Deep Learning Model for Predicting Clinical Service Line Financial Performance Using Case Mix, Resource Consumption, Length of Stay, Payer Contracts, Procedure Volume, and Operational Throughput Measures