US20220114342A1 - Natural language processing of electronic records - Google Patents
Natural language processing of electronic records Download PDFInfo
- Publication number
- US20220114342A1 US20220114342A1 US17/369,908 US202117369908A US2022114342A1 US 20220114342 A1 US20220114342 A1 US 20220114342A1 US 202117369908 A US202117369908 A US 202117369908A US 2022114342 A1 US2022114342 A1 US 2022114342A1
- Authority
- US
- United States
- Prior art keywords
- records
- clinical
- journal
- intensity
- category
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003058 natural language processing Methods 0.000 title description 7
- 238000000034 method Methods 0.000 claims description 26
- 238000005303 weighing Methods 0.000 claims description 4
- 238000012015 optical character recognition Methods 0.000 claims 6
- 230000008569 process Effects 0.000 description 12
- 238000012545 processing Methods 0.000 description 12
- 238000004590 computer program Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 11
- 230000003542 behavioural effect Effects 0.000 description 9
- 208000019901 Anxiety disease Diseases 0.000 description 8
- 230000036506 anxiety Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 5
- 206010038743 Restlessness Diseases 0.000 description 4
- 230000003993 interaction Effects 0.000 description 4
- 210000003205 muscle Anatomy 0.000 description 4
- 206010010144 Completed suicide Diseases 0.000 description 3
- 230000006399 behavior Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 235000013305 food Nutrition 0.000 description 3
- 238000002560 therapeutic procedure Methods 0.000 description 3
- 206010012239 Delusion Diseases 0.000 description 2
- 208000004547 Hallucinations Diseases 0.000 description 2
- 238000013542 behavioral therapy Methods 0.000 description 2
- 231100000868 delusion Toxicity 0.000 description 2
- 208000024714 major depressive disease Diseases 0.000 description 2
- 230000036651 mood Effects 0.000 description 2
- 238000002610 neuroimaging Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 208000020016 psychiatric disease Diseases 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 235000003930 Aegle marmelos Nutrition 0.000 description 1
- 244000058084 Aegle marmelos Species 0.000 description 1
- 208000007415 Anhedonia Diseases 0.000 description 1
- 206010002942 Apathy Diseases 0.000 description 1
- 208000020925 Bipolar disease Diseases 0.000 description 1
- 206010012335 Dependence Diseases 0.000 description 1
- 208000007590 Disorders of Excessive Somnolence Diseases 0.000 description 1
- 206010016759 Flat affect Diseases 0.000 description 1
- 206010022998 Irritability Diseases 0.000 description 1
- ZDXPYRJPNDTMRX-VKHMYHEASA-N L-glutamine Chemical compound OC(=O)[C@@H](N)CCC(N)=O ZDXPYRJPNDTMRX-VKHMYHEASA-N 0.000 description 1
- 208000016285 Movement disease Diseases 0.000 description 1
- 206010049816 Muscle tightness Diseases 0.000 description 1
- 208000000112 Myalgia Diseases 0.000 description 1
- 208000002193 Pain Diseases 0.000 description 1
- 206010033864 Paranoia Diseases 0.000 description 1
- 208000027099 Paranoid disease Diseases 0.000 description 1
- 206010037213 Psychomotor retardation Diseases 0.000 description 1
- 208000013738 Sleep Initiation and Maintenance disease Diseases 0.000 description 1
- 206010042008 Stereotypy Diseases 0.000 description 1
- 208000010513 Stupor Diseases 0.000 description 1
- 201000007201 aphasia Diseases 0.000 description 1
- 235000019789 appetite Nutrition 0.000 description 1
- 230000036528 appetite Effects 0.000 description 1
- 230000004596 appetite loss Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 208000029560 autism spectrum disease Diseases 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 206010007776 catatonia Diseases 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 235000019788 craving Nutrition 0.000 description 1
- 230000000994 depressogenic effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 208000035475 disorder Diseases 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000003400 hallucinatory effect Effects 0.000 description 1
- 206010020765 hypersomnia Diseases 0.000 description 1
- 206010022437 insomnia Diseases 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 235000021266 loss of appetite Nutrition 0.000 description 1
- 208000019017 loss of appetite Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 208000013465 muscle pain Diseases 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 208000019116 sleep disease Diseases 0.000 description 1
- 208000022925 sleep disturbance Diseases 0.000 description 1
- 230000000392 somatic effect Effects 0.000 description 1
- 201000009032 substance abuse Diseases 0.000 description 1
- 231100000736 substance abuse Toxicity 0.000 description 1
- 208000011117 substance-related disease Diseases 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
Definitions
- This specification relates to natural-language analysis by a computer on historical records including electronic medical records.
- Natural language processing is a subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to process and analyze large amounts of natural language data. Challenges in natural language processing frequently involve speech recognition, natural language understanding, and natural-language generation.
- a natural language or ordinary language is any language that has evolved naturally in humans through use and repetition. Natural languages can take different forms, such as speech or signing. They are distinguished from constructed and formal languages such as those used to program computers or to study logic.
- This specification describes technologies for classifying subjects into behavioral categories based on natural language notes. These technologies generally involve classifying notes about a subject into one or more pre-existing classifications using natural-language analysis of the notes.
- one innovative aspect of the subject matter described in this specification can be embodied in methods that include the actions of accessing, from computer storage, electronic records for a given subject, wherein the electronic records include natural language notes about the subject; identifying tokens in the natural language notes; generating, for each token, a corresponding intensity score representing an intensity of match between the token and a particular dimension, wherein the intensity scores are each values on a first scale, wherein each dimension is one of a plurality of dimensions of a category out of a plurality of categories; generating rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale; compiling, for each dimension of each category, a dimension-score based on the intensity scores; and categorizing the subject into at least one category based on the dimension scores.
- inventions of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods.
- a system of one or more computers to be configured to perform particular operations or actions means that the system has installed on it software, firmware, hardware, or a combination of them that in operation cause the system to perform the operations or actions.
- one or more computer programs to be configured to perform particular operations or actions means that the one or more programs include instructions that, when executed by data processing apparatus, cause the apparatus to perform the operations or actions.
- the electronic records are electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients.
- the electronic records comprise documents written by the subject.
- the natural language notes comprise natural language notes written by a clinician of the subject. Identifying tokens in the natural language notes comprises: generating a list of all possible-tokens in the natural language notes; and selecting some, but not all of the possible-tokens based on a match with a particular dimension.
- the particular dimension is one dimension that is part of a category from a list of possible categories used by a clinician of the subject, each category having a plurality of dimensions.
- the second scale is anxiety, the second scale spans from 0 to 9 and the particular category is “Major Depressive Disorder”.
- the tokens comprise words, strings, and portions of words.
- the first scale spans from 0 to 1.
- Generating rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale comprises only generating rescaled-intensity scores for intensity scores that are greater than a threshold value.
- the subject matter described in this specification can be implemented in particular embodiments so as to realize one or more of the following advantages.
- the technology of computer-based natural-language processing is advanced.
- the use of natural-language notes as an input to behavior categorization can allow for new and previously-unrealized uses of existing digital data to provide medical services.
- This technology can provide categorization in clinically accepted categories such as those published in widely used diagnostic manuals. Therefore, this technology can advantageously aid existing clinician by providing information in formats and structures that are already familiar to them, using data that was never originally organized into such formats and structures.
- the resulting scores can then be used in combination with brain imaging data (e.g., connectome data) to train one or more machine learning model(s) to automatically tag brain imaging data and/or the associated subjects for follow-up for at least one condition.
- brain imaging data e.g., connectome data
- This technology can be used for better diagnostic assessments of patients. For example, by using objective computations instead of subjective analysis, patients can be provided with a minimum level of care regardless of the skill level of a clinician or sequence of clinicians. This may be of particular value to patients that move, change insurance providers, or are otherwise forced to change clinicians before any one clinician is able to form a complete picture of the patient.
- this technology can produce, for the first time, analysis from large-scale mining of records (e.g., EMRs) that can aid in the discovery and development of new clinical patterns that would be impossible to identify without the use of this technology.
- EMRs electronic medical record
- researchers may perform automatic and very fast analysis of subpopulations to study the effect of various interventions (e.g., to compare the effectiveness of competing drugs or behavioral therapies). This can increase the speed and flexibility of both hypothesis generation and hypothesis testing.
- this technology may be used to continually or periodically monitor patients that may be known to be, or not yet known to be, at risk of negative impacts.
- alerts may be set up to alert patients and/or caregivers if a patient is trending toward a negative impact such as risk of suicide, violence, substance abuse, etc.
- alerts may be raised early enough with a caregiver to prompt the caregiver to supplement their own subjective assessments with this technology's objective assessments to produce better outcomes than would be possible with subjective assessments alone.
