Welcome to the Lexigram documentation site. Below you'll find our API documentation and a few interactive components that allow you to play directly with our data endpoints (no code required) before obtaining your very own API Key and unleashing the power of Lexigram on your data! Please direct any questions or requests to support@lexigram.io.
\nBelow you'll find our API documentation and a few interactive components that allow you to interact directly with our data endpoints (no code required). To try it out, get an API key from the Get Access link in the menu. Please direct any questions or requests to support@lexigram.io.
\nYou can use your API Key (an encrypted JWT), available from https://app.lexigram.io, to obtain an expiring token that can be used to authenticate your user with the API.
\nTo obtain the token, make an empty GET request to the /auth/token using the Authorization header with a value that is set to Bearer {your_api_key}. In the response, the expiring token will be in the token property. You can then use that in the Authorization header until it expires and you have to request another token.
The matched concepts returned by the extraction endpoints are grouped according to the following main types:
\nFor each of the main classification types, there are different sublevels. For instance, Glaucoma has main classification Disease and subclassification Eye Diseases.
\nIt is possible to have multiple subclassifications depending on how many hierarchy branches connect into a concept. For instance, Tachycardia has main classification Disease and subclassifications Cardiovascular Diseases and Symptom. For the Disease main classification, one can expect to see as many subclassifications as medical specialties and/or medical systems.
\nThe Drug main classification is similarly subclassificasified into active ingredients of drugs, according to the organ or system on which they act. For instance, the drug Tegretol is subclassified as Nervous System. Also, multiple subclassifications can be applied to drugs. For instance: Diclofenac Potassium is subclassified as Sensory Organs, Musculoskeletal and Dermatological.
\nProcedure is subclassified as: Diagnosis, Therapeutics, Anesthesia, Surgical, Investigative and Dentistry. For instance, Blood Transfusion is Therapeutics and Auscultation is Diagnosis.
\nAnatomy is subclassified with each anatomical system plus additional categories for things like Fluids, Cells, Tissues and Viruses and Bacterias. The Anatomy can have multiple subclassifications, for instance Oropharynx has Stomatognathic System and Respiratory System.
\nMental Health is subclassified with Mental Disorders, Behaviors, Psychological Phenomena and Disciplines. Psychological Phenomena are concepts like Sleep, Hearing or Appetite. Behaviours include concepts like Anxiety, Affect, Depression and Feelings. Mental Disorders encompasses the psychiatric illnesses manifested by breakdowns in the adaptational process. Examples of Mental Disorders are: Depressive Disorder, Dementia or Alcoholism.
\nThe “Other” classification are other related medical concepts that cannot be classified in the the aforementioned main classification categories. For example: Office Visit is classified as Other with subclassifications Organization and Administration.
\nNotes gathered from an EHR will frequently be divided into distinct sections. These sections can help to provide additional information about the concepts that they contain. For example, your application may want to increase or decrease the relevance of matches found in certain sections.
\nSections are provided in a top-level attribute on the extraction response, titled sections. Using the information in each section object, you can align matched concepts to their corresponding section.
For the input text Chief Complaint: worsening chest pain and shortness of breath, we would see the following section extracted:
\"sections\": [\n {\n \"id\": \"chief-complaint\",\n \"label\": \"chief complaint\",\n \"type\": \"SECTION\",\n \"sectionId\": 0,\n \"explanation\": {\n \"begin\": 0,\n \"end\": 15,\n \"matchedTokens\": [\n {\n \"token\": \"Chief\",\n \"position\": 0\n },\n {\n \"token\": \"Complaint\",\n \"position\": 2\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": []\n }\n ]\n Using the section.explanation.end position, we can see that any concept match that has concept.explanation.begin greater than 15 is contained within this section.
For documents with multiple sections, the begin and end boundary for each individual section can be determined by using the explanation.end information from the target section and the explanation.begin information from the next section.
