Abstract: Systems/techniques that facilitate smart training and smart deployment of machine learning models are provided. In various embodiments, a system can access a first set of data candidates that are available for training of a machine learning model. In various aspects, the system can compute at least one feature distribution of the first set of data candidates. In various instances, the system can identify, in the first set of data candidates, a strict subset of data candidates, wherein at least one feature distribution of the strict subset of data candidates matches the at least one feature distribution of the first set of data candidates. In various cases, the system can train the machine learning model on the strict subset of data candidates.

Abstract: Techniques are described for domain adaptation of image processing models using post-processing model correction According to an embodiment, a method comprises training, by a system operatively coupled to a processor, a post-processing model to correct an image-based inference output of a source image processing model that results from application of the source image processing model to a target image from a target domain that differs from a source domain, wherein the source image processing model was trained on source images from the source domain. In one or more implementations, the source imaging processing model comprises an organ segmentation model and the post-processing model can comprise a shape-autoencoder. The method further comprises applying, by the system, the source image processing model and the post-processing model to target images from the target domain to generate optimized image-based inference outputs for the target images.

Abstract: Systems and methods are provided for perioperative care in a medical facility. In an example, a system includes a display and a computing device operably coupled to the display and storing instructions executable to output, to the display, a graphical user interface (GUI) that includes real-time medical device data of a patient, at least some of the real-time medical device data displayed via the GUI as a plurality of patient monitoring parameter tiles, the GUI including a risk score indicative of a relative likelihood that the patient will exhibit a condition within a period of time, and responsive to a user input, display, on the GUI, a set of trend lines each showing values for a respective patient monitoring parameter over a time range, each trend line of the set of trend lines selected based on a contribution of each respective patient monitoring parameter to the risk score.

Abstract: Systems/techniques that facilitate learning-based domain transformation for medical images are provided. In various embodiments, a system can access a medical image. In various aspects, the medical image can depict an anatomical structure according to a first medical scanning domain. In various instances, the system can generate, via execution of a machine learning model, a predicted image based on the medical image. In various aspects, the predicted image can depict the anatomical structure according to a second medical scanning domain that is different from the first medical scanning domain. In some cases, the first and second medical scanning domains can be first and second energy levels of a computed tomography (CT) scanning modality. In other cases, the first and second medical scanning domains can be first and second contrast phases of the CT scanning modality.

Abstract: This application provides methods and systems for automated condition-based maintenance of mechanical systems. Example systems may at least one memory coupled to one or more computer processors that are configured to receive first data from the mechanical system indicative of performance of a first component of the mechanical system, determine, using the first data, a first performance metric for the first component, determine, using the first performance metric, a probability value that a fault has occurred at the first component, and determine, using the probability value, a predicted length of time until failure of the first component.

Abstract: Techniques are described for domain adaptation of image processing models using post-processing model correction According to an embodiment, a method comprises training, by a system operatively coupled to a processor, a post-processing model to correct an image-based inference output of a source image processing model that results from application of the source image processing model to a target image from a target domain that differs from a source domain, wherein the source image processing model was trained on source images from the source domain. In one or more implementations, the source imaging processing model comprises an organ segmentation model and the post-processing model can comprise a shape-autoencoder. The method further comprises applying, by the system, the source image processing model and the post-processing model to target images from the target domain to generate optimized image-based inference outputs for the target images.

Abstract: A method for predicting an anomaly in a combustor (16) is presented. The method includes receiving signals representative of parameters in one or more combustion cans (22, 24) of the combustor, generating a plurality of patterns based on a permutation entropy window and the signals, identifying a plurality of pattern categories in the plurality of patterns, determining a permutation entropy based on the plurality of patterns and the plurality of pattern categories, and predicting an anomaly in the combustor based on the permutation entropy. The method further includes comparing the plurality of pattern categories to determined permutations of pattern categories if the anomaly is present in the combustor, and predicting a category of the anomaly based on the comparison of the plurality of pattern categories to the determined permutations of pattern categories.

