以雲端平台結合IoT裝置實現心房顫動的大規模篩檢

Cloud-based IoT devices for large-scale atrial fibrillation screening

曾繼禾(Chi-Ho Tseng)

指導教授 ： 林致廷

共同指導教授 ： 羅孟宗(Men-Tzung Lo)

國立臺灣大學/電機資訊學院/生醫電子與資訊學研究所/博士(2020年)

https://doi.org/10.6342/NTU202002978

摘要

心房顫動是一種常見的心律不整疾病，患有心房顫動的患者其中風的風險比一般人要高五倍，其在65歲以上人口的患病率則約為10％。然而心房顫動的發生可能是偶發性的，並且有時並無伴隨相關症狀，導致心房顫動難以在早期被診斷出來。而近期的幾項研究採用了新型的微型化非連續式心電圖量測儀器，在普查式篩檢和家庭用量測方式都可以顯著的提高檢測率。在本研究中為了開發有效且可持續性的策略來檢測尚未診斷到的心房顫動患者，我們提出了一種結合物聯網裝置與雲端基礎架構的系統用於心律不整篩檢，尤其是針對心房顫動。與行動裝置結合後，可以不受環境限制隨時隨地在不同使用情境下連接並使用該系統。例如在健康檢查中即時反饋篩檢結果或是用於7-14天的家用心電圖篩檢。本系統於2018年至2019年間與桃園市衛生局、臺北醫學大學附設醫院和新竹國泰醫院等單位合作用於心律不整篩檢超過2000多人，檢出率為13.4％。並於2019年後與陸續與台北市、桃園市、嘉義縣、台南市、屏東縣等縣市衛生單位合作展開大規模心房顫動篩檢，至目前為止已完成超過8729位無心房顫動病史的民眾篩檢，其中65歲以上的心房顫動檢出率為2.6%。

關鍵字

心房顫動 ； 物聯網 ； 雲端服務 ； 醫療資訊系統 ； 健康照護

並列摘要

Atrial fibrillation (AFib) is the most common arrhythmia, and the patients with AFib have five times higher risk for stroke. The prevalence is around 10% for the population over 65 years old. However, the occurrence of AFib can be episodic and sometimes asymptomatic which leads to underdiagnoses of AFib. New miniaturized intermittent ECG devices were adopted in several studies, and both population screening and home-based monitoring can significantly increase the detection rate. To develop an efficient and sustainable strategy for detecting undiagnosed AFib, we propose a cloud-based AI system for arrhythmia screening, especially for AFib. This system can be ubiquitously connected by incorporating with mobile devices in different scenarios such as health examination with real-time feedback or home-based monitor for the patients with 7-14 days screening. The system has been applied for Arrhythmia screening. During from 2018 to 2019, we cooperate with Department of Public Health, Taoyuan, Taipei Medical University Hospital, and Hsinchu Cathay General hospital, over 2000 people had been screened with 13.4% arrhythmia detection rate. Starting in 2019, we continue cooperate with health institutions of Taipei, Chiayi, Tainan, and Pingtung on large-scale AFib screening. So far, over 8729 people had been screened and 2.6% of AFib detection rate over 65.

並列關鍵字

atrial fibrilliation ； IoT ； cloud services ； medical information system ； health care

參考文獻

[1] Siontis, K.C., et al., “Typical, atypical, and asymptomatic presentations of new-onset atrial fibrillation in the community: Characteristics and prognostic implications,” Heart Rhythm, 2016. 13(7): p. 1418-24. doi: 10.1016/j.hrthm.2016.03.003.

[2] Bleifer, S.B., et al., “Diagnosis of occult arrhythmias by Holter electrocardiography,” Progress in Cardiovascular Diseases, 1974. 16(6): p. 569-599. doi: https://doi.org/10.1016/0033-0620(74)90019-X.

[3] Kuchar, D.L., C.W. Thorburn, and N.L. Sammel, “Prediction of serious arrhythmic events after myocardial infarction: Signal-averaged electrocardiogram, holter monitoring and radionuclide ventriculography,” Journal of the American College of Cardiology, 1987. 9(3): p. 531. doi: 10.1016/S0735-1097(87)80045-1.

[4] Lipski, J., et al., “Value of holter monitoring in assessing cardiac arrhythmias in symptomatic patients,” The American Journal of Cardiology, 1976. 37(1): p. 102-107. doi: https://doi.org/10.1016/0002-9149(76)90507-5.

[5] Rizos, T., et al., “Continuous Stroke Unit Electrocardiographic Monitoring Versus 24-Hour Holter Electrocardiography for Detection of Paroxysmal Atrial Fibrillation After Stroke,” Stroke, 2012. 43(10): p. 2689-2694. doi: doi:10.1161/STROKEAHA.112.654954.