Patient charts frequently contain data presented as lists, including problem, medication, medical history, and complaint lists. Lexigram examines text provided to the API in order to expose these lists, and surfaces information about the list itself, its position, contents, and matched concepts.
\nLists are a beta feature and are not enabled by default. They can be enabled by passing the parameter with-lists set to true for GET requests. For a POST, include \"withLists\": true in the JSON request body.
Information about lists is returned on a top-level lists attribute. For the example text PATIENT CONDITIONS 1) Hypertension 2) ADHD 3) Diabetes, the response would include:
\"lists\": [\n {\n \"id\": 0,\n \"explanation\": {\n \"begin\": 22,\n \"end\": 54,\n \"itemCount\": 3\n },\n \"items\": [\n {\n \"number\": 1,\n \"label\": \"1)\",\n \"explanation\": {\n \"begin\": 22,\n \"end\": 34,\n \"matchesIndex\": [\n 0\n ]\n }\n },\n {\n \"number\": 2,\n \"label\": \"2)\",\n \"explanation\": {\n \"begin\": 38,\n \"end\": 42,\n \"matchesIndex\": [\n 1\n ]\n }\n },\n {\n \"number\": 3,\n \"label\": \"3)\",\n \"explanation\": {\n \"begin\": 46,\n \"end\": 54,\n \"matchesIndex\": [\n 2\n ]\n }\n }\n ]\n }\n]\nAs with Sections, the overlapping position of the list item begin and end can be used to determine which concept matches fall within which list item. For example, this match is returned for the example text above:
\n{\n \"id\": \"lxg-3ac43c43af84\",\n \"label\": \"Hypertension\",\n \"type\": \"ENTITY\",\n \"explanation\": {\n \"begin\": 22,\n \"end\": 34,\n \"matchedTokens\": [\n {\n \"token\": \"Hypertension\",\n \"position\": 6\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": [],\n \"classification\": {\n \"main\": \"Disease\",\n \"subs\": [\n {\n \"id\": \"lxg-7f69fbb334e4\",\n \"label\": \"Vascular Diseases\"\n },\n {\n \"id\": \"lxg-0eeec8ac29e4\",\n \"label\": \"Cardiovascular Diseases\"\n }\n ]\n }\n}\nWe can use the match begin and end to determine that this match falls within the list item with number equal to one.
The Lexigram API will extract and normalize patient demographics, which include gender, age, ethnicity, and race. These demographics are frequently included in reporting and are used in calculating risk stratification and other patient data calculation algorithms.
\nThese demographics frequently appear clustered in charts, and to avoid ambiguity, the Lexigram API will look for these clusters when extracting demographic data. For example, the statement The patient is a 55 year old Caucasian male, would result in the following concept matches:
{\n \"matches\": [\n {\n \"id\": \"age\",\n \"label\": \"age\",\n \"type\": \"DEMOGRAPHIC\",\n \"explanation\": {\n \"begin\": 17,\n \"end\": 28,\n \"matchedTokens\": [\n {\n \"token\": \"55\",\n \"position\": 8\n },\n {\n \"token\": \"year\",\n \"position\": 10\n },\n {\n \"token\": \"old\",\n \"position\": 12\n }\n ],\n \"subtype\": \"age\",\n \"value\": \"55\",\n \"fuzzyValues\": []\n },\n \"contexts\": []\n },\n {\n \"id\": \"white\",\n \"label\": \"white\",\n \"type\": \"DEMOGRAPHIC\",\n \"explanation\": {\n \"begin\": 29,\n \"end\": 38,\n \"matchedTokens\": [\n {\n \"token\": \"Caucasian\",\n \"position\": 14\n }\n ],\n \"subtype\": \"race\",\n \"value\": \"white\",\n \"fuzzyValues\": []\n },\n \"contexts\": []\n },\n {\n \"id\": \"male\",\n \"label\": \"male\",\n \"type\": \"DEMOGRAPHIC\",\n \"explanation\": {\n \"begin\": 39,\n \"end\": 43,\n \"matchedTokens\": [\n {\n \"token\": \"male\",\n \"position\": 16\n }\n ],\n \"subtype\": \"gender\",\n \"value\": \"male\",\n \"fuzzyValues\": []\n },\n \"contexts\": []\n }\n ]\n}\nDates in patient charts are used to track the onset of conditions, periods of medication usage, and other information about patient conditions. The API will extract dates in a variety of formats that are presented in the text.
