Diffuse　correlation　spectroscopy　(DCS)　derives　blood　flow　index　(BFI)　by　measuring　the　temporal　intensity　fluctuations　of　multiply　scattered　light.　Blood　flow　index　(BFI)　and　especially　its　variations　were　demonstrated　to　be　approximately　proportional　to　absolute　blood　flow.　In　this　paper,　we　have　proposed　a　predictive　method　for　calculating　blood　flow　as　well　as　oxygen　saturation,　based　on　a　deep　neural　network　of　long　short-term　memory　(LSTM)　architecture.　The　simulated　multiwavelength　normalized　intensity　autocorrelation　function　data　for　various　blood　flows　and　oxygen　saturations　were　used　to　train　the　LSTM　architecture.　The　results　validated　the　feasibility　of　the　proposed　method　for　quantification　of　blood　flow　and　oxygen　saturation　simultaneously　in　DCS.　The　proposed　approach　would　be　an　alternative　method　for　oxygen　metabolism　monitoring.