- FIG. 1 shows an example system usable to classify a subject.
- FIG. 2 shows an example data pipeline that receives data and produces data stored to disk and/or displayed in a graphical user interface (GUI).
- GUI graphical user interface
- FIG. 3 shows an example process for generating classifications.
- FIG. 4 shows an example GUI.
- a computer system accesses historical electronic records such as medical records of a patient receiving behavioral therapy from a clinician.
- the computer system applies natural language processing procedures to identify terms in the records that are related to various dimensions of behavioral classifications, and scores these words based on intensity. This produces values in a first scale, and the computer system rescales them to different scales that are used in pre-existing dimensions of the behavioral classification. For example, the computer system can use these processes to categorize a subject into a behavioral category from Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, among other sources of classifications.
- FIG. 1 shows an example system 100 usable to classify a subject.
- a computer system 102 accesses electronic records, applies natural-language processing to the records, and classifies a subject discussed in the natural language of the records.
- the computer system 102 includes one or more real or virtual computing devices with memory and processors that execute instructions to perform computations.
- the system can also include networking components to connect various elements of the system, etc.
- the computer system 102 can include a distributed set of servers and one or more clients that send data to the servers and request execution on the data.
- the computer system 102 receives data about a subject.
- the data about the subject can include natural-language text written by one or more people about the subject.
- a clinician treating the subject can keep notes from therapy sessions stored as natural-language notes 104 .
- the subject may keep a journal of their behaviors and experiences, stored in journal entries 106 .
- the data 104 and 106 may be generated informally or ‘spur-of-the-moment’ in natural language without an attempt to specifically record clinically-accepted key words associated with a particular behavioral classification.
- technology described in this document can convert those natural-language elements into data that can be used to classify a subject into a pre-existing, formal category behavior.
- the computer system 102 can generate natural-language intensity values 108 .
- the natural-language intensity values 108 record a measure of how intense a particular token (e.g., word, phrase) is in natural use (e.g., as opposed to how intense the same token is understood to be in specialized jargon).
- “happy” may have a lower intensity score than “ecstatic”
- “busy” may have a lower intensity value than “frantic”.
- a single intensity scale may be used for every token, even tokens of different domains. That is to say, an intensity scale of 0 to 1 may be used, with “happy” and “busy” having values of 0.2 while “ecstatic” and “frantic” may having values of 0.8.
- a lexical and rule-based sentiment analysis tool may be trained to sum valence scores of each work in a lexicon, adjusted according to a rule-set, and then normalized between ⁇ 1 and 1 or 0 and 1.
- these lexicons, rulesets, and normalization ranges may be configured to match specific implementation details chosen in other parts of the system to ensure compatibility.
- the computer system 102 can generate rescaled-intensity scores 110 .
- the computer system 102 can adjust the intensity scores 108 from their original scale to the scale of the dimension they apply to. For example, “happy” may be placed on a scale from 1-5 and receive a rescaled-intensity score 110 of 1.0 (i.e. the original 0.2 multiplied by the scale length of 5 here), while “busy may be placed on a scale from 1-10 and receive a rescaled-intensity score of 2.0 (i.e. the original 0.2 multiplied by the scale length of 10 here).
- the computer system 102 can categorize the subject into one or more of the classifications. For example, for a particular classification (e.g., Major Depressive Disorder), there may be 9 dimensions, and an aggregate score of 5 or more across all dimensions may classify the subject into the classification of depressed. In some cases, a classification strength may be determined. For example, a score of 3- may be considered marginal or subclinical, a score of 5-7 may be considered mid-grade, and a score of 7 and above (e.g., 9) can be considered high-grade.
- a classification strength may be determined. For example, a score of 3- may be considered marginal or subclinical, a score of 5-7 may be considered mid-grade, and a score of 7 and above (e.g., 9) can be considered high-grade.
- the computer system may place the classification into dates or date ranges 112 .
- the tokens from a single document (and thus at a single time) or from documents within a time range may be collected and used to categorize the subject. Then, the subject may be placed in that classification for that date or date range. In this way, the classified dates 112 may be collected to create classification over time 114 for a subject.
- FIG. 2 shows an example data pipeline 200 that receives data and produces data stored to disk and/or displayed in a graphical user interface (GUI).
- the data pipeline 200 may be used, for example, by the computer system 100 in order to categorize a subject into one or more behavioral categories based on natural-language data.
- the pipeline 202 describes a number of modules, shown in rectangular boxes with sharp corners, which receive and generate data, shown in rectangular boxes with rounded corners.
- the pipeline 200 receives the data 104 and 106 .
- This data 104 and 106 can be collected from one source, or may be collected from multiple sources.
- One example source of the data 104 and 106 is electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients.
- EMRs electronic medical records
- the subject may be a particular patient receiving therapy from a clinic that treats hundreds of patients.
- Each of the patients may have EMR that are specific to the patient, and the pipeline 200 may receive only those EMR associated with the subject, that the subject has consented for use in this way.
- One example source of the data 104 and 106 is electronic records of notes written by a clinician for the subject. In some cases, these can be included as part of the EMR discussed previously. In some cases, these may be stored in formats other than EMRs. For example, a clinician may record notes during therapy with a pen and paper and then summarize the notes electronically afterward.
- the data 104 and 106 can be generated by scanning the pen-and-paper notes, applying object character recognition (OCR) to the scanned documents, and using the OCR text.
- OCR object character recognition
- One example source of the data 104 and 106 is documents written by the subject.
- a subject may keep a journal in which they document their emotions, actions, environment, etc. These journal entries may be used as the data 104 and 106 .
- other writings of the subject such as email, social-media posts, or fiction may be collected and used as the data 104 and 106 .
- a tokenizer 202 can receive the data 104 and 106 to generate a set of tokens 204 .
- tokens 204 are collections of characters partitioned into units on which natural-language processes can be performed.
- tokens may be single words, strings of works, and/or portions of words.
- the tokenizer 202 can receive a complete text expected to contain many tokens (e.g., a narrative entry in an EMR, a journal entry for a given day) and iterate through the characters of the text to identify a break between tokens. For example, space characters and certain punctuation may be identified by the tokenizer 202 to identify the beginning or end of a token.
- the some or all of the data 104 and 106 may include timestamps or other temporal information.
- the tokenizer 202 can identify, for each token, a timestamp from the electronic data 104 and 106 .
- the tokens 204 can be tagged with the timestamp of the source data 104 and 106 in which it occurs.
- a natural language intensity scorer 206 can receive the tokens 204 and, for each token, generate a natural language intensity score 108 .
- the natural language intensity scorer 206 may store a table of token or sub-token strings along with a corresponding intensity value along a first scale (e.g. 0 to 1, 1 to 100).
- a first scale e.g. 0 to 1, 1 to 100.
- a scale converter 208 can convert the natural language intensity scores 108 to the rescaled intensity scores 110 .
- the scale converter 208 can receive, as input, a natural language intensity score 108 , the original scale range for the natural language intensity score 108 , and target range to rescale too.
- This target range may be, for example, the scale of one of the dimensions of a classification from a group of pre-defined behavioral classification.
- this process may be repeated so that each and every natural language intensity score 108 has corresponding 110 rescaled intensity scores 110 for each dimension.
- each token 204 may be examined to match it with only a subset of related dimensions, and rescaled intensity scores 110 may be generated only for the matching dimensions.
- a dimension of clinical depression may be “anxiety”.
- the tokens “anxious” and “panic” may be matched to the anxiety dimension, while the tokens “shoe” and “sky” may not be matched. Examples of tokens used for matching can be found later in this document.
- a categorizer 210 can use the rescaled intensity scores 110 and their corresponding timestamps from the data 104 and 106 to create the categorized dates 110 .
- the categorizer may record that, for all dates, the subject is not classified into any of the pre-existing behavioral classifications. Often, a subject will be classified into only one or a small percentage of the possible classifications. However, there may be subjects for whom the data 104 and 106 indicates classification into many categories.
- the categorizer 210 can generate categorized dates 110 that include a strength value (e.g., mild vs heavy).
- the categorizer 210 may include stored data that defines dimensions that are each part of a category from a list of possible categories used by a clinician of the subject, each category having a plurality of dimensions.
- One example category, depression is listed at the end of this document along with its dimensions.
- a token or sub-token is listed.
- the intensity of the token is included in the running score for that dimension for that category for the subject. For example, the token “anxious” would increase the value for the “anxiety” dimension, while “calm” would decrease the “anxiety” dimension.
- One such scheme for this inclusion of value is an average of intensity of all tokens.
- Another scheme may include weighing some tokens, for example, by a recency value, by their location in a text, etc.
