心房顫動是因心房在正常起搏點以外的其它位置出現不正常的快速電氣活動，蓋過原本節律所產生。 心房顫動可導致病人心臟衰竭、提高中風等多項併發症風險。傳統治療方式多以服用抗心律不整藥物 為主，但藥物控制療效往往不如預期，且治標不治本。導管電氣燒灼術由於具備微創、安全、併發症 少與操作容易等優點而迅速普及，儼然已成為藥物治療之外的主流療法。然而目前慢性心房顫動患者 （指持續出現心房顫動超過7 天以上）因心房顫動發生時間增長，可能同時存在多種機轉與多個快速 放電區域影響，使心房電訊號轉變為複雜碎裂形態，因其非線性、非穩態特性分析難度大增，致使此 類病患之導管電氣燒灼術治療成功率偏低，仍有技術瓶頸亟待突破。多導極型導管可同時記錄來自多 個位置的數據，並且因此可同時提供空間及時間資訊。此種資訊有助於表現出心房基質之電氣特性。 本計畫利用非接觸型三維立體定位系統與多導級型導管並結合複雜系統與統計物理理論，開發出相似 度指標與轉子源即時辨識等技術，以旋度（曲率力），散度來描述平均的波傳播特性並結合熵函數梯 度場證實「轉子源（Rotor）」的存在。此外，藉由異構計算結合不同平台和處理器的優勢以達到計算 效能的最佳化，達成以多導級型導管即時三維造影之目標。最終更可望取代現行的複雜碎裂心內訊號 分析，成為目前立體定位及訊號分析整合系統不可或缺的功能。 ;Atrial fibrillation (AF) is the most common cardiac arrhythmia in clinical practice. Due to its being minimally invasive, safe, and relatively easy to perform with few complications, the procedure of catheter ablation has become a prominent choice to terminate AF in addition to medication. The procedure traditionally involves in isolation of the pulmonary veins (PVs) and is the mainstream non-pharmacologic therapy for AF. While PV isolation commonly performed in clinical practice around the world, the success rate of this procedure decreases significantly for patients with persistent AF (AF consistently occurs for more than 7 days) and yet, no satisfactory improvement has been made by the physicians and researchers. A plausible reason is as the atrial fibrillation persists, wide-distributed rapid discharging areas in combination of multiple new evolving mechanisms might all contribute to maintain the presence of atrial fibrillation. In this case, the atrial electrogram will exhibit quite complicated and irregular patterns, termed complex fractionated atrial electrogram (CFAE). The CFAE eletrograms features non-stationary and nonlinear behaviors diminishing frequency gradient (no distinct dominant frequency) and make frequency domain analysis significantly more difficult to localize the ablation site. The multiple lead catheters can simultaneously record data from different locations, and thus can provide both spatial and temporal information that help to reveal the electrical properties of the atrial matrix. Recent study documented the importance of rotors in maintaining persistent AF. Identifying the pivot of rotor was well accepted to present the phase singularity (PS). However as the atrial fibrillation progressed from paroxysmal to persistent, the areas with CFAE electrograms occupied by high density of irregular wavebreaks can randomly cause the PS. Those issues will highly compromise the capability of phase mapping to identify rotors. In this project, we aim to combine the theory of complex systems and statistical physics with a non-contact 3 dimensional guiding system to demonstrate the existence of “rotors” as one of the main causes of AF and to develop analytic tools such as the similarity index and recurrence plot to identify the AF sources. We use physical concepts such as the divergence and the curl (rotating force) of a vector field to describe the characteristics of the average wave propagation and also the gradient field of an entropy function to establish the existence and effects of the rotor sources. In addition, through a heterogeneous computing structure we can incorporate the advantages of various platforms and processors to achieve optimal computing efficiency. It is expected that our procedures can be translated into an essential algorithm for AF ablation in the latest 3 dimensional guiding system.
