Technology Introduction:
We developed a Deep Heterogeneous Multimodal Learning-Based Technique for Predicting Hospital Readmission and Mortality Risk in heart failure patients. This technique integrates heterogeneous modalities, including clinical data, electrocardiograms, and chest X-rays, to predict short-term and long-term risks for mortality and readmission. The relevant clinical application of this technique has been recognized by the 20th National Innovation Award and under validation in multiple hospitals.
Industry Applicability:
Heart failure prevalence has been rising in increased aged populations and become one of the main factors of hospitalizations for patients aged 65 and older. Due to high readmission and mortality rates, accurate risk prediction tools are under high demands. Our technique fuses heterogeneous modalities to predict short-term and long-term risks for readmission and mortality in heart failure patients, carring high application values in smart medicine areas, including effective clinical decision support, cost reduction and improved outcomes on chronic disease care.