Abstract: Systems and methods are provided for perioperative care in a medical facility. In an example, a system includes a display and a computing device operably coupled to the display and storing instructions executable to output, to the display, a graphical user interface (GUI) that includes real-time medical device data of a patient, at least some of the real-time medical device data displayed via the GUI as a plurality of patient monitoring parameter tiles, the GUI including a risk score indicative of a relative likelihood that the patient will exhibit a condition within a period of time, and responsive to a user input, display, on the GUI, a set of trend lines each showing values for a respective patient monitoring parameter over a time range, each trend line of the set of trend lines selected based on a contribution of each respective patient monitoring parameter to the risk score.

Abstract: A monitoring system of a combined cycle power plant having a topping cycle performed by a gas turbine and a bottoming cycle performed by a steam turbine, the monitoring system. The monitoring system includes a processor that receives a first set of data indicating a first measurement value of an operating parameter of the combined cycle power plant. The processor determines a baseline efficiency of the process based on the first measurement value using a bottoming cycle performance (BCP) model. The processor determines a threshold based on the baseline efficiency. The processor receives a second set of data indicating a second measurement value of the operating parameter. The processor compare the threshold to an operational efficiency based on the second measurement value using the BCP model, and provides an indication of a decrease in efficiency of the process if the operational efficiency is below the threshold.

Abstract: A non-transitory computer readable medium stores instructions that are executed by a processor of a monitoring that when executed, the processor receives a state of at least one operational parameter of the turbomachinery. Then, the processor determines whether the state is within a region of accuracy of a stored record of a stored state stored in the memory. When the state is within the region of accuracy, the processor determines at least one operating condition from the stored record. Alternatively, when the state is not within the first region of accuracy, the processor stores the state as a new record in the memory.

Abstract: This application provides methods and systems for automated condition-based maintenance of mechanical systems. Example systems may at least one memory coupled to one or more computer processors that are configured to receive first data from the mechanical system indicative of performance of a first component of the mechanical system, determine, using the first data, a first performance metric for the first component, determine, using the first performance metric, a probability value that a fault has occurred at the first component, and determine, using the probability value, a predicted length of time until failure of the first component.

Abstract: A non-transitory computer-readable storage medium storing one or more processor-executable instructions wherein the one or more instructions, when executed by a processor of a controller, cause acts to be performed including receiving signals representative of pressure between respective compressor blade tips and a casing of a compressor at one or more stages, generating multiple patterns based on a permutation entropy window and the signals, identifying multiple pattern categories in the multiple patterns, determining a permutation entropy based on the multiple patterns and the multiple pattern categories, predicting an anomaly in the compressor based on the permutation entropy, comparing the multiple pattern categories to determined permutations of pattern categories when an anomaly is present in the compressor, and predicting a category of the anomaly based on the comparison of the multiple pattern categories to the determined permutation of pattern categories.

Abstract: A fault detection method for a selective catalytic reduction (SCR) system comprising an SCR reactor, includes receiving a plurality of operating parameters (702) of the SCR reactor from a plurality of sensors. The method also includes estimating a state of an adaptive reactor model (704) representative of the SCR reactor based on the plurality of operating parameters. The method also includes generating a feature parameter (706) based on the plurality of operating parameters and the estimated state of the adaptive reactor model. The method includes determining a fault in the SCR system (708) based on the feature parameter.

Abstract: A system includes a controller coupled to an inlet bleed heat system of a compressor. The controller is configured to detect an icing condition of an inlet portion of the compressor by determining a current compressor efficiency value of the compressor, a current compressor flow value of the compressor, or any combination thereof, comparing the current compressor efficiency value to a compressor efficiency model of the compressor, the current compressor flow value to a compressor flow model of the compressor, or any combination thereof, and providing an icing indication to the inlet bleed heat system if a first difference between the current compressor efficiency value and the compressor efficiency model is greater than a first threshold, a second difference between the current compressor flow value and the compressor flow model is greater than a second threshold, or any combination thereof.

