 本論文的主要貢獻如下：第一，本論文利用適於頸部CT影像軟組織分析方法，改善頸部CT影像中組織分區的缺點，可有效分辨異常淋巴結及軟組織。第二，提出一可協助病灶偵測之流程雛形及方法。本論文最終之實驗結果顯示本輔助診斷系統在頸部CT影像上能有效地找出疑似病灶之可疑區塊，為醫師提供更簡便及有效的輔助診斷資訊。

In the medical images, computerized tomography (CT) technology is used in radiological examinations, and neck CT scanning inspection is often applied in research of neck tumors or masses as well as in searching for the cervical lymph node to be tumescent; however, due to the fact that tissue of neck CT is complex, doctors are often confused about the correct location and size of the lesion when they are diagnosing, and then trapped into a stalemated situation. The purpose of this study is the use of image processing and information technology for automatically computer-aided medical analysis on neck CT images, and displays the results on a computer screen or medical instruments. By this way, doctors are allowed to understand the patient's situation more quickly. Moreover, this technology not only alleviates the burden on doctors, but also promotes the efficiency. The main contribution of this paper is as follows: First, this paper uses a suitable analysis method on soft tissues of the neck CT image to improve the shortcoming when partitioning them, and effectively distinguish the abnormal lymph nodes and soft tissue. Second, this paper proposes a prototype and methods to assist the process of lesion detection. In this thesis, the final experimental results show that the computer-aided diagnosis system in the neck CT imaging can effectively identify the suspicious block of suspected lesions and provide doctors with simpler and effective assistant diagnostic information.