\nFor example, the following statements will all result in a date extraction that has been normalized to 2018-11-13:
The date extraction appears as below in the concept matches:
\n{\n \"id\": \"date\",\n \"label\": \"date\",\n \"type\": \"DATE\",\n \"explanation\": {\n \"begin\": 42,\n \"end\": 54,\n \"matchedTokens\": [\n {\n \"token\": \"Nov\",\n \"position\": 12\n },\n {\n \"token\": \"13\",\n \"position\": 14\n },\n {\n \"token\": \",\",\n \"position\": 15\n },\n {\n \"token\": \"2018\",\n \"position\": 17\n }\n ],\n \"subtype\": \"date\",\n \"value\": \"2018-11-13\",\n \"fuzzyValues\": []\n },\n \"contexts\": []\n}\nDated problem and medication lists are also good areas to surface information on a patient's medical history, and can provide insight into the progression of a condition of interest.
\nUnfortunately, the appearance of dates in chart text, especially for charts that have been faxed, scanned, or otherwise processed via OCR, is incredibly ambiguous. Frequently, the date a chart was printed or faxed will appear alongside the date of a visit or the start of a prescription. Dates of physician attestations, physician and nurse's note sign-offs, and many other dates that are not relevant to patient conditions can appear in the data.
\nBecause of this, we caution users to be mindful of how they are utilizing the date extractions we present via the API. We are happy to discuss issues related to date extraction and how to best situate these extractions in your application in more detail, please contact support@lexigram.io to arrange a call.
\nConcept matches are enriched with a context object that contain additional information of the \"surroundings\" for a particular extraction, e.g. no evidence of pneumonia is considered a negated definite statement and the extraction match will contain the following context object:
\"contexts\": [\n {\n \"type\": \"negative\",\n \"subtype\": \"definite\",\n \"explanation\": {\n \"begin\": 0,\n \"end\": 14,\n \"matchedTokens\": [\n {\n \"token\": \"no\",\n \"position\": 0\n },\n {\n \"token\": \"evidence\",\n \"position\": 2\n },\n {\n \"token\": \"of\",\n \"position\": 4\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"N0000057\",\n \"triggerLabel\": \"no evidence\"\n }\n }\n]\nThe two main data attributes in that object are type and subtype which contain the context classification. Use type and subtype if your applications only care about blacklisting or whitelisting certain types of context, or if you are using the context as a feature for a downstream machine learning process. The other attributes in that object represent the text provenance of where the context trigger was found.
The list of contexts with examples are described in the sections below.
\nPhysicians use specific language to discuss the conditions that a patient may or may not have. The following are categories of negation and hypothetical contexts that illustrate the variety of ways in which these conditions are documented.
\n{ \"type\" : \"affirmed\", \"subtype\": \"definite\"}\n{ \"type\" : \"negative\", \"subtype\": \"definite\"}\n{ \"type\" : \"speculative\", \"subtype\": \"affirmed\"}\n{ \"type\" : \"speculative\", \"subtype\": \"negative\"}\n{ \"type\" : \"speculative\", \"subtype\": \"uncertain\"}\n{ \"type\" : \"speculative\", \"subtype\": \"future\"}\n{ \"type\" : \"speculative\", \"subtype\": \"indication\"}\nThe API will extract vitals and lab values, making it simple to track weight, height, BMI, blood pressure readings, HbA1c, blood glucose, metabolic panels, lipid panels, and more.
