Multimodal tensor-based method for integrative and continuous patient monitoring during postoperative cardiac care

Larry Hernandez, Renaid B. Kim, Neriman Tokcan, Harm Derksen, Ben E. Biesterveld, Alfred Croteau, Aaron M. Williams, Michael Mathis, Kayvan Najarian, Jonathan Gryak

Abstract: Highlights•Effective postoperative care requires the synthesis of multiple data modalities.•Advanced techniques can extract complex patterns from physiological signals.•Tensor-based methods reduce model inputs while maintaining/increasing performance.•A multimodal approach can predict hemodynamic decompensation several hours in advance.