果蠅心臟功能參數快速量化之自動化訊號處理
Abstract
心跳週期(Heart period, HP)、心跳率(Heart rate, HR)、舒張期(Diastolic interval, DI)、收縮期(Systolic interval, SI)、舒張和收縮直徑(Diastolic and Systolic diameter, EDD, ESD)是心臟跳動的重要參數。利用以上參數可計算出關鍵的衡量指數：心律不整指數 (Arrhythmicity index, AI)和心肌收縮比率(Fractional shortening, FS)。
於本論文中，提出兩種評估果蠅心臟功能的訊號處理方法。第一種為「Brightness結合M-mode法」。每張圖的亮度在心臟舒張時變暗，在心臟收縮時變亮，所以利用亮度改變可決定心臟此時的狀態。在此方法中，能自動化取得的參數有心跳率、心跳週期、舒張直徑、收縮直徑，再藉由M-mode圖手動求舒張期、收縮期。
第二種為「Random Walker結合Auto Threshold法」。這是一種專門針對果蠅心跳收縮舒張參數所開發的快速分析的全新自動化演算法。此方法利用Random Walker分割演算法圈出心臟外圍。得到每個時間點的心臟開閉狀態後，再採用auto threshold決定收縮與舒張的分界。一旦確立後，上述每一項心臟參數都可自動化取得。這套流程十分可靠，且省時、容易操作，適合大量資料和長時間分析。
我們的演算方法結合掃頻式光學同調斷層掃描術(Swept source optical coherence tomography, SS-OCT)紀錄果蠅心跳，並利用此二種演算流程，探討在w1118、Acer 9i和w1118、AcerΔ168不同品系果蠅的心臟功能。
Heart period (HP), heart rate (HR), diastolic and systolic intervals (DI, SI), diastolic and systolic diameter (EDD, ESD) are important parameters for heart beat. These parameters can be used to calculate the decisive factor, such as arrhythmicity index (AI) and fractional shortening (FS), for determining the cardiac function of the heart. In this thesis, we proposed two processing algorithms for assessing heart condition of drosophila. The first one is called “ROI Brightness Combines with M-mode Algorithm”. The frame brightness decreases during diastole and increases during systole, so the change in brightness can determine the state of the heart. But in this algorithm we can only get HR, HP, EDD, and ESD automatically, and use M-mode graph to calculate DI and SI manually. The second algorithm is called “Random Walker Combines with Auto Threshold Algorithm”. This is a new and fully automatic algorithm that is specialized for rapid analysis of beat-to-beat contraction-relaxation parameters of the heart in Drosophila. It uses random walker segmentation algorithm to precisely circle the heart edge. After get each state of heart opening, we adopt auto threshold method to clearly distinguish the heart diastolic and systolic condition. Once obtained the dividing line between diastole and systole, we can quantify every heart-beat parameters of drosophila automatically. This methodology is robust, timesaving, and easy to use. It’s very suitable for large samples and over significant periods of time. Our methodology combines high-speed swept-source optical coherence tomography (SS-OCT) for recording of beating hearts. We use these two methods to point out the different cardiac functions in w1118, Acer 9i and AcerΔ168 strain of drosophila.
Heart period (HP), heart rate (HR), diastolic and systolic intervals (DI, SI), diastolic and systolic diameter (EDD, ESD) are important parameters for heart beat. These parameters can be used to calculate the decisive factor, such as arrhythmicity index (AI) and fractional shortening (FS), for determining the cardiac function of the heart. In this thesis, we proposed two processing algorithms for assessing heart condition of drosophila. The first one is called “ROI Brightness Combines with M-mode Algorithm”. The frame brightness decreases during diastole and increases during systole, so the change in brightness can determine the state of the heart. But in this algorithm we can only get HR, HP, EDD, and ESD automatically, and use M-mode graph to calculate DI and SI manually. The second algorithm is called “Random Walker Combines with Auto Threshold Algorithm”. This is a new and fully automatic algorithm that is specialized for rapid analysis of beat-to-beat contraction-relaxation parameters of the heart in Drosophila. It uses random walker segmentation algorithm to precisely circle the heart edge. After get each state of heart opening, we adopt auto threshold method to clearly distinguish the heart diastolic and systolic condition. Once obtained the dividing line between diastole and systole, we can quantify every heart-beat parameters of drosophila automatically. This methodology is robust, timesaving, and easy to use. It’s very suitable for large samples and over significant periods of time. Our methodology combines high-speed swept-source optical coherence tomography (SS-OCT) for recording of beating hearts. We use these two methods to point out the different cardiac functions in w1118, Acer 9i and AcerΔ168 strain of drosophila.
Description
Keywords
心跳參數, 隨機沃克分割, 自動位準, 掃頻式光學同調斷層掃描術, Heart-beat parameters, Random walker segmentation, Auto threshold, Swept source optical coherence tomography, Angiotensin converting enzyme-related