Abstract: A fluid transport system includes at least one flow control device and a multiphase pump configured to transport fluid. At least one pump sensing device is configured to measure at least one operating characteristic of the multiphase pump. A controller is programmed with a pump map including a correlation of the at least one operating characteristic of the multiphase pump with at least one operating characteristic of the fluid. The controller is configured to determine an estimated value of the at least one operating characteristic of the fluid based on the measured value of the at least one operating characteristic of the multiphase pump and the pump map. At least one regulating device coupled to at least one flow control device is modulated based on the estimated value of the at least one operating characteristic of the fluid.

Abstract: A method for predicting an anomaly in a combustor (16) is presented. The method includes receiving signals representative of parameters in one or more combustion cans (22, 24) of the combustor, generating a plurality of patterns based on a permutation entropy window and the signals, identifying a plurality of pattern categories in the plurality of patterns, determining a permutation entropy based on the plurality of patterns and the plurality of pattern categories, and predicting an anomaly in the combustor based on the permutation entropy. The method further includes comparing the plurality of pattern categories to determined permutations of pattern categories if the anomaly is present in the combustor, and predicting a category of the anomaly based on the comparison of the plurality of pattern categories to the determined permutations of pattern categories.

Abstract: A system includes a controller coupled to an inlet bleed heat system of a compressor. The controller is configured to detect an icing condition of an inlet portion of the compressor by determining a current compressor efficiency value of the compressor, a current compressor flow value of the compressor, or any combination thereof, comparing the current compressor efficiency value to a compressor efficiency model of the compressor, the current compressor flow value to a compressor flow model of the compressor, or any combination thereof, and providing an icing indication to the inlet bleed heat system if a first difference between the current compressor efficiency value and the compressor efficiency model is greater than a first threshold, a second difference between the current compressor flow value and the compressor flow model is greater than a second threshold, or any combination thereof.

Abstract: A non-transitory computer-readable storage medium storing one or more processor-executable instructions wherein the one or more instructions, when executed by a processor of a controller, cause acts to be performed including receiving signals representative of pressure between respective compressor blade tips and a casing of a compressor at one or more stages, generating multiple patterns based on a permutation entropy window and the signals, identifying multiple pattern categories in the multiple patterns, determining a permutation entropy based on the multiple patterns and the multiple pattern categories, predicting an anomaly in the compressor based on the permutation entropy, comparing the multiple pattern categories to determined permutations of pattern categories when an anomaly is present in the compressor, and predicting a category of the anomaly based on the comparison of the multiple pattern categories to the determined permutation of pattern categories.

Abstract: A non-transitory computer readable medium stores instructions that are executed by a processor of a monitoring that when executed, the processor receives a state of at least one operational parameter of the turbomachinery. Then, the processor determines whether the state is within a region of accuracy of a stored record of a stored state stored in the memory. When the state is within the region of accuracy, the processor determines at least one operating condition from the stored record. Alternatively, when the state is not within the first region of accuracy, the processor stores the state as a new record in the memory.

Abstract: A monitoring system of a combined cycle power plant having a topping cycle performed by a gas turbine and a bottoming cycle performed by a steam turbine, the monitoring system. The monitoring system includes a processor that receives a first set of data indicating a first measurement value of an operating parameter of the combined cycle power plant. The processor determines a baseline efficiency of the process based on the first measurement value using a bottoming cycle performance (BCP) model. The processor determines a threshold based on the baseline efficiency. The processor receives a second set of data indicating a second measurement value of the operating parameter. The processor compare the threshold to an operational efficiency based on the second measurement value using the BCP model, and provides an indication of a decrease in efficiency of the process if the operational efficiency is below the threshold.