\nWhen doing these extractions, the API will align the measurement type (height), the value (180), and the unit (centimeters). For this example, using the text Height 180 cm will result in the following extraction:
{\n \"matches\": [\n {\n \"id\": \"height\",\n \"label\": \"height\",\n \"type\": \"MEASUREMENT\",\n \"explanation\": {\n \"begin\": 0,\n \"end\": 10,\n \"matchedTokens\": [\n {\n \"token\": \"Height\",\n \"position\": 0\n },\n {\n \"token\": \"180\",\n \"position\": 2\n }\n ],\n \"subtype\": \"height\",\n \"value\": \"180.0\",\n \"unit\": \"centimeter\",\n \"fuzzyValues\": []\n },\n \"contexts\": []\n }\n ]\n}\nWhen working with charts that have gone through an OCR process, it is common to find several numeric values clustered in a section describing vitals, without clear positioning to indicate which value corresponds with which measurement type. In these cases, the API will return a list of values in the fuzzyValues attribute, allowing the user to further refine these values as needed.
Where available, the API will extract detailed information about medication from patient charts. These are extracted as distinct contexts for a given drug and includes type (branded vs generic), formulation, ingredients, dosage strength, route, and form.
\nMedications that contain multiple ingredients will include each ingredient, along with the ingredient strength for the given formulation, provided that a strength is present and can be extracted.
\nFor a generic drug, this is the information that is extracted (input: Lisinopril 20 mg oral tab):
{\n \"matches\": [\n {\n \"id\": \"lxg-41490a0f260b\",\n \"label\": \"Lisinopril\",\n \"type\": \"DRUG\",\n \"explanation\": {\n \"begin\": 0,\n \"end\": 10,\n \"matchedTokens\": [\n {\n \"token\": \"Lisinopril\",\n \"position\": 0\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": [\n {\n \"type\": \"medication\",\n \"subtype\": \"form\",\n \"explanation\": {\n \"begin\": 22,\n \"end\": 25,\n \"matchedTokens\": [\n {\n \"token\": \"tab\",\n \"position\": 8\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"385055001\",\n \"triggerLabel\": \"tablet\"\n }\n },\n {\n \"type\": \"medication\",\n \"subtype\": \"ingredients\",\n \"ingredients\": [\n {\n \"name\": \"Lisinopril\",\n \"id\": \"lxg-41490a0f260b\",\n \"strength\": {\n \"value\": 20,\n \"unit\": \"milligram\"\n }\n }\n ]\n },\n {\n \"type\": \"medication\",\n \"subtype\": \"route\",\n \"explanation\": {\n \"begin\": 17,\n \"end\": 21,\n \"matchedTokens\": [\n {\n \"token\": \"oral\",\n \"position\": 6\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"26643006\",\n \"triggerLabel\": \"oral\"\n }\n },\n {\n \"type\": \"medication\",\n \"subtype\": \"generic\",\n \"explanation\": {\n \"begin\": 0,\n \"end\": 10,\n \"matchedTokens\": [\n {\n \"token\": \"Lisinopril\",\n \"position\": 0\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"lxg-41490a0f260b\",\n \"triggerLabel\": \"Lisinopril\"\n }\n }\n ],\n \"classification\": {\n \"main\": \"Drug\",\n \"subs\": [\n {\n \"id\": \"lxg-a4ede3447f55\",\n \"label\": \"AGENTS ACTING ON THE RENIN-ANGIOTENSIN SYSTEM\"\n },\n {\n \"id\": \"lxg-8b7dbf4bc1d9\",\n \"label\": \"ACE INHIBITORS, PLAIN\"\n },\n {\n \"id\": \"lxg-04b05fd06b05\",\n \"label\": \"CARDIOVASCULAR SYSTEM\"\n }\n ]\n }\n }\n ]\n}\nFor a branded drug, this is the information that is extracted (input: HYDROcodone/acetaminophen (NORCO 10) 10mg/325mg Tab 60 Tab 0):
