Computer aided diagnosis and monitoring of heart failure patients
AssigneeKoninklijke Philips N.V.
InventorsMarija Despenic, Michael Alex Van Hartskamp
Life of this patent
Estimating the term…
Abstract
A system and computer program are provided for computer aided diagnosis or monitoring of heart failure patients. The system and computer program obtain patient data representing measurements of a select and limited set of health indicators of a patient. A trained classifier is applied to the patient data to obtain a classification of the occurrence or degree of heart failure for the patient. The limited set of health indicators is specifically selected from a large set of health indicators which are known to be predictive of heart failure on the basis of a clinical study which demonstrated the accuracy of the classification using this select set.
Claims18
A system for computer aided diagnosis or monitoring of heart failure patients, the system comprising:
- a memory comprising instruction data representing a set of instructions, and classifier data representing a trained classifier trained on training data in a patient population;
- a processor configured to communicate with the memory and to execute the set of instructions, wherein the set of instructions, when executed by the processor, cause the processor to:
- obtain a select set of health indicators from a large set of health indicators that are predictive of heart failure, the select set of health indicators consisting of fatigue level, orthopnea, dyspnea, and oxygen saturation;
- obtain patient data representing measurements of the select set of health indicators from a patient, so as to obtain patient fatigue level data, patient orthopnea data, patient dyspnea data and patient oxygen saturation data;
- provide a classification of an occurrence or degree of heart failure for the patient using the select set of health indicators and the trained classifier, wherein the measurements of the select set of health indicators from the patient are obtained in an outpatient setting, without any involvement or oversight by a medical expert or technician,
- retrieve the trained classifier from the memory, wherein each sample of the training data represents the measurements of the same select set of health indicators, consisting of fatigue level, orthopnea, dyspnea and oxygen saturation, of at least one studied patient in the patient population and has a label indicating an occurrence or degree of heart failure of the at least one studied patient, the training data including population fatigue level data, population orthopnea data, population dyspnea data, and population oxygen saturation data;
- wherein the patient data is different from the training data on which the trained classifier is trained,
- apply the trained classifier to the patient data, including the patient fatigue level data, the patient orthopnea data, the patient dyspnea data, and the patient oxygen saturation data, to obtain the classification of the occurrence or the degree of heart failure for the patient;
- and
- generate output data representing the classification of the occurrence or the degree of heart failure for the patient.
A non-transitory computer-readable medium storing computer readable instructions which, when executed by a processor system, cause the processor system to carry out operations for a computer aided diagnosis or monitoring of heart failure patients by:
- obtaining a select set of health indicators from a large set of health indicators that are predictive of heart failure, the select set of health indicators consisting of fatigue level, orthopnea, dyspnea, and oxygen saturation;
- accessing patient data representing measurements of the select set of health indicators, including fatigue level, orthopnea, dyspnea and oxygen saturation, from a patient, so as to obtain patient fatigue level data, patient orthopnea data, patient dyspnea data and patient oxygen saturation data;
- providing a classification of an occurrence or degree of heart failure for the patient using a trained classifier trained on training data in a patient population, wherein the measurements of the select set of health indicators, including the patient fatigue level data, the patient orthopnea data, the patient dyspnea data, and the patient oxygen saturation data, from the patient are obtained in an outpatient setting, without any involvement or oversight by a medical expert or technician,
- accessing the trained classifier, wherein each sample of the training data represents the measurements of the same select set of health indicators, consisting of fatigue level, orthopnea, dyspnea and oxygen saturation, of at least one studied patient in the patient population and has a label indicating an occurrence or degree of heart failure of the at least one studied patient, the training data including population fatigue level data, population orthopnea data, population dyspnea data, and population oxygen saturation data;
- wherein the patient data is different from the training data on which the trained classifier is trained;
- applying the trained classifier to the patient data, including the patient fatigue level data, the patient orthopnea data, the patient dyspnea data, and the patient oxygen saturation data, to obtain the classification of the occurrence or the degree of heart failure for the patient;
- and
- generating output data representing the classification of the occurrence or the degree of heart failure for the patient.
Dependencies are inferred from the claim text.