Identification of Atrial Fibrillation Using Complex Network Similarity

Ide ABSTRAC We investigat network is est dium signals r larity. Finally The method is sitivity, specif cates that our 1. Introduc Atrial fibrillat menon in clin tivities take pl heart rhythm. heart and ma will induce hi important to more serious. With the d complex netw analyze the dy the real world complex syste lyzed by using In [9], the the two-dimen culated as the diac arrhythm racy of the de spectively. Be thod needs to As we have electrical activ will become w ity between t calculated to CT te the use of c timated via the recorded from , epicardium s s validated usi ficity and accu approach may Recurrence Com Atrial Fibrillatio ction tion (AF) is a nic [1,2]. Duri lace in the hea AF deteriorat ay cause embo igh rates of mo detect AF an development work [3-8] has ynamical prop d. Because the em, the heart e g complex netw distribution en nsional recons e feature of ele mia and the sen etection were ecause of the be proposed. e known, for vities, the sim weaker, so feat two recurrenc identify AF. T complex netwo e joint recurren m dogs into the signals are cla ing 1000 samp uracy of the id y lay a foundati mplex Network on a common arrh ing AF, chaot art which resul tes the functio olic events an orbidity and m nd prevent it of nonlinear become a pop perties of comp e heart has be electrical activ works. ntropy of the p structed phase ectrocardiogram nsitivity, speci 86.0%, 96.0% low sensitivit complex netw milarity betwee tures represent ce complex ne This will be p Published Online f Atrial Network ajuan Zhang atory of ASIC an Electronic Engine 210720090@fu Received ork similarity nce plot and H recurrence co assified utilizin ples including dentification ar ion for the pre k; Phase Space hythmia pheno ic electrical ac lt in an irregul on of the who nd stroke whic mortality. So it from becomin dynamics, th pular method t plex systems i een proven to vity can be an point density i space was ca m to detect ca ficity and accu % and 92.7% r ty, a novel m works of chaot n two network ting the simila etworks can b proven with th for the identif Hamming dista omplex networ ng the classific 500 atrial fibri re 98.2%, …