{\n \"matches\": [\n {\n \"id\": \"lxg-322820454fbc\",\n \"label\": \"Norco\",\n \"type\": \"DRUG\",\n \"explanation\": {\n \"begin\": 27,\n \"end\": 32,\n \"matchedTokens\": [\n {\n \"token\": \"NORCO\",\n \"position\": 3\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": [\n {\n \"type\": \"medication\",\n \"subtype\": \"form\",\n \"explanation\": {\n \"begin\": 48,\n \"end\": 51,\n \"matchedTokens\": [\n {\n \"token\": \"Tab\",\n \"position\": 9\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"385055001\",\n \"triggerLabel\": \"tablet\"\n }\n },\n {\n \"type\": \"medication\",\n \"subtype\": \"ingredients\",\n \"ingredients\": [\n {\n \"name\": \"Acetaminophen 325 MG\",\n \"id\": \"lxg-a499f0cc3e75\",\n \"strength\": {\n \"value\": 325,\n \"unit\": \"milligram\"\n }\n },\n {\n \"name\": \"Hydrocodone Bitartrate 10 MG\",\n \"id\": \"lxg-a96f583b72f5\",\n \"strength\": {\n \"value\": 10,\n \"unit\": \"milligram\"\n }\n }\n ]\n },\n {\n \"type\": \"medication\",\n \"subtype\": \"brand\",\n \"explanation\": {\n \"begin\": 27,\n \"end\": 32,\n \"matchedTokens\": [\n {\n \"token\": \"NORCO\",\n \"position\": 3\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"lxg-322820454fbc\",\n \"triggerLabel\": \"Norco\"\n }\n }\n ],\n \"classification\": {\n \"main\": \"Drug\",\n \"subs\": [\n {\n \"id\": \"lxg-261ecc3eb835\",\n \"label\": \"ANALGESICS\"\n },\n {\n \"id\": \"lxg-f9460f837191\",\n \"label\": \"COUGH SUPPRESSANTS, EXCL. COMBINATIONS WITH EXPECTORANTS\"\n },\n {\n \"id\": \"lxg-e0fa575f13a0\",\n \"label\": \"NERVOUS SYSTEM\"\n },\n {\n \"id\": \"lxg-493406c1bb02\",\n \"label\": \"RESPIRATORY SYSTEM\"\n },\n {\n \"id\": \"lxg-b165765b7c80\",\n \"label\": \"COUGH AND COLD PREPARATIONS\"\n },\n {\n \"id\": \"lxg-7240c6128ed3\",\n \"label\": \"OTHER ANALGESICS AND ANTIPYRETICS\"\n }\n ]\n }\n }\n ]\n}\nMany diseases or procedures occur in the context of a particular body part or system. Where those occur in the text, the API will present the anatomy part as a context of the disease.
\nFor example, the statement The patient has a skin abscess will result in the concept match abscess contextualized with the anatomy concept skin:
{\n \"id\": \"lxg-88b88605857f\",\n \"label\": \"Abscess\",\n \"type\": \"ENTITY\",\n \"explanation\": {\n \"begin\": 23,\n \"end\": 30,\n \"matchedTokens\": [\n {\n \"token\": \"abscess\",\n \"position\": 10\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": [\n {\n \"type\": \"anatomy\",\n \"subtype\": \"anatomy\",\n \"explanation\": {\n \"begin\": 18,\n \"end\": 22,\n \"matchedTokens\": [\n {\n \"token\": \"skin\",\n \"position\": 8\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"anatomy-skin\",\n \"triggerLabel\": \"skin\"\n }\n }\n ],\n \"classification\": {\n \"main\": \"Disease\",\n \"subs\": [\n {\n \"id\": \"lxg-ecb807e05bc2\",\n \"label\": \"Infection\"\n },\n {\n \"id\": \"lxg-4ebfcb6de390\",\n \"label\": \"Pathological Conditions, Signs and Symptoms\"\n },\n {\n \"id\": \"lxg-7a3777afaaa9\",\n \"label\": \"Pathologic Processes\"\n },\n {\n \"id\": \"lxg-ae7931d19353\",\n \"label\": \"Suppuration\"\n },\n {\n \"id\": \"lxg-8187cb28c903\",\n \"label\": \"Inflammation\"\n },\n {\n \"id\": \"lxg-16235b458c6e\",\n \"label\": \"Bacterial Infections and Mycoses\"\n }\n ]\n }\n}\nWhen working with a patient, physicians will capture a patient and family history, which can help provide insight to the patient's current status. When extracting conditions from chart data, it is important to avoid attributing conditions that exist in a patient's family history to the patient themselves.