- a timeline organizer 212 can use the categorized dates 110 to generate categories over time 214 for the subject. For example, the timeline organizer may arrange the categories, and corresponding strengths, in a timeline to show change in categorizations and strength over time for the subject based on the data 104 and 106 .
- a GUI engine 216 can generate one or more GUIs showing the categories over time 214 .
- the GUI engine 216 may generate webpages or native application interfaces shown numerical information, graphs, or other data.
- One such example GUI is shown in FIG. 5 .
- the GUI engine can send an alarm when a category reaches a threshold level set on a category by category basis.
- a datastore 218 can store the categories over time 214 , and any other data in the pipeline 200 , for use by other computing systems.
- the datastore 218 may be part of an EMR for the subject in a patient-management system.
- FIG. 3 shows an example process 300 for generating classifications.
- the process 300 can be performed, for example, by the computer system 100 using the pipeline 200 and will therefore be described with reference to the computer system 100 using the pipeline 200 .
- other systems may be used to perform the process 300 or similar processes.
- Records are accessed 302 .
- the computer system 102 and access, from computer storage, electronic records for a given subject, wherein the electronic records include natural language notes about the subject.
- These can include the data 104 and 106 as well as data from other sources.
- Tokens are identified 304 .
- the computer system 102 can identify tokens in the natural language notes.
- the tokenizer 202 can parse text document looking for stop characters that indicate the end and beginning of a token.
- Example stop characters include, but are not limited to, periods, spaces, and commas.
- the tokenizer 202 can initially generate a list of all possible-tokens in the data 104 and 106 . Then, the tokenizer 202 can select some, but not all of the possible-tokens. For example, the tokenizer 202 (or another element of the pipeline 200 ) can identify tokens that are a match with a particular dimension or all dimensions of one or all possible categories.
- Intensity scores are generated 306 .
- the computer system 102 can generate, for each token or each token matching a dimension, a corresponding intensity score representing an intensity of match between the token and a particular dimension.
- these intensity scores may be domain-neutral and represent a general intensity and do not necessarily indicate a closeness or nearness-of-fit with a particular category or dimension of category.
- Rescaled-intensity scores are generated 308 .
- the computer system 102 can generate rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale. That is to say, for each scored token, the score can be adjusted based on the scale used by the categories that are of interest and used by the computer system 102 .
- this can include only generating rescaled-intensity scores for intensity scores that are greater than a threshold value. For example, tokens with intensity values below a particular threshold (e.g., 0.1 on a scale of 0 to 1) may be discarded. This may exclude, for example, data in which a large number of low intensity tokens are found.
- a threshold value e.g., 0.1 on a scale of 0 to 1.
- Dimension scores are compiled 310 .
- the computer system 102 can compile for each dimension of each category, a dimension-score based on the intensity scores. In some cases, this may include adding together the intensity scores of tokens matching the dimensions. In some cases, this may include finding an average of the intensity scores of tokens matching the dimension.
- the subject is categorized 312 .
- the computer system 102 can categorize the subject into at least one category based on the dimension scores. This may include, for example, identify categories for which the various dimensions all have dimension-scores above corresponding threshold values.
- FIG. 4 shows an example GUI 400 .
- the GUI 400 is shown here as a webpage rendered in a web browser, but other formats of GUIs are possible.
- the GUI 400 includes a timeline 402 along which a trend line 404 of classification strength is shown for a given classification (e.g., depression, addiction, autism spectrum disorder, bipolar disorder, etc).
- a given classification e.g., depression, addiction, autism spectrum disorder, bipolar disorder, etc.
- Embodiments of the subject matter and the functional operations described in this specification can be implemented in digital electronic circuitry, in tangibly-embodied computer software or firmware, in computer hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them.
- Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible non-transitory storage medium for execution by, or to control the operation of, data processing apparatus.
- the computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them.
- the program instructions can be encoded on an artificially-generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus.
- data processing apparatus refers to data processing hardware and encompasses all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers.
- the apparatus can also be, or further include, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit).
- the apparatus can optionally include, in addition to hardware, code that creates an execution environment for computer programs, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
- a computer program which may also be referred to or described as a program, software, a software application, an app, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
- a program may, but need not, correspond to a file in a file system.
- a program can be stored in a portion of a file that holds other programs or data, e.g., one or more scripts stored in a markup language document, in a single file dedicated to the program in question, or in multiple coordinated files, e.g., files that store one or more modules, sub-programs, or portions of code.
- a computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a data communication network.
- the processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform functions by operating on input data and generating output.
- the processes and logic flows can also be performed by special purpose logic circuitry, e.g., an FPGA or an ASIC, or by a combination of special purpose logic circuitry and one or more programmed computers.
- Computers suitable for the execution of a computer program can be based on general or special purpose microprocessors or both, or any other kind of central processing unit.
- a central processing unit will receive instructions and data from a read-only memory or a random access memory or both.
- the essential elements of a computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and data.
- the central processing unit and the memory can be supplemented by, or incorporated in, special purpose logic circuitry.
- a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks.
- a computer need not have such devices.
- a computer can be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a portable storage device, e.g., a universal serial bus (USB) flash drive, to name just a few.
- PDA personal digital assistant
- GPS Global Positioning System
- USB universal serial bus
- Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks.
- semiconductor memory devices e.g., EPROM, EEPROM, and flash memory devices
- magnetic disks e.g., internal hard disks or removable disks
- magneto-optical disks e.g., CD-ROM and DVD-ROM disks.
- embodiments of the subject matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user can provide input to the computer.
- a display device e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor
- keyboard and a pointing device e.g., a mouse or a trackball
- Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input.
- a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user's device in response to requests received from the web browser.
- a computer can interact with a user by sending text messages or other forms of message to a personal device, e.g., a smartphone, running a messaging application, and receiving responsive messages from the user in return.
- Embodiments of the subject matter described in this specification can be implemented in a computing system that includes a back-end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front-end component, e.g., a client computer having a graphical user interface, a web browser, or an app through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back-end, middleware, or front-end components.
- the components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (LAN) and a wide area network (WAN), e.g., the Internet.
- LAN local area network
- WAN wide area network
- the computing system can include clients and servers.
- a client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other.
- a server transmits data, e.g., an HTML page, to a user device, e.g., for purposes of displaying data to and receiving user input from a user interacting with the device, which acts as a client.
- Data generated at the user device e.g., a result of the user interaction, can be received at the server from the device.
Abstract
Electronic records are accessed from computer storage for a given subject, wherein the electronic records include natural language notes about the subject. Tokens are identified in the natural language notes. For each token, a corresponding intensity score is generated representing an intensity of match between the token and a particular dimension, wherein the intensity scores are each values on a first scale, wherein each dimension is one of a plurality of dimensions of a category out of a plurality of categories; generating rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale. For each dimension of each category, a dimension-score is compiled based on the intensity scores; and categorizing the subject into at least one category based on the dimension scores. The subject is categorized into at least one category based on the dimension scores.
Description
- This application is a continuation of and claims priority to U.S. patent application Ser. No. 17/066,335, filed Oct. 8, 2020, the entire contents of which are incorporated herein by reference.
- This specification relates to natural-language analysis by a computer on historical records including electronic medical records.
- Natural language processing (NLP) is a subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to process and analyze large amounts of natural language data. Challenges in natural language processing frequently involve speech recognition, natural language understanding, and natural-language generation.
- A natural language or ordinary language is any language that has evolved naturally in humans through use and repetition. Natural languages can take different forms, such as speech or signing. They are distinguished from constructed and formal languages such as those used to program computers or to study logic.
- This specification describes technologies for classifying subjects into behavioral categories based on natural language notes. These technologies generally involve classifying notes about a subject into one or more pre-existing classifications using natural-language analysis of the notes.
- In general, one innovative aspect of the subject matter described in this specification can be embodied in methods that include the actions of accessing, from computer storage, electronic records for a given subject, wherein the electronic records include natural language notes about the subject; identifying tokens in the natural language notes; generating, for each token, a corresponding intensity score representing an intensity of match between the token and a particular dimension, wherein the intensity scores are each values on a first scale, wherein each dimension is one of a plurality of dimensions of a category out of a plurality of categories; generating rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale; compiling, for each dimension of each category, a dimension-score based on the intensity scores; and categorizing the subject into at least one category based on the dimension scores.
- Other embodiments of this aspect include corresponding computer systems, apparatus, and computer programs recorded on one or more computer storage devices, each configured to perform the actions of the methods. For a system of one or more computers to be configured to perform particular operations or actions means that the system has installed on it software, firmware, hardware, or a combination of them that in operation cause the system to perform the operations or actions. For one or more computer programs to be configured to perform particular operations or actions means that the one or more programs include instructions that, when executed by data processing apparatus, cause the apparatus to perform the operations or actions.
