Electroencephalogram(EEG)　signals　can　reflect　the　electronic　activities　of　the　brain　and　are　commonly　used　for　disease　diagnosis.　For　this　reason,　as　part　of　the　George　B.　Moody　PhysioNet　Challenge　2023,　we　BJUT-bme　team　developed　a　deep　learning　model　for　automated　analysis　of　EEG　signals　to　predict　patients　＇　clinical　outcomes　after　12,　24,　48,　and　72　hours　following　cardiac　arrest.　This　model　used　a　convolutional　neural　network　with　R(2+1)D　module　and　extracted　the　power　spectrum　features　of　EEG　and　brain　functional　connectivity　features　as　inputs　to　the　model,　then　output　prediction　outcomes.　We　got　scores　of　0.602,　0.522,　and　0.406　separately　on　the　training　set,　validation　set,　and　test　set.　But　our　team　was　unable　to　be　ranked.　©　2023　CinC.