\nThe following categories allow users to determine the distinct historical context being used in physical documentation.
\n{ \"type\" : \"history\", \"subtype\": \"patient\"}\nIt also propagates section context to entities in close proximity to the header:
\nPATIENT HISTORY:\n\n- Alcoholism\n- Depression\n{ \"type\" : \"history\", \"subtype\": \"family\"}\nAnd it also propagates with section headers:
\nFAMILY HISTORY:\n\n- Depression\n- Breast Cancer\n- Coronary artery disease\nKnowing a patient's allergies can be an important part of ensuring patient safety. The API will surface allergies to known substances by providing an allergy context.
\nExample for input Patient has allergy to penicillin:
{\n \"matches\": [\n {\n \"id\": \"lxg-f8833f946643\",\n \"label\": \"Penicillin\",\n \"type\": \"ENTITY\",\n \"explanation\": {\n \"begin\": 23,\n \"end\": 33,\n \"matchedTokens\": [\n {\n \"token\": \"penicillin\",\n \"position\": 8\n }\n ],\n \"fuzzyValues\": []\n },\n \"contexts\": [\n {\n \"type\": \"allergy\",\n \"subtype\": \"allergy\",\n \"explanation\": {\n \"begin\": 12,\n \"end\": 22,\n \"matchedTokens\": [\n {\n \"token\": \"allergy\",\n \"position\": 4\n },\n {\n \"token\": \"to\",\n \"position\": 6\n }\n ],\n \"fuzzyValues\": [],\n \"triggerId\": \"allergy-forward\",\n \"triggerLabel\": \"allergy to\"\n }\n }\n ]\n }\n ]\n}\nClinical concepts are frequently mentioned in the name of medical establishments, for example \"The Diabetes Center of Dallas\" or \"Heart Wellness Center\". The API will contextualize these with type clinic-name and negative. Note: this feature is experimental and needs to be enabled by passing withClinicNames: true in the JSON body.
{\n \"matches\": [\n {\n \"id\": \"lxg-45528ef7190c\",\n \"label\": \"Heart\",\n \"type\": \"ENTITY\",\n \"explanation\": {\n \"fuzzyValues\": [],\n \"begin\": 0,\n \"end\": 5,\n \"matchedTokens\": [\n {\n \"token\": \"Heart\",\n \"position\": 0\n }\n ]\n },\n \"contexts\": [\n {\n \"type\": \"clinic-name\",\n \"subtype\": \"possible-clinic-name\",\n \"explanation\": {\n \"fuzzyValues\": [],\n \"begin\": -1,\n \"end\": 30,\n \"matchedTokens\": [\n {\n \"token\": \"internal-contextualizer-context\",\n \"position\": -1\n }\n ],\n \"triggerId\": \"cn-ctxt\",\n \"triggerLabel\": \"cn-ctxt\"\n }\n },\n {\n \"type\": \"negative\",\n \"subtype\": \"possible-clinic-name\",\n \"explanation\": {\n \"fuzzyValues\": [],\n \"begin\": -1,\n \"end\": 30,\n \"matchedTokens\": [\n {\n \"token\": \"internal-contextualizer-context\",\n \"position\": -1\n }\n ],\n \"triggerId\": \"cn-ctxt\",\n \"triggerLabel\": \"cn-ctxt\"\n }\n }\n ],\n \"classification\": {\n \"main\": \"Anatomy\",\n \"subs\": [\n {\n \"id\": \"lxg-032b4d0893da\",\n \"label\": \"Anatomy\"\n },\n {\n \"id\": \"lxg-180eae181bfe\",\n \"label\": \"Cardiovascular System\"\n }\n ]\n }\n }\n ],\n \"codes\": []\n}\nWhen described in narrative, conditions can have a variety of modifiers that provide additional information about the condition. Sometimes modifiers are substantial enough that they are considered a distinct condition. For example, the modifier pre when used with diabetic is a concept prediabetic, however pre used with hypertension is surfaced as the condition hypertension with a context pre.