- The foregoing and other embodiments can each optionally include one or more of the following features, alone or in combination. In particular, one embodiment includes all the following features in combination.
- The electronic records are electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients. The electronic records comprise documents written by the subject. The natural language notes comprise natural language notes written by a clinician of the subject. Identifying tokens in the natural language notes comprises: generating a list of all possible-tokens in the natural language notes; and selecting some, but not all of the possible-tokens based on a match with a particular dimension. The particular dimension is one dimension that is part of a category from a list of possible categories used by a clinician of the subject, each category having a plurality of dimensions. The second scale is anxiety, the second scale spans from 0 to 9 and the particular category is “Major Depressive Disorder”. The tokens comprise words, strings, and portions of words. The first scale spans from 0 to 1. Generating rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale comprises only generating rescaled-intensity scores for intensity scores that are greater than a threshold value.
- The subject matter described in this specification can be implemented in particular embodiments so as to realize one or more of the following advantages. The technology of computer-based natural-language processing is advanced. The use of natural-language notes as an input to behavior categorization can allow for new and previously-unrealized uses of existing digital data to provide medical services. This technology can provide categorization in clinically accepted categories such as those published in widely used diagnostic manuals. Therefore, this technology can advantageously aid existing clinician by providing information in formats and structures that are already familiar to them, using data that was never originally organized into such formats and structures. The resulting scores can then be used in combination with brain imaging data (e.g., connectome data) to train one or more machine learning model(s) to automatically tag brain imaging data and/or the associated subjects for follow-up for at least one condition. This technology can be used for better diagnostic assessments of patients. For example, by using objective computations instead of subjective analysis, patients can be provided with a minimum level of care regardless of the skill level of a clinician or sequence of clinicians. This may be of particular value to patients that move, change insurance providers, or are otherwise forced to change clinicians before any one clinician is able to form a complete picture of the patient. For researchers, this technology can produce, for the first time, analysis from large-scale mining of records (e.g., EMRs) that can aid in the discovery and development of new clinical patterns that would be impossible to identify without the use of this technology. For example, researchers may perform automatic and very fast analysis of subpopulations to study the effect of various interventions (e.g., to compare the effectiveness of competing drugs or behavioral therapies). This can increase the speed and flexibility of both hypothesis generation and hypothesis testing. For both patients and caregivers, this technology may be used to continually or periodically monitor patients that may be known to be, or not yet known to be, at risk of negative impacts. That is to say, alerts may be set up to alert patients and/or caregivers if a patient is trending toward a negative impact such as risk of suicide, violence, substance abuse, etc. By continual monitoring of records, alerts may be raised early enough with a caregiver to prompt the caregiver to supplement their own subjective assessments with this technology's objective assessments to produce better outcomes than would be possible with subjective assessments alone.
- The details of one or more embodiments of the subject matter of this specification are set forth in the accompanying drawings and the description below. Other features, aspects, and advantages of the subject matter will become apparent from the description, the drawings, and the claims.
-
FIG. 1 shows an example system usable to classify a subject. -
FIG. 2 shows an example data pipeline that receives data and produces data stored to disk and/or displayed in a graphical user interface (GUI). -
FIG. 3 shows an example process for generating classifications. -
FIG. 4 shows an example GUI. - Like reference numbers and designations in the various drawings indicate like elements.
- A computer system accesses historical electronic records such as medical records of a patient receiving behavioral therapy from a clinician. The computer system applies natural language processing procedures to identify terms in the records that are related to various dimensions of behavioral classifications, and scores these words based on intensity. This produces values in a first scale, and the computer system rescales them to different scales that are used in pre-existing dimensions of the behavioral classification. For example, the computer system can use these processes to categorize a subject into a behavioral category from Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, among other sources of classifications.
-
FIG. 1 shows anexample system 100 usable to classify a subject. In thesystem 100, acomputer system 102 accesses electronic records, applies natural-language processing to the records, and classifies a subject discussed in the natural language of the records. - The
computer system 102 includes one or more real or virtual computing devices with memory and processors that execute instructions to perform computations. The system can also include networking components to connect various elements of the system, etc. In some cases, thecomputer system 102 can include a distributed set of servers and one or more clients that send data to the servers and request execution on the data. - The
computer system 102 receives data about a subject. In general, the data about the subject can include natural-language text written by one or more people about the subject. For example, a clinician treating the subject can keep notes from therapy sessions stored as natural-language notes 104. In another example, the subject may keep a journal of their behaviors and experiences, stored injournal entries 106. Thedata computer system 102 can examine thedata - The
computer system 102 can generate natural-language intensity values 108. In general, the natural-language intensity values 108 record a measure of how intense a particular token (e.g., word, phrase) is in natural use (e.g., as opposed to how intense the same token is understood to be in specialized jargon). In one example, “happy” may have a lower intensity score than “ecstatic”, and similarly “busy” may have a lower intensity value than “frantic”. In many cases, a single intensity scale may be used for every token, even tokens of different domains. That is to say, an intensity scale of 0 to 1 may be used, with “happy” and “busy” having values of 0.2 while “ecstatic” and “frantic” may having values of 0.8. For example, a lexical and rule-based sentiment analysis tool may be trained to sum valence scores of each work in a lexicon, adjusted according to a rule-set, and then normalized between −1 and 1 or 0 and 1. As will be understood, these lexicons, rulesets, and normalization ranges may be configured to match specific implementation details chosen in other parts of the system to ensure compatibility. - The
computer system 102 can generate rescaled-intensity scores 110. As various dimensions of various categories may use different scales, thecomputer system 102 can adjust the intensity scores 108 from their original scale to the scale of the dimension they apply to. For example, “happy” may be placed on a scale from 1-5 and receive a rescaled-intensity score 110 of 1.0 (i.e. the original 0.2 multiplied by the scale length of 5 here), while “busy may be placed on a scale from 1-10 and receive a rescaled-intensity score of 2.0 (i.e. the original 0.2 multiplied by the scale length of 10 here). - With the rescaled-
intensity scores 110, thecomputer system 102 can categorize the subject into one or more of the classifications. For example, for a particular classification (e.g., Major Depressive Disorder), there may be 9 dimensions, and an aggregate score of 5 or more across all dimensions may classify the subject into the classification of depressed. In some cases, a classification strength may be determined. For example, a score of 3- may be considered marginal or subclinical, a score of 5-7 may be considered mid-grade, and a score of 7 and above (e.g., 9) can be considered high-grade. - The computer system may place the classification into dates or date ranges 112. For example, if the
data classified dates 112 may be collected to create classification overtime 114 for a subject. -
FIG. 2 shows anexample data pipeline 200 that receives data and produces data stored to disk and/or displayed in a graphical user interface (GUI). Thedata pipeline 200 may be used, for example, by thecomputer system 100 in order to categorize a subject into one or more behavioral categories based on natural-language data. Thepipeline 202 describes a number of modules, shown in rectangular boxes with sharp corners, which receive and generate data, shown in rectangular boxes with rounded corners. - The
pipeline 200 receives thedata data data pipeline 200 may receive only those EMR associated with the subject, that the subject has consented for use in this way. - One example source of the
data data - One example source of the
data data data - A
tokenizer 202 can receive thedata tokens 204. In general,tokens 204 are collections of characters partitioned into units on which natural-language processes can be performed. In some examples, tokens may be single words, strings of works, and/or portions of words. Thetokenizer 202 can receive a complete text expected to contain many tokens (e.g., a narrative entry in an EMR, a journal entry for a given day) and iterate through the characters of the text to identify a break between tokens. For example, space characters and certain punctuation may be identified by thetokenizer 202 to identify the beginning or end of a token. - In some cases, the some or all of the
data tokenizer 202 can identify, for each token, a timestamp from theelectronic data tokens 204 can be tagged with the timestamp of thesource data - A natural
language intensity scorer 206 can receive thetokens 204 and, for each token, generate a naturallanguage intensity score 108. For example, the naturallanguage intensity scorer 206 may store a table of token or sub-token strings along with a corresponding intensity value along a first scale (e.g. 0 to 1, 1 to 100). When a token 204 is received, the naturallanguage intensity scorer 206 can match the token to a token or sub-token on the table and return the corresponding value. - A scale converter 208 can convert the natural language intensity scores 108 to the rescaled intensity scores 110. For example, the scale converter 208 can receive, as input, a natural
language intensity score 108, the original scale range for the naturallanguage intensity score 108, and target range to rescale too. This target range may be, for example, the scale of one of the dimensions of a classification from a group of pre-defined behavioral classification. In some cases, this process may be repeated so that each and every naturallanguage intensity score 108 has corresponding 110 rescaled intensity scores 110 for each dimension. In some cases, each token 204 may be examined to match it with only a subset of related dimensions, and rescaled intensity scores 110 may be generated only for the matching dimensions. For example, a dimension of clinical depression may be “anxiety”. In such a case, the tokens “anxious” and “panic” may be matched to the anxiety dimension, while the tokens “shoe” and “sky” may not be matched. Examples of tokens used for matching can be found later in this document. - A categorizer 210 can use the rescaled intensity scores 110 and their corresponding timestamps from the
data data dates 110 that include a strength value (e.g., mild vs heavy). - The categorizer 210 may include stored data that defines dimensions that are each part of a category from a list of possible categories used by a clinician of the subject, each category having a plurality of dimensions. One example category, depression, is listed at the end of this document along with its dimensions. For each dimension, a token or sub-token is listed. When a token 202 matches a token or sub-token in the list, the intensity of the token is included in the running score for that dimension for that category for the subject. For example, the token “anxious” would increase the value for the “anxiety” dimension, while “calm” would decrease the “anxiety” dimension. One such scheme for this inclusion of value is an average of intensity of all tokens. Another scheme may include weighing some tokens, for example, by a recency value, by their location in a text, etc.
