Abstract: Various methods and systems are provided for predicting a discharge date of a patient of a healthcare facility. In one embodiment, a method for a patient management system of a healthcare facility comprises receiving a selected confidence level from a user of the patient management system; predicting a date for discharging a patient of the healthcare facility using a trained discharge date prediction model, based on a set of patient data of the patient, the discharge date prediction model trained on historical patient data of the healthcare facility; based on the received confidence level and an output of the trained discharge date prediction model, predict a discharge date window of the patient; generating a predicted discharge date window element summarizing the discharge window prediction in a user interface (UI) of the patient management system; and displaying the UI on a display device of the patient management system.

Abstract: Techniques are described for predicting information regarding expected time of occurrence of clinical events based on longitudinal patient data. According to an embodiment, a method can include clustering, by a system comprising a processor, training data samples corresponding to different patients that experienced a clinical event into different patient groups as a function of different defined timeframes within which the clinical event occurred. The method further comprises employing, by the system, a first machine learning process to train a classification model using the training data samples to predict the different patient groups to which the training data samples respectively belong, and employing, by the system, a second machine learning process to train a clinical time to event model using the training data samples to predict an expected duration of time until occurrence of the clinical event as a function of the different patient groups.

Abstract: Techniques are described for managing tasks of a dynamic system with limited resources using a machine-learning and combinatorial optimization framework. In one embodiment, a computer-implemented method is provided that comprises employing, by a system operatively coupled to a processor, one or more first machine learning models to determine a total demand for tasks of a dynamic system within a defined time frame based on state information regarding a current state of the dynamic system, wherein the state information comprises task information regarding currently pending tasks of the tasks. The method further comprises, employing, by the system, one or more second machine learning models to determine turnaround times for completing the tasks based on the state information, and determining, by the system, a prioritization order for performing the currently pending tasks based on the total demand and the turnaround times.

Abstract: Techniques are described for identifying complex patients and forecasting patient outcomes based on a variety of factors including medical, socio-economic, mental and behavioral. According to an embodiment, a method can include employing one or more machine learning models to identify complex patients and predict patient outcomes like length of stay, potential discharge trajectories with likelihoods, discharge destinations, readmission likelihood and safety. These models are applied to respective patients that are currently admitted to a hospital and expected to be placed after discharge from the hospital, wherein the one or more discharge forecasting machine learning models predict the discharge destinations based on clinical data points and non-clinical data points collected for the respective patients.

Abstract: Systems and techniques for monitoring, predicting and/or alerting for census periods in medical inpatient units are presented. A system can include a grouping component that defines a group of beds at a medical facility based on at least one grouping factor, and a group stability component that determines a measure of occupancy variability for the group based on historical census data for respective beds in the group. The system can further include a model selection component that selects one or more census forecasting models for the group based on the measure of occupancy variability, and a patient census component that applies the one or more census forecasting models to current patient flow data for the medical facility to forecast an expected occupancy level for the group during one or more future periods of time.

Abstract: Techniques are described for optimizing patient placement and sequencing in dynamic medical environment. In various embodiments, a method includes receiving current state information regarding a current state of a medical facility system in real-time, including operating conditions data regarding current operating conditions of the medical facility system and patient case data regarding active patient cases and pending patient cases of the medical facility system. The method further includes forecasting future state information for the medical facility system based on the current state information using a machine learning framework, including forecasted timeline information regarding future timing of workflow events of the active patient cases and pending patient cases.

Abstract: Systems and techniques for monitoring, predicting and/or alerting for census periods in medical inpatient units are presented. A system can perform a first machine learning process to learn patterns in patient flow data related to a set of patient identities and a set of operations associated with a set of medical inpatient units. The system can also perform a second machine learning process to detect abnormalities associated with the patterns in the patient flow data. Furthermore, the system can determine patient census data associated with a prediction for a total number of patient identities in the set of medical inpatient units during a period of time based on the patterns and the abnormalities. The system can also generate an alert for a user interface in response to a determination that the patient census data satisfies a defined criterion.

Abstract: Techniques are described for identifying complex patients and forecasting patient outcomes based on a variety of factors including medical, socio-economic, mental and behavioral. According to an embodiment, a method can include employing one or more machine learning models to identify complex patients and predict patient outcomes like length of stay, potential discharge trajectories with likelihoods, discharge destinations, readmission likelihood and safety. These models are applied to respective patients that are currently admitted to a hospital and expected to be placed after discharge from the hospital, wherein the one or more discharge forecasting machine learning models predict the discharge destinations based on clinical data points and non-clinical data points collected for the respective patients.

Abstract: Techniques are described for managing tasks of a dynamic system with limited resources using a machine-learning and combinatorial optimization framework. In one embodiment, a computer-implemented method is provided that comprises employing, by a system operatively coupled to a processor, one or more first machine learning models to determine a total demand for tasks of a dynamic system within a defined time frame based on state information regarding a current state of the dynamic system, wherein the state information comprises task information regarding currently pending tasks of the tasks. The method further comprises, employing, by the system, one or more second machine learning models to determine turnaround times for completing the tasks based on the state information, and determining, by the system, a prioritization order for performing the currently pending tasks based on the total demand and the turnaround times.