The following modifiers are tracked:
\nchronic\nacute\nsub-acute\ntemporary\nbenign\nmalignant\nsomatic\nvisceral\nunknown\ninfantile\nadult\nchildhood\nfamilial\nessential\nprimary\nsecondary\nlocalised\nborderline\nchronic active\nchronic persistent\ndouble\nfloating\nsymmetric\nasymmetric\nsevere\npre\npost\nUsing the Authentication approach described below, you can access the following endpoints. To see specific requests and responses that illustrate the different features of the API, select an Example from the drop-down below.
\n","schema":"https://schema.getpostman.com/json/collection/v2.0.0/collection.json","toc":[{"content":"Quick Start","slug":"quick-start"},{"content":"Concept Classification","slug":"concept-classification"},{"content":"Sections","slug":"sections"},{"content":"Lists","slug":"lists"},{"content":"Demographics","slug":"demographics"},{"content":"Dates","slug":"dates"},{"content":"Contexts Explained","slug":"contexts-explained"},{"content":"Negation and Hypothetical","slug":"negation-and-hypothetical"},{"content":"Vitals and Lab Values","slug":"vitals-and-lab-values"},{"content":"Medication","slug":"medication"},{"content":"Anatomy","slug":"anatomy"},{"content":"History","slug":"history"},{"content":"Allergies","slug":"allergies"},{"content":"Clinic Names","slug":"clinic-names"},{"content":"Modifiers","slug":"modifiers"},{"content":"Endpoint Usage","slug":"endpoint-usage"}],"owner":"8682281","collectionId":"955b9476-b925-427d-ab32-dc2ae6d8a23d","publishedId":"Tzz4QKEK","public":true,"customColor":{"top-bar":"5a55a5","right-sidebar":"303030","highlight":"5a55a5"},"publishDate":"2022-01-10T20:00:13.000Z"},"item":[{"name":"Extract Entities","id":"ee8d9d95-50d5-4fbf-8833-d80226b9c8b0","request":{"method":"POST","header":[{"key":"Content-Type","value":"application/json"}],"body":{"mode":"raw","raw":"{\n \"text\": \"The patient has diabetes\",\n \"withContext\": true,\n \"withMatchLogic\": \"ignore-length\",\n \"withText\": true\n}"},"url":"https://api.lexigram.io/v4/extract/entities","description":"The core of our data extraction API. Matches the input text against concepts in the Lexigraph to find and extract concept matches. Maintains pointers to the exact character offsets matched in the text. Identifies the types of the concept (drug, anatomy, problem, etc.), its context (negation, speculation, patient/family history, demographics), and a sectionId if it appears in a known section.
\n","auth":{"type":"bearer","bearer":{"token":"Returns HTML enriched versions of the input text surrounding found concepts in span tags to aid visualization of the data.
\n","auth":{"type":"bearer","bearer":{"token":"