- A
timeline organizer 212 can use the categorized dates 110 to generate categories over time 214 for the subject. For example, the timeline organizer may arrange the categories, and corresponding strengths, in a timeline to show change in categorizations and strength over time for the subject based on thedata - A
GUI engine 216 can generate one or more GUIs showing the categories over time 214. For example, theGUI engine 216 may generate webpages or native application interfaces shown numerical information, graphs, or other data. One such example GUI is shown inFIG. 5 . The GUI engine can send an alarm when a category reaches a threshold level set on a category by category basis. - A
datastore 218 can store the categories over time 214, and any other data in thepipeline 200, for use by other computing systems. For example, thedatastore 218 may be part of an EMR for the subject in a patient-management system. -
FIG. 3 shows anexample process 300 for generating classifications. Theprocess 300 can be performed, for example, by thecomputer system 100 using thepipeline 200 and will therefore be described with reference to thecomputer system 100 using thepipeline 200. However, other systems may be used to perform theprocess 300 or similar processes. - Records are accessed 302. For example, the
computer system 102 and access, from computer storage, electronic records for a given subject, wherein the electronic records include natural language notes about the subject. These can include thedata - Tokens are identified 304. For example, the
computer system 102 can identify tokens in the natural language notes. In some cases, thetokenizer 202 can parse text document looking for stop characters that indicate the end and beginning of a token. Example stop characters include, but are not limited to, periods, spaces, and commas. - The
tokenizer 202 can initially generate a list of all possible-tokens in thedata tokenizer 202 can select some, but not all of the possible-tokens. For example, the tokenizer 202 (or another element of the pipeline 200) can identify tokens that are a match with a particular dimension or all dimensions of one or all possible categories. - Intensity scores are generated 306. For example, the
computer system 102 can generate, for each token or each token matching a dimension, a corresponding intensity score representing an intensity of match between the token and a particular dimension. As will be understood, these intensity scores may be domain-neutral and represent a general intensity and do not necessarily indicate a closeness or nearness-of-fit with a particular category or dimension of category. - Rescaled-intensity scores are generated 308. For example, the
computer system 102 can generate rescaled-intensity scores from the intensity scores by rescaling the intensity scores from the first scale to a second scale different from the first scale. That is to say, for each scored token, the score can be adjusted based on the scale used by the categories that are of interest and used by thecomputer system 102. - In some cases, this can include only generating rescaled-intensity scores for intensity scores that are greater than a threshold value. For example, tokens with intensity values below a particular threshold (e.g., 0.1 on a scale of 0 to 1) may be discarded. This may exclude, for example, data in which a large number of low intensity tokens are found.
- Dimension scores are compiled 310. For example, the
computer system 102 can compile for each dimension of each category, a dimension-score based on the intensity scores. In some cases, this may include adding together the intensity scores of tokens matching the dimensions. In some cases, this may include finding an average of the intensity scores of tokens matching the dimension. - The subject is categorized 312. For example, the
computer system 102 can categorize the subject into at least one category based on the dimension scores. This may include, for example, identify categories for which the various dimensions all have dimension-scores above corresponding threshold values. -
FIG. 4 shows anexample GUI 400. TheGUI 400 is shown here as a webpage rendered in a web browser, but other formats of GUIs are possible. TheGUI 400 includes atimeline 402 along which atrend line 404 of classification strength is shown for a given classification (e.g., depression, addiction, autism spectrum disorder, bipolar disorder, etc). - Embodiments of the subject matter and the functional operations described in this specification can be implemented in digital electronic circuitry, in tangibly-embodied computer software or firmware, in computer hardware, including the structures disclosed in this specification and their structural equivalents, or in combinations of one or more of them. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible non-transitory storage medium for execution by, or to control the operation of, data processing apparatus. The computer storage medium can be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or a combination of one or more of them. Alternatively or in addition, the program instructions can be encoded on an artificially-generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus.
- The term “data processing apparatus” refers to data processing hardware and encompasses all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, or multiple processors or computers. The apparatus can also be, or further include, special purpose logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit). The apparatus can optionally include, in addition to hardware, code that creates an execution environment for computer programs, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of them.
- A computer program, which may also be referred to or described as a program, software, a software application, an app, a module, a software module, a script, or code, can be written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages; and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data, e.g., one or more scripts stored in a markup language document, in a single file dedicated to the program in question, or in multiple coordinated files, e.g., files that store one or more modules, sub-programs, or portions of code. A computer program can be deployed to be executed on one computer or on multiple computers that are located at one site or distributed across multiple sites and interconnected by a data communication network.
- The processes and logic flows described in this specification can be performed by one or more programmable computers executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows can also be performed by special purpose logic circuitry, e.g., an FPGA or an ASIC, or by a combination of special purpose logic circuitry and one or more programmed computers.
- Computers suitable for the execution of a computer program can be based on general or special purpose microprocessors or both, or any other kind of central processing unit. Generally, a central processing unit will receive instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a central processing unit for performing or executing instructions and one or more memory devices for storing instructions and data. The central processing unit and the memory can be supplemented by, or incorporated in, special purpose logic circuitry. Generally, a computer will also include, or be operatively coupled to receive data from or transfer data to, or both, one or more mass storage devices for storing data, e.g., magnetic, magneto-optical disks, or optical disks. However, a computer need not have such devices. Moreover, a computer can be embedded in another device, e.g., a mobile telephone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a Global Positioning System (GPS) receiver, or a portable storage device, e.g., a universal serial bus (USB) flash drive, to name just a few.
- Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks.
- To provide for interaction with a user, embodiments of the subject matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor, for displaying information to the user and a keyboard and a pointing device, e.g., a mouse or a trackball, by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback; and input from the user can be received in any form, including acoustic, speech, or tactile input. In addition, a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user's device in response to requests received from the web browser. Also, a computer can interact with a user by sending text messages or other forms of message to a personal device, e.g., a smartphone, running a messaging application, and receiving responsive messages from the user in return.
- Embodiments of the subject matter described in this specification can be implemented in a computing system that includes a back-end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front-end component, e.g., a client computer having a graphical user interface, a web browser, or an app through which a user can interact with an implementation of the subject matter described in this specification, or any combination of one or more such back-end, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network (LAN) and a wide area network (WAN), e.g., the Internet.
- The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. In some embodiments, a server transmits data, e.g., an HTML page, to a user device, e.g., for purposes of displaying data to and receiving user input from a user interacting with the device, which acts as a client. Data generated at the user device, e.g., a result of the user interaction, can be received at the server from the device.
- While this specification contains many specific implementation details, these should not be construed as limitations on the scope of any invention or on the scope of what may be claimed, but rather as descriptions of features that may be specific to particular embodiments of particular inventions. Certain features that are described in this specification in the context of separate embodiments can also be implemented in combination in a single embodiment. Conversely, various features that are described in the context of a single embodiment can also be implemented in multiple embodiments separately or in any suitable subcombination. Moreover, although features may be described above as acting in certain combinations and even initially be claimed as such, one or more features from a claimed combination can in some cases be excised from the combination, and the claimed combination may be directed to a subcombination or variation of a subcombination.
- Similarly, while operations are depicted in the drawings in a particular order, this should not be understood as requiring that such operations be performed in the particular order shown or in sequential order, or that all illustrated operations be performed, to achieve desirable results. In certain circumstances, multitasking and parallel processing may be advantageous. Moreover, the separation of various system modules and components in the embodiments described above should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.