Abstract: Systems and techniques for monitoring, predicting and/or alerting for census periods in medical inpatient units are presented. A system can perform a first machine learning process to learn patterns in patient flow data related to a set of patient identities and a set of operations associated with a set of medical inpatient units. The system can also perform a second machine learning process to detect abnormalities associated with the patterns in the patient flow data. Furthermore, the system can determine patient census data associated with a prediction for a total number of patient identities in the set of medical inpatient units during a period of time based on the patterns and the abnormalities. The system can also generate an alert for a user interface in response to a determination that the patient census data satisfies a defined criterion.

Abstract: Techniques are described that employ a multifactorial, machine-learning based system and prioritization framework for optimizing patient placement to beds at a medical facility. In one embodiment, a computer-implemented is provided that comprises receiving, by a system operatively coupled to a processor, a patient placement request requesting placement of a patient to a hospital bed of the healthcare facility, wherein the request is associated with information identifying a medical service for the patient and a bed type. The method further comprises, selecting, by the system, a placement prioritization model from a set of placement prioritization models based on the medical service and the bed type, and employing, by the system, the prioritization model and state information regarding a current state of the healthcare facility to determine a prioritization score reflective of a priority level of the patient placement request.

Abstract: A method and system for controlling demand events in a utility network with multiple customer sites. The value of a demand response parameter threshold for invoking a demand response event is calculated based on the number of available demand response events and the number of opportunities remaining to issue the available demand response events. This parameter represents the utility objectives for using the demand response program (e.g., cost savings, reliability, avoided costs). A current value of the demand response parameter is compared to the threshold value, and a determination is made whether or not to call a demand response event for the current opportunity, or to save the event for a future opportunity based upon this comparison.

Abstract: A system and method of using one or more DC-DC/DC-AC converters and/or alternative devices allows strings of multiple module technologies to coexist within the same PV power plant. A computing (optimization) framework estimates the percentage allocation of PV power plant capacity to selected PV module technologies. The framework and its supporting components considers irradiation, temperature, spectral profiles, cost and other practical constraints to achieve the lowest levelized cost of electricity, maximum output and minimum system cost. The system and method can function using any device enabling distributed maximum power point tracking at the module, string or combiner level.

Abstract: Systems and methods for determining how to stack containers/trailers on a vehicle consist at a terminal/yard at least to maintain aerodynamic efficiency. Embodiments of the present invention provide a terminal management software application configured to determine how to stack containers/trailers on vehicles of a vehicle consist, taking into account a resultant aerodynamic efficiency of the vehicle consist during transit as well as other factors.

Abstract: Snapshot data may be received indicative of a current state of resources that deliver healthcare to a plurality of patients associated with a healthcare enterprise. The received snapshot data may be automatically used to initialize a healthcare enterprise simulation model. The healthcare enterprise simulation model may then be executed to automatically generate a predicted future state of the resources at a predetermined point in time. Some embodiments may automatically suggest mitigation strategies based on simulated scenarios reflecting potential bottlenecks and appropriate actions that may be taken by the enterprise. The system may continuously and automatically monitor forecast accuracy to detect potential anomalies.

Abstract: According to some embodiments, first current resource data indicative of a current state of first resources that are used to deliver healthcare to a plurality of patients associated with a first healthcare enterprise is continuously and automatically received. The first current resource data may be automatically used to update a healthcare enterprise simulation model. The healthcare enterprise simulation model may automatically generate a predicted future state of the first resources, wherein the predicted future state of the first resources is based at least in part on second resource data indicative of a state of second resources that are used to deliver healthcare to a plurality of patients associated with a second healthcare enterprise remote from and networked with the first healthcare enterprise.

Abstract: Current resource data may be continuously and automatically received indicative of a current state of resources that are used to deliver healthcare to a plurality of patients associated with a healthcare enterprise. The current resource data may be automatically used to update a healthcare enterprise simulation model. The healthcare enterprise simulation model may be executed to automatically generate a predicted future state of the resources. A resource request may then be received, and a resource assignment engine may automatically assign a particular resource to the resource request based at least in part on the predicted future state of the resources.

Abstract: Systems and methods for determining how to stack containers/trailers on a vehicle consist at a terminal/yard at least to maintain aerodynamic efficiency. Embodiments of the present invention provide a terminal management software application configured to determine how to stack containers/trailers on vehicles of a vehicle consist, taking into account a resultant aerodynamic efficiency of the vehicle consist during transit as well as other factors.

Abstract: A system and method of using one or more DC-DC/DC-AC converters and/or alternative devices allows strings of multiple module technologies to coexist within the same PV power plant. A computing (optimization) framework estimates the percentage allocation of PV power plant capacity to selected PV module technologies. The framework and its supporting components considers irradiation, temperature, spectral profiles, cost and other practical constraints to achieve the lowest levelized cost of electricity, maximum output and minimum system cost. The system and method can function using any device enabling distributed maximum power point tracking at the module, string or combiner level.

Abstract: A method and system for controlling demand events in a utility network with multiple customer sites. The value of a demand response parameter threshold for invoking a demand response event is calculated based on the number of available demand response events and the number of opportunities remaining to issue the available demand response events. This parameter represents the utility objectives for using the demand response program (e.g., cost savings, reliability, avoided costs). A current value of the demand response parameter is compared to the threshold value, and a determination is made whether or not to call a demand response event for the current opportunity, or to save the event for a future opportunity based upon this comparison.