Hypothesis-driven, real-time analysis of physiological data streams using textual representations

Abstract

A method of analyzing physiological data streams. According to the method, physiological data is received into a computerized machine. The physiological data comprises numerical data and medical symptoms of a patient. Features are extracted from the physiological data based on development of the physiological data over a period of time. The features are converted into a textual representation using natural language generation. Input terms for an information retrieval system operating on the computerized machine are automatically generated based on the features. The input terms are input to the information retrieval system. A corpus of data is automatically searched to retrieve results to the input terms using the information retrieval system.

What is claimed is: 
     
         1 . A computer system for exploration of time series physiological data streams comprising:
 an input/output port receiving physiological data, said physiological data comprising numerical data and medical symptoms of a patient;   a processor operatively connected to said input/output port, said processor:
 automatically extracting features from said physiological data based on development of said physiological data over a period of time, 
 automatically converting said features into a textual representation based on natural language generation, 
 automatically generating input terms for an information retrieval system based on said textual representation, 
 inputting said input terms to said information retrieval system, and 
   automatically searching a corpus of data to retrieve results to said input terms using said information retrieval system.   
     
     
         2 . The computer system according to  claim 1 , said processor further generating hypotheses related to a medical condition of said patient based on said results. 
     
     
         3 . The computer system according to  claim 2 , said processor further generating a confidence score for said hypotheses. 
     
     
         4 . The computer system according to  claim 3 , said processor further ranking said hypotheses according to said confidence score for said hypotheses and recommending at least one analysis based on said hypotheses. 
     
     
         5 . The computer system according to  claim 2 , said processor further recommending at least one analysis based on said hypotheses. 
     
     
         6 . The computer system according to  claim 3 , said processor further storing a history of said results to said input terms, said hypotheses, and said confidence score for said hypotheses, in a non-transitory storage medium; and
 said processor further correlating previously generated hypotheses and corresponding confidence scores stored in said non-transitory storage medium with at least one analysis based on said hypotheses.   
     
     
         7 . The computer system according to  claim 1 , further comprising a user interface further outputting said results to said input terms and a link to said corpus of data indicating how said corpus of data contributed to said results. 
     
     
         8 . The computer system according to  claim 1 , said user interface comprising a graphic user interface. 
     
     
         9 . A device for analyzing physiological data streams comprising:
 a receiver receiving data from said physiological data streams, said data comprising information of a medical condition of a patient;   a question-answering system performing a plurality of question answering processes;   a processor connected to said question-answering system, said processor:
 extracting a list of features from said data in said physiological data streams based on development of said physiological data over a period of time; 
 generating at least one query based on said list of features by converting said list of features to natural language; 
 presenting said at least one query to said question-answering system; 
 receiving at least one response to said at least one query; and 
 developing a first hypothesis concerning a diagnosis of said medical condition of said patient based on said at least one response; and 
   a network interface outputting said list of features to external sources separate from said question-answering system and receiving at least one additional hypothesis concerning a diagnosis of said medical condition of said patient from said external sources,   said processor recommending at least one of a treatment for said medical condition of said patient and at least one additional analysis based on said first hypothesis and said at least one additional hypothesis by comparing said first hypothesis and said at least one additional hypothesis to said medical symptoms.   
     
     
         10 . The device according to  claim 9 , further comprising a user interface operatively connected to said processor, said user interface outputting at least one of said first hypothesis, said at least one additional hypothesis, said at least one treatment for said medical condition of said patient, and said at least one additional analysis. 
     
     
         11 . The device according to  claim 9 , said processor further generating a confidence score for said first hypothesis and said at least one additional hypothesis. 
     
     
         12 . The device according to  claim 11 , said processor further ranking said at least one additional analysis according to said confidence scores for said first hypothesis and said at least one additional hypothesis. 
     
     
         13 . The device according to  claim 9 , further comprising a non-transitory storage medium. 
     
     
         14 . The device according to  claim 13 , said processor further storing a history of said at least one response to said natural language query, said first hypothesis, said at least one additional hypothesis, and said confidence scores for said first hypothesis and said at least one additional hypothesis, in said non-transitory storage medium; and
 said processor further correlating previously generated hypotheses and corresponding confidence scores stored in said non-transitory storage medium with said at least one additional analysis based on said hypotheses.   
     
     
         15 . A non-transitory computer readable storage medium readable by a computerized device, said non-transitory computer readable storage medium storing instructions executable by said computerized device to perform a method comprising:
 receiving physiological data, said physiological data comprising numerical data and medical symptoms of a patient;   extracting features from said physiological data based on development of said physiological data over a period of time;   converting said features into a textual representation using natural language generation;   automatically generating input terms for an information retrieval system based on said features;   inputting said input terms to said information retrieval system; and   automatically searching a corpus of data to retrieve results to said input terms, using said information retrieval system.   
     
     
         16 . The non-transitory computer readable storage medium according to  claim 15 , said method further comprising generating hypotheses related to a medical condition of said patient based on said results. 
     
     
         17 . The non-transitory computer readable storage medium according to  claim 16 , said method further comprising generating a confidence score for said hypotheses. 
     
     
         18 . The non-transitory computer readable storage medium according to  claim 17 , said method further comprising ranking said hypotheses according to said confidence score for said hypotheses and recommending at least one analysis according to said confidence score for said hypotheses. 
     
     
         19 . The non-transitory computer readable storage medium according to  claim 17 , said method further comprising:
 storing, in a non-transitory storage medium, a history of said results to said input terms, said hypotheses, and said confidence score for said hypotheses; and   correlating previously generated hypotheses and corresponding confidence scores stored in said non-transitory storage medium with said at least one analysis based on said hypotheses.   
     
     
         20 . The non-transitory computer readable storage medium according to  claim 16 , said method further comprising recommending at least one analysis based on said hypotheses.