- Particular embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. For example, the actions recited in the claims can be performed in a different order and still achieve desirable results. As one example, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some cases, multitasking and parallel processing may be advantageous.
- One example category, depression, is listed at the end of this document along with its dimensions. For each dimension, token(s) and/or sub-token(s) is/are listed:
- anxiety:
-
- positive:
- anxious
- anxiety
- worri
- panic
- fear
- uneasiness
- negative:
- calm
- still
- delusions:
- positive:
- delusion
- false beli
- fantasy
- paranoia
- preoccupation
- suspicious
- persecut
- hostile
- somatic concern
- unusual tought
- positive:
- disorg_speech:
- negative:
- coherent
- positive:
- disorganized speech
- derailment
- incoherence
- negative:
- disorganized:
- negative:
- organized
- positive:
- disorganized
- catatonic
- stupor
- excit
- restless
- disorientation
- bizarre behavio
- conceptual disorg
- anxiety
- grandios
- mannerism
- poor attention
- thought disorder
- movement disorder
- negative:
- fatigue:
- negative:
- energy
- positive:
- fatigue
- anergia
- tired
- negative:
- food:
- negative:
- weight
- appetite
- food
- positive:
- weight
- loss of appetite
- food
- cravings
- negative:
- hallucinations:
- positive:
- hallucination
- hallucinate
- hallucinatory
- positive:
- interest:
- negative:
- interest
- pleasure
- sex
- hobbies
- sport
- positive:
- anhedonia
- negative:
- irritability:
- positive:
- irrita
- positive:
- mood:
- negative:
- happy
- mood
- interest
- positive:
- depress
- sad
- irrita
- anxi
- cry
- glum
- downcast
- unhappy
- sadness
- sad
- tearful
- empty
- empti
- hopeless
- frustration
- negative:
- muscle:
- positive:
- muscle pain
- muscle tension
- stiff muscle
- muscles tension
- muscles pain
- positive:
- negsym:
- negative:
- move
- speak
- motivated
- spontaneity
- conversation
- coperative
- attention
- positive:
- affective flattening
- alogia
- avolition
- mute
- unmotivated
- uncooperative
- apath
- withdraw
- blunt
- poor rapport
- stereotyped thinking
- motor retardation
- depression
- poor attention
- avoidance
- immobility
- flat affect
- reduced feeling
- negative:
- psych:
- positive:
- psychomotor retardation
- slow
- positive:
- restless:
- negative:
- rest
- relax
- positive:
- restless
- on the edge
- negative:
- sleep:
- negative:
- sleep
- slept
- staying asleep
- falling asleep
- positive:
- insomnia
- hypersomnia
- difficulty falling asleep
- difficulty speeling
- unstatisfying sleep
- restless
- sleep disturbance
- negative:
- suicide:
- negative:
- live
- living
- positive:
- death
- suicide
- self-harm
- death
- negative:
- concentration:
- negative:
- concentrate
- think
- focus
- positive:
- indecisiv
- inattentive
- mind going blank
- negative:
- worthless:
- negative:
- esteem
- positive:
- worthless
- hopeless
- guilt
- negative:
- positive:
Claims (43)
1. A method comprising:
accessing, from computer storage, clinical records for a patient prepared by a clinician treating the patient, the clinical records comprising natural language clinical-notes about the subject;
accessing, from computer storage, journals records prepared from journaling by the patient, the journal records comprising natural language entries;
combining the clinical records with the journal records to produce a plurality of clinical-journal tokens from at least one of the clinical records and the journal records, each of the clinical-journal tokens having an associated time based on the at least one of the clinical records and the journal records; and
creating a categorizing time-series for the subject based on the clinical journal tokens, the categorizing time-series showing intensity of a clinically-diagnostic category for the subject over a time period in which the clinical records were prepared and the journal records were prepared.
2. The method of claim 1 , wherein the clinical records are electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients and wherein the journal records are stored in an electronic journal for a single one of the many patients.
3. (canceled)
4. (canceled)
5. (canceled)
6. (canceled)
7. (canceled)
8. The method of claim 1 , wherein the clinical-journal tokens comprise words, strings, and portions of words.
9. (canceled)
10. (canceled)
11. The method of claim 1 , wherein:
the clinical records and the journal records include timestamps;
identifying, for each clinical-journal token, a timestamp from the clinical records or the journal records; and
the method further comprises associating the associated time for each clinical-journal token based on one of the timestamps.
12. A system comprising:
one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
accessing, from computer storage, clinical records for a patient prepared by a clinician treating the patient, the clinical records comprising natural language clinical-notes about the subject;
accessing, from computer storage, journals records prepared from journaling by the patient, the journal records comprising natural language entries;
combining the clinical records with the journal records to produce a plurality of clinical-journal tokens from at least one of the clinical records and the journal records, each of the clinical journal tokens having an associated time based on the at least one of the clinical records and the journal records; and
creating a categorizing time-series for the subject based on the clinical journal tokens, the categorizing time-series showing intensity of a clinically-diagnostic category for the subject over a time period in which the clinical records were prepared and the journal records were prepared.
13. The system of claim 12 , wherein the clinical records are electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients and wherein the journal records are stored in an electronic journal for a single one of the many patients.
14. (canceled)
15. (canceled)
16. (canceled)
17. (canceled)
18. (canceled)
19. (canceled)
20. One or more non-transitory computer-readable storage media encoded with instructions that, when executed by one or more computers, cause the one or more computers to perform operations comprising:
accessing, from computer storage, clinical records for a patient prepared by a clinician treating the patient, the clinical records comprising natural language clinical-notes about the subject;
accessing, from computer storage, journals records prepared from journaling by the patient, the journal records comprising natural language entries;
combining the clinical records with the journal records to produce a plurality of clinical-journal tokens from at least one of the clinical records and the journal records, each of the clinical journal tokens having an associated time based on the at least one of the clinical records and the journal records; and
creating a categorizing time-series for the subject based on the clinical journal tokens, the categorizing time-series showing intensity of a clinically-diagnostic category for the subject over a time period in which the clinical records were prepared and the journal records were prepared.
21. The method of claim 1 , wherein combining the clinical records with the journal records to produce a plurality of clinical journal tokens comprises:
generating, for each clinical-journal token, a plurality of intensity scores, each intensity score being a numeric-value on a single scale such that none of the intensity values are greater than a maximum for the single scale and such that none of the intensity values are less a minimum for the single scale; and
translating the intensity scores to translated-scores in category-specific scales, at least some of the category-specific scales having a maximum for the category-specific scale different than the maximum for the single scale.
22. The method of claim 1 , wherein generating a time-series icon by:
for at least a part of the time period, generating an element with a geometric parameter reflective of the intensity of the clinically-diagnostic category, such that changes in the intensity correspond to changes in the geometric parameter; and
causing the element to be rendered in a graphical user interface (GUI) as a time-series icon to show to a user the changes in the intensity by way of the geometric parameter.
23. The method of claim 22 , wherein the element is a line chart and wherein the geometric parameter is a y-value in the line chart reflective of the intensity of the clinically-diagnosed category.
24. The method of claim 22 , wherein causing the element to be rendered comprises serving to a client device a webpage with instructions to instruct the client device to render the element in a web browser.
25. The system of claim 12 , wherein the clinical journal tokens comprise words, strings, and portions of words.
26. The system of claim 12 , wherein:
the clinical records and the journal records include timestamps;
identifying, for each clinical-journal token, a timestamp from the clinical records or the journal records; and
the operations further comprises associating the associated time for each clinical journal token based on one of the timestamps.
27. The system of claim 12 , wherein combining the clinical records with the journal records to produce a plurality of clinical journal tokens comprises:
generating, for each clinical-journal token, a plurality of intensity scores, each intensity score being a numeric-value on a single scale such that none of the intensity values are greater than a maximum for the single scale and such that none of the intensity values are less a minimum for the single scale; and
translating the intensity scores to translated-scores in category-specific scales, at least some of the category-specific scales having a maximum for the category-specific scale different than the maximum for the single scale.
28. The system of claim 12 , wherein generating a time-series icon by:
for at least a part of the time period, generating an element with a geometric parameter reflective of the intensity of the clinically-diagnostic category, such that changes in the intensity correspond to changes in the geometric parameter; and
causing the element to be rendered in a graphical user interface (GUI) as a time-series icon to show to a user the changes in the intensity by way of the geometric parameter.
29. The system of claim 28 , wherein the element is a line chart and wherein the geometric parameter is a y-value in the line chart reflective of the intensity of the clinically-diagnosed category.
30. The system of claim 28 , wherein causing the element to be rendered comprises serving to a client device a webpage with instructions to instruct the client device to render the element in a web browser.
31. The media of claim 20 , wherein the clinical records are electronic medical records (EMRs) stored in a patient-management system that manages EMRs for many patients and wherein the journal records are stored in an electronic journal for a single one of the many patients.
32. The media of claim 20 , wherein the clinical journal tokens comprise words, strings, and portions of words.
33. The media of claim 20 , wherein:
the clinical records and the journal records include timestamps;
identifying, for each clinical-journal token, a timestamp from the clinical records or the journal records; and
the operations further comprises associating the associated time for each clinical journal token based on one of the timestamps.
34. The media of claim 20 , wherein combining the clinical records with the journal records to produce a plurality of clinical journal tokens comprises:
generating, for each clinical-journal token, a plurality of intensity scores, each intensity score being a numeric-value on a single scale such that none of the intensity values are greater than a maximum for the single scale and such that none of the intensity values are less a minimum for the single scale; and
translating the intensity scores to translated-scores in category-specific scales, at least some of the category-specific scales having a maximum for the category-specific scale different than the maximum for the single scale.
35. The media of claim 20 , wherein generating a time-series icon by:
for at least a part of the time period, generating an element with a geometric parameter reflective of the intensity of the clinically-diagnostic category, such that changes in the intensity correspond to changes in the geometric parameter; and
causing the element to be rendered in a graphical user interface (GUI) as a time-series icon to show to a user the changes in the intensity by way of the geometric parameter.
36. The media of claim 35 , wherein the element is a line chart and wherein the geometric parameter is a y-value in the line chart reflective of the intensity of the clinically-diagnosed category.
37. The media of claim 35 , wherein causing the element to be rendered comprises serving to a client device a webpage with instructions to instruct the client device to render the element in a web browser.
38. The method of claim 1 , wherein at least one of accessing the clinical records and accessing the journal records comprises performing optical character recognition (OCR) on one or more scanned documents.
39. The method of claim 1 , wherein combining the clinical records with the journal records comprises weighing the clinical records and journal records by their type.
40. The system of claim 12 , wherein at least one of accessing the clinical records and accessing the journal records comprises performing optical character recognition (OCR) on one or more scanned documents.
41. The system of claim 12 , wherein combining the clinical records with the journal records comprises weighing the clinical records and journal records by their type.
42. The media of claim 20 , wherein combining the clinical records with the journal records comprises weighing the clinical records and journal records by their type
43. The media of claim 20 , wherein at least one of accessing the clinical records and accessing the journal records comprises performing optical character recognition (OCR) on one or more scanned documents.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US17/369,908 US20220114342A1 (en) | 2020-10-08 | 2021-07-07 | Natural language processing of electronic records |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US17/066,335 US11080484B1 (en) | 2020-10-08 | 2020-10-08 | Natural language processing of electronic records |
US17/369,908 US20220114342A1 (en) | 2020-10-08 | 2021-07-07 | Natural language processing of electronic records |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/066,335 Continuation US11080484B1 (en) | 2020-10-08 | 2020-10-08 | Natural language processing of electronic records |
Publications (1)
Publication Number | Publication Date |
---|---|
US20220114342A1 true US20220114342A1 (en) | 2022-04-14 |
Family
ID=77063320
Family Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/066,335 Active US11080484B1 (en) | 2020-10-08 | 2020-10-08 | Natural language processing of electronic records |
US17/369,908 Pending US20220114342A1 (en) | 2020-10-08 | 2021-07-07 | Natural language processing of electronic records |
Family Applications Before (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US17/066,335 Active US11080484B1 (en) | 2020-10-08 | 2020-10-08 | Natural language processing of electronic records |
Country Status (2)
Country | Link |
---|---|
US (2) | US11080484B1 (en) |
WO (1) | WO2022073058A1 (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11080484B1 (en) * | 2020-10-08 | 2021-08-03 | Omniscient Neurotechnology Pty Limited | Natural language processing of electronic records |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050176680A1 (en) * | 2003-12-11 | 2005-08-11 | Sepracor, Inc. | Combination of sedative and a neurotransmitter modulator, and methods for improving sleep quality and treating depression |
US20140201095A1 (en) * | 2013-01-14 | 2014-07-17 | Intermedhx, Llc | Healthcare assurance system |
US20190220777A1 (en) * | 2018-01-16 | 2019-07-18 | Jpmorgan Chase Bank, N.A. | System and method for implementing a client sentiment analysis tool |
US20190385711A1 (en) * | 2018-06-19 | 2019-12-19 | Ellipsis Health, Inc. | Systems and methods for mental health assessment |
US20190392936A1 (en) * | 2018-04-24 | 2019-12-26 | Arrix, Inc. | Systems and methods for medication management |
US20200090067A1 (en) * | 2018-09-13 | 2020-03-19 | International Business Machines Corporation | State of emotion time series |
US20200321125A1 (en) * | 2019-04-03 | 2020-10-08 | Crystal Christmas | Patient controlled integrated and comprehensive health record management system |
US20200402625A1 (en) * | 2019-06-21 | 2020-12-24 | nference, inc. | Systems and methods for computing with private healthcare data |
US20230241338A1 (en) * | 2020-07-08 | 2023-08-03 | Fisher & Paykel Healthcare Limited | Improvements relating to respiratory support |
Family Cites Families (49)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6006188A (en) * | 1997-03-19 | 1999-12-21 | Dendrite, Inc. | Speech signal processing for determining psychological or physiological characteristics using a knowledge base |
PL211975B1 (en) * | 2002-12-27 | 2012-07-31 | Otsuka Pharma Co Ltd | Carbostyril derivatives and serotonin reuptake inhibitors for treatment of mood disorders |
US20070142395A1 (en) * | 2005-12-16 | 2007-06-21 | Vela Pharmaceuticals, Inc. | Treatment of sexual dysfunction with 1-[3-[4-(3-chlorophenyl)-1-piperazinyl]propyl]-5-methoxy-3,4-dihydro-2(1H)-quinolone or salt thereof |
US7937345B2 (en) * | 2006-07-12 | 2011-05-03 | Kofax, Inc. | Data classification methods using machine learning techniques |
WO2009029689A1 (en) * | 2007-08-27 | 2009-03-05 | Google Inc. | Distinguishing accessories from products for ranking search results |
US8996538B1 (en) * | 2009-05-06 | 2015-03-31 | Gracenote, Inc. | Systems, methods, and apparatus for generating an audio-visual presentation using characteristics of audio, visual and symbolic media objects |
CN102385858B (en) * | 2010-08-31 | 2013-06-05 | 国际商业机器公司 | Emotional voice synthesis method and system |
US9159068B2 (en) * | 2010-10-12 | 2015-10-13 | International Business Machines Corporation | Service management using user experience metrics |
AU2012225661A1 (en) | 2011-03-07 | 2013-09-19 | Health Fidelity, Inc. | Systems and methods for processing patient history data |
US8606575B1 (en) * | 2011-09-06 | 2013-12-10 | West Corporation | Method and apparatus of providing semi-automated classifier adaptation for natural language processing |
US8463595B1 (en) * | 2012-03-06 | 2013-06-11 | Reputation.Com, Inc. | Detailed sentiment analysis |
US9009027B2 (en) * | 2012-05-30 | 2015-04-14 | Sas Institute Inc. | Computer-implemented systems and methods for mood state determination |
WO2014026152A2 (en) | 2012-08-10 | 2014-02-13 | Assurerx Health, Inc. | Systems and methods for pharmacogenomic decision support in psychiatry |
WO2015119991A1 (en) | 2014-02-05 | 2015-08-13 | 3M Innovative Properties Company | Natural language processing for medical records |
US20160350486A1 (en) | 2014-02-05 | 2016-12-01 | 3M Innovative Properties Company | Natural language processing for medical records |
US20150309987A1 (en) * | 2014-04-29 | 2015-10-29 | Google Inc. | Classification of Offensive Words |
EP2985711A1 (en) * | 2014-08-14 | 2016-02-17 | Accenture Global Services Limited | System for automated analysis of clinical text for pharmacovigilance |
US10269342B2 (en) * | 2014-10-29 | 2019-04-23 | Hand Held Products, Inc. | Method and system for recognizing speech using wildcards in an expected response |
US10684998B2 (en) * | 2014-11-21 | 2020-06-16 | Microsoft Technology Licensing, Llc | Automatic schema mismatch detection |
US10431214B2 (en) * | 2014-11-26 | 2019-10-01 | Voicebox Technologies Corporation | System and method of determining a domain and/or an action related to a natural language input |
KR102315574B1 (en) * | 2014-12-03 | 2021-10-20 | 삼성전자주식회사 | Apparatus and method for classification of data, apparatus and method for segmentation of region of interest |
US9836456B2 (en) * | 2015-01-12 | 2017-12-05 | Google Llc | Techniques for providing user image capture feedback for improved machine language translation |
WO2016156975A1 (en) * | 2015-02-17 | 2016-10-06 | La Valeriane Sas | Elder care assessment and interactive case management system |
US10380657B2 (en) * | 2015-03-04 | 2019-08-13 | International Business Machines Corporation | Rapid cognitive mobile application review |
US20160314410A1 (en) * | 2015-04-23 | 2016-10-27 | Rovi Guides, Inc. | Systems and methods for improving accuracy in media asset recommendations based on data from one data space |
US9613300B2 (en) * | 2015-06-17 | 2017-04-04 | Canon Kabushiki Kaisha | Material classification using multiview capture |
US10835174B2 (en) * | 2016-01-12 | 2020-11-17 | Covidien Lp | System and method for monitoring cerebral activity |
US20170236520A1 (en) * | 2016-02-16 | 2017-08-17 | Knuedge Incorporated | Generating Models for Text-Dependent Speaker Verification |
US10268769B2 (en) * | 2016-08-29 | 2019-04-23 | International Business Machines Corporation | Sentiment analysis |
US20180068222A1 (en) * | 2016-09-07 | 2018-03-08 | International Business Machines Corporation | System and Method of Advising Human Verification of Machine-Annotated Ground Truth - Low Entropy Focus |
US10593429B2 (en) | 2016-09-28 | 2020-03-17 | International Business Machines Corporation | Cognitive building of medical condition base cartridges based on gradings of positional statements |
US20180150882A1 (en) * | 2016-11-28 | 2018-05-31 | Mastercard International Incorporated | Systems and Methods for Use in Determining Consumer Interest in Products Based on Intensities of Facial Expressions |
US11205103B2 (en) * | 2016-12-09 | 2021-12-21 | The Research Foundation for the State University | Semisupervised autoencoder for sentiment analysis |
EP3416123A1 (en) * | 2017-06-16 | 2018-12-19 | KBC Groep NV | System for identification of fraudulent transactions |
US20200286614A1 (en) * | 2017-09-08 | 2020-09-10 | The General Hospital Corporation | A system and method for automated labeling and annotating unstructured medical datasets |
US10387572B2 (en) * | 2017-09-15 | 2019-08-20 | International Business Machines Corporation | Training data update |
US10558920B2 (en) * | 2017-10-02 | 2020-02-11 | Servicenow, Inc. | Machine learning classification with confidence thresholds |
TWI661319B (en) * | 2017-11-30 | 2019-06-01 | 財團法人資訊工業策進會 | Apparatus, method, and computer program product thereof for generatiing control instructions based on text |
JP7060434B2 (en) * | 2018-04-18 | 2022-04-26 | トヨタ自動車株式会社 | State estimator |
US10699104B2 (en) * | 2018-05-03 | 2020-06-30 | International Business Machines Corporation | Image obtaining based on emotional status |
US11756667B2 (en) * | 2018-05-30 | 2023-09-12 | Siemens Healthcare Gmbh | Decision support system for medical therapy planning |
CN113015977A (en) | 2018-06-29 | 2021-06-22 | 人工智能技术公司 | Deep learning based diagnosis and referral of diseases and conditions using natural language processing |
US20200007634A1 (en) * | 2018-06-29 | 2020-01-02 | Microsoft Technology Licensing, Llc | Cross-online vertical entity recommendations |
US20200065868A1 (en) * | 2018-08-23 | 2020-02-27 | Walmart Apollo, Llc | Systems and methods for analyzing customer feedback |
JP6739846B2 (en) * | 2018-09-03 | 2020-08-12 | MiZ株式会社 | Hydrogen-containing composition for improving depressive symptoms |
MX2021006985A (en) * | 2018-12-14 | 2021-09-08 | Praxis Prec Medicines Inc | Methods for the treatment of depression. |
US10937416B2 (en) * | 2019-02-01 | 2021-03-02 | International Business Machines Corporation | Cross-domain multi-task learning for text classification |
US11138378B2 (en) * | 2019-02-28 | 2021-10-05 | Qualtrics, Llc | Intelligently summarizing and presenting textual responses with machine learning |
US11080484B1 (en) * | 2020-10-08 | 2021-08-03 | Omniscient Neurotechnology Pty Limited | Natural language processing of electronic records |
-
2020
- 2020-10-08 US US17/066,335 patent/US11080484B1/en active Active
-
2021
- 2021-07-07 US US17/369,908 patent/US20220114342A1/en active Pending
- 2021-09-20 WO PCT/AU2021/051092 patent/WO2022073058A1/en active Application Filing
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050176680A1 (en) * | 2003-12-11 | 2005-08-11 | Sepracor, Inc. | Combination of sedative and a neurotransmitter modulator, and methods for improving sleep quality and treating depression |
US20140201095A1 (en) * | 2013-01-14 | 2014-07-17 | Intermedhx, Llc | Healthcare assurance system |
US20190220777A1 (en) * | 2018-01-16 | 2019-07-18 | Jpmorgan Chase Bank, N.A. | System and method for implementing a client sentiment analysis tool |
US20190392936A1 (en) * | 2018-04-24 | 2019-12-26 | Arrix, Inc. | Systems and methods for medication management |
US20190385711A1 (en) * | 2018-06-19 | 2019-12-19 | Ellipsis Health, Inc. | Systems and methods for mental health assessment |
US20200090067A1 (en) * | 2018-09-13 | 2020-03-19 | International Business Machines Corporation | State of emotion time series |
US20200321125A1 (en) * | 2019-04-03 | 2020-10-08 | Crystal Christmas | Patient controlled integrated and comprehensive health record management system |
US20200402625A1 (en) * | 2019-06-21 | 2020-12-24 | nference, inc. | Systems and methods for computing with private healthcare data |
US20230241338A1 (en) * | 2020-07-08 | 2023-08-03 | Fisher & Paykel Healthcare Limited | Improvements relating to respiratory support |
Also Published As
Publication number | Publication date |
---|---|
US11080484B1 (en) | 2021-08-03 |
WO2022073058A1 (en) | 2022-04-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Jackson et al. | From text to thought: How analyzing language can advance psychological science | |
Fontaine et al. | The world of emotions is not two-dimensional | |
Denecke et al. | Sentiment analysis in medical settings: New opportunities and challenges | |
Nelson et al. | Universality revisited | |
Figueroa et al. | Active learning for clinical text classification: is it better than random sampling? | |
Kaggal et al. | Toward a learning health-care system–knowledge delivery at the point of care empowered by big data and NLP | |
Schuller et al. | A review on five recent and near-future developments in computational processing of emotion in the human voice | |
Dehghani et al. | TACIT: An open-source text analysis, crawling, and interpretation tool | |
Taikh et al. | Semantic classification of pictures and words | |
Ashby et al. | The early processing of metaphors and similes: Evidence from eye movements | |
Al-Bahrani et al. | Survivability prediction of colon cancer patients using neural networks | |
D’mello et al. | Psychological measurement in the information age: Machine-learned computational models | |
Henry et al. | Cue coalitions and additivity in predictive processing: The interaction between case and prosody in L2 German | |
Sawalha et al. | Detecting presence of PTSD using sentiment analysis from text data | |
Liu et al. | Semiautomatic annotation of mooc forum posts | |
Sauter | Is there a role for language in emotion perception? | |
Constâncio et al. | Deception detection with machine learning: A systematic review and statistical analysis | |
Vandenbussche et al. | Using natural language processing to automatically classify written self-reported narratives by patients with migraine or cluster headache | |
Denecke et al. | Sentiment analysis from medical texts | |
Pang et al. | The factor structure of the COPE questionnaire in a sample of clinically depressed and anxious adults | |
Németh et al. | Bio, psycho, or social: supervised machine learning to classify discursive framing of depression in online health communities | |
Roy et al. | There’s so much to do and not enough time to do it! A case for sentiment analysis to derive meaning from open text using student reflections of engineering activities | |
US20220114342A1 (en) | Natural language processing of electronic records | |
Xu et al. | Suicide note sentiment classification: a supervised approach augmented by web data | |
Bartolotti et al. | Listening to speech and non-speech sounds activates phonological and semantic knowledge differently |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
AS | Assignment |
Owner name: OMNISCIENT NEUROTECHNOLOGY PTY LIMITED, AUSTRALIA Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SUGHRUE, MICHAEL EDWARD;DOYEN, STEPHANE PHILIPPE;NICHOLAS, PETER JAMES;REEL/FRAME:058780/0844 Effective date: 20201012 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |