{ "resourceType" : "CodeSystem", "id" : "AIdeviceTypeCS", "meta" : { "security" : [{ "system" : "http://terminology.hl7.org/CodeSystem/v3-ObservationValue", "code" : "AIAST", "display" : "Artificial Intelligence asserted" }] }, "language" : "en", "text" : { "status" : "generated", "div" : "

Generated Narrative: CodeSystem AIdeviceTypeCS

Security Label: Artificial Intelligence asserted (Details: ObservationValue code AIAST = 'Artificial Intelligence asserted')

Properties

This code system defines the following properties for its concepts

NameCodeURIType
Not Selectableabstracthttp://hl7.org/fhir/concept-properties#notSelectableboolean

Concepts

This case-sensitive code system http://hl7.org/fhir/uv/aitransparency/CodeSystem/AIdeviceTypeCS defines the following codes in a Grouped By hierarchy:

LvlCodeDisplayDefinitionNot Selectable
1Artificial-Intelligence All kinds of Artificial IntelligenceAny type of Artificial Intelligence system, undifferentiated.
2\u00a0\u00a0AI-By-Model Classification by Technical ModelThis category classifies AI systems based on the underlying technical models they employ.true
3\u00a0\u00a0\u00a0\u00a0Rule-Based-Systems Rule-Based SystemsSystems that apply explicitly authored rules, logic, knowledge representations, or inference procedures to input data.
3\u00a0\u00a0\u00a0\u00a0Machine-Learning-Models Machine Learning ModelsThey include supervised learning models (e.g., Support Vector Machines (SVM), Random Forests (RF)), which can be used for disease classification and risk prediction; unsupervised learning models (e.g., K-means clustering), which can discover hidden characteristics of patient subgroups; and reinforcement learning models, which can be applied in dynamic treatment plan management.
4\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Deep-Learning-Models Deep Learning ModelsExamples include Convolutional Neural Networks (CNNs), which perform excellently in medical image analysis; Recurrent Neural Networks (RNNs) and their variant LSTMs, which are suitable for processing time-series physiological signal data; Generative Adversarial Networks (GANs), which can be used to synthesize training data and alleviate the scarcity of medical data; and Transformer models, which are widely used in multiple tasks such as medical imaging, text analysis, and physiological signal prediction.
5\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Foundation-Models Foundation ModelsDeep learning models trained on broad data at scale that can be adapted to a range of downstream tasks.true
6\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Large-Language-Models Large Language ModelsThese models, such as GPT-4 and PaLM, are trained on massive text datasets and can perform various natural language processing tasks, including medical text understanding, generation, and question answering.
6\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Vision-Foundation-Models Vision Foundation ModelsFoundation models trained primarily to process and represent visual data for adaptation to vision tasks.
6\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0Vision-Language-Models Vision-Language ModelsFoundation models that jointly process visual and language data.
3\u00a0\u00a0\u00a0\u00a0Hybrid-Models Hybrid ModelsThese models combine multiple AI techniques to leverage their respective strengths. For instance, combining CNNs and RNNs can effectively process medical image sequences; integrating machine learning and deep learning models can enhance disease prediction accuracy; and combining rule-based systems with machine learning can improve interpretability and reliability in clinical decision support.
2\u00a0\u00a0AI-By-Behavior Classification by Output BehaviorGroups AI systems by whether identical inputs and execution conditions are expected to produce identical outputs.true
3\u00a0\u00a0\u00a0\u00a0Deterministic-AI Deterministic AIAn AI system that produces the same output for the same input when its model, configuration, state, and execution environment are unchanged.
3\u00a0\u00a0\u00a0\u00a0Non-Deterministic-AI Non-deterministic AIAn AI system whose output may vary for the same input and execution conditions, for example because it uses probabilistic sampling or stochastic processing.
2\u00a0\u00a0AI-By-Scenario Classification by Application ScenarioThis category classifies AI systems based on their application scenarios in the medical field.true
3\u00a0\u00a0\u00a0\u00a0Intelligent-Diagnosis-and-Treatment Intelligent Diagnosis and TreatmentBy analyzing massive volumes of medical data, these AI systems assist doctors in making more accurate diagnostic and treatment decisions.
3\u00a0\u00a0\u00a0\u00a0Medical-Image-Analysis Medical Image AnalysisLeveraging deep learning technologies, these AI tools automatically identify lesion areas in medical images.
3\u00a0\u00a0\u00a0\u00a0Personalized-Treatment Personalized TreatmentThese AI systems create precise patient profiles to formulate personalized treatment plans.
3\u00a0\u00a0\u00a0\u00a0Clinical-Monitoring-and-Early-Warning Clinical Monitoring and Early WarningMonitors longitudinal or real-time patient data to identify deterioration, adverse events, or other conditions requiring attention.
3\u00a0\u00a0\u00a0\u00a0Clinical-Documentation-and-Workflow Clinical Documentation and WorkflowCreates, summarizes, extracts, codes, routes, or quality-checks clinical and administrative information to support healthcare workflows.
3\u00a0\u00a0\u00a0\u00a0Drug-Discovery-and-Development Drug Discovery and DevelopmentAI in this category accelerates the screening of candidate drugs and optimizes the design of clinical trials.
3\u00a0\u00a0\u00a0\u00a0Medical-Quality-Control Medical Quality ControlThese AI tools are used to generate standardized medical document templates and detect defects in medical documents and images.
3\u00a0\u00a0\u00a0\u00a0Patient-Services Patient ServicesAI systems here provide patients with services such as intelligent medical guidance, symptom self-assessment, and medical consultation.
3\u00a0\u00a0\u00a0\u00a0Population-Health-and-Operations Population Health and Healthcare OperationsSupports population stratification, public health, resource planning, scheduling, logistics, or other healthcare operational activities.
2\u00a0\u00a0AI-By-DataType Classification by Processed Data TypeThis category classifies AI systems based on the types of medical data they primarily process.true
3\u00a0\u00a0\u00a0\u00a0AI-for-Medical-Imaging-Data AI for Medical Imaging DataIt mainly processes medical imaging data such as X-rays, MRIs, and CT scans.
3\u00a0\u00a0\u00a0\u00a0AI-for-Physiological-Signal-Data AI for Physiological Signal DataThis type of AI deals with physiological signal data like electrocardiograms (ECG) and electroencephalograms (EEG).
3\u00a0\u00a0\u00a0\u00a0AI-for-Medical-Text-Data AI for Medical Text DataIt processes text data such as electronic health records (EHRs) and medical abstracts.
3\u00a0\u00a0\u00a0\u00a0AI-for-Structured-Clinical-Data Structured Clinical DataProcesses coded, tabular, or relational health data such as diagnoses, medications, laboratory results, observations, claims, or FHIR resources.
3\u00a0\u00a0\u00a0\u00a0AI-for-Genomic-and-Omics-Data Genomic and Omics DataProcesses genomic, transcriptomic, proteomic, metabolomic, or related molecular data.
3\u00a0\u00a0\u00a0\u00a0AI-for-Audio-Data Audio DataProcesses speech, heart sounds, respiratory sounds, or other clinically relevant audio.
3\u00a0\u00a0\u00a0\u00a0AI-for-Video-Data Video DataProcesses temporal image sequences such as endoscopy, ultrasound cine loops, gait recordings, or procedure video.
3\u00a0\u00a0\u00a0\u00a0AI-for-Multimodal-Data Multimodal DataJointly processes two or more data modalities, such as images and text or physiological signals and structured clinical data.
" }, "extension" : [{ "url" : "http://hl7.org/fhir/StructureDefinition/structuredefinition-wg", "valueCode" : "ehr" }, { "url" : "http://hl7.org/fhir/StructureDefinition/structuredefinition-fmm", "valueInteger" : 2, "_valueInteger" : { "extension" : [{ "url" : "http://hl7.org/fhir/StructureDefinition/structuredefinition-conformance-derivedFrom", "valueCanonical" : "http://hl7.org/fhir/uv/aitransparency/ImplementationGuide/hl7.fhir.uv.aitransparency" }] } }, { "url" : "http://hl7.org/fhir/StructureDefinition/structuredefinition-standards-status", "valueCode" : "trial-use", "_valueCode" : { "extension" : [{ "url" : "http://hl7.org/fhir/StructureDefinition/structuredefinition-conformance-derivedFrom", "valueCanonical" : "http://hl7.org/fhir/uv/aitransparency/ImplementationGuide/hl7.fhir.uv.aitransparency" }] } }], "url" : "http://hl7.org/fhir/uv/aitransparency/CodeSystem/AIdeviceTypeCS", "identifier" : [{ "system" : "urn:ietf:rfc:3986", "value" : "urn:oid:2.16.840.1.113883.4.642.40.79.16.1" }], "version" : "1.0.0-current", "name" : "AIdeviceTypeCS", "title" : "Device type for Artificial Intelligence", "status" : "active", "experimental" : false, "date" : "2026-08-06T14:31:47+00:00", "publisher" : "HL7 International / Electronic Health Records", "contact" : [{ "name" : "HL7 International / Electronic Health Records", "telecom" : [{ "system" : "url", "value" : "http://www.hl7.org/Special/committees/ehr" }, { "system" : "email", "value" : "ehr@lists.hl7.org" }] }], "description" : "This CodeSystem contains codes for the Device.type that indicate that the Device is an AI. The codes here were created by AI.", "jurisdiction" : [{ "coding" : [{ "system" : "http://unstats.un.org/unsd/methods/m49/m49.htm", "code" : "001" }] }], "caseSensitive" : true, "hierarchyMeaning" : "grouped-by", "content" : "complete", "count" : 32, "property" : [{ "code" : "abstract", "uri" : "http://hl7.org/fhir/concept-properties#notSelectable", "type" : "boolean" }], "concept" : [{ "code" : "Artificial-Intelligence", "display" : "All kinds of Artificial Intelligence", "definition" : "Any type of Artificial Intelligence system, undifferentiated.", "concept" : [{ "code" : "AI-By-Model", "display" : "Classification by Technical Model", "definition" : "This category classifies AI systems based on the underlying technical models they employ.", "property" : [{ "code" : "abstract", "valueBoolean" : true }], "concept" : [{ "code" : "Rule-Based-Systems", "display" : "Rule-Based Systems", "definition" : "Systems that apply explicitly authored rules, logic, knowledge representations, or inference procedures to input data." }, { "code" : "Machine-Learning-Models", "display" : "Machine Learning Models", "definition" : "They include supervised learning models (e.g., Support Vector Machines (SVM), Random Forests (RF)), which can be used for disease classification and risk prediction; unsupervised learning models (e.g., K-means clustering), which can discover hidden characteristics of patient subgroups; and reinforcement learning models, which can be applied in dynamic treatment plan management.", "concept" : [{ "code" : "Deep-Learning-Models", "display" : "Deep Learning Models", "definition" : "Examples include Convolutional Neural Networks (CNNs), which perform excellently in medical image analysis; Recurrent Neural Networks (RNNs) and their variant LSTMs, which are suitable for processing time-series physiological signal data; Generative Adversarial Networks (GANs), which can be used to synthesize training data and alleviate the scarcity of medical data; and Transformer models, which are widely used in multiple tasks such as medical imaging, text analysis, and physiological signal prediction.", "concept" : [{ "code" : "Foundation-Models", "display" : "Foundation Models", "definition" : "Deep learning models trained on broad data at scale that can be adapted to a range of downstream tasks.", "property" : [{ "code" : "abstract", "valueBoolean" : true }], "concept" : [{ "code" : "Large-Language-Models", "display" : "Large Language Models", "definition" : "These models, such as GPT-4 and PaLM, are trained on massive text datasets and can perform various natural language processing tasks, including medical text understanding, generation, and question answering." }, { "code" : "Vision-Foundation-Models", "display" : "Vision Foundation Models", "definition" : "Foundation models trained primarily to process and represent visual data for adaptation to vision tasks." }, { "code" : "Vision-Language-Models", "display" : "Vision-Language Models", "definition" : "Foundation models that jointly process visual and language data." }] }] }] }, { "code" : "Hybrid-Models", "display" : "Hybrid Models", "definition" : "These models combine multiple AI techniques to leverage their respective strengths. For instance, combining CNNs and RNNs can effectively process medical image sequences; integrating machine learning and deep learning models can enhance disease prediction accuracy; and combining rule-based systems with machine learning can improve interpretability and reliability in clinical decision support." }] }, { "code" : "AI-By-Behavior", "display" : "Classification by Output Behavior", "definition" : "Groups AI systems by whether identical inputs and execution conditions are expected to produce identical outputs.", "property" : [{ "code" : "abstract", "valueBoolean" : true }], "concept" : [{ "code" : "Deterministic-AI", "display" : "Deterministic AI", "definition" : "An AI system that produces the same output for the same input when its model, configuration, state, and execution environment are unchanged." }, { "code" : "Non-Deterministic-AI", "display" : "Non-deterministic AI", "definition" : "An AI system whose output may vary for the same input and execution conditions, for example because it uses probabilistic sampling or stochastic processing." }] }, { "code" : "AI-By-Scenario", "display" : "Classification by Application Scenario", "definition" : "This category classifies AI systems based on their application scenarios in the medical field.", "property" : [{ "code" : "abstract", "valueBoolean" : true }], "concept" : [{ "code" : "Intelligent-Diagnosis-and-Treatment", "display" : "Intelligent Diagnosis and Treatment", "definition" : "By analyzing massive volumes of medical data, these AI systems assist doctors in making more accurate diagnostic and treatment decisions." }, { "code" : "Medical-Image-Analysis", "display" : "Medical Image Analysis", "definition" : "Leveraging deep learning technologies, these AI tools automatically identify lesion areas in medical images." }, { "code" : "Personalized-Treatment", "display" : "Personalized Treatment", "definition" : "These AI systems create precise patient profiles to formulate personalized treatment plans." }, { "code" : "Clinical-Monitoring-and-Early-Warning", "display" : "Clinical Monitoring and Early Warning", "definition" : "Monitors longitudinal or real-time patient data to identify deterioration, adverse events, or other conditions requiring attention." }, { "code" : "Clinical-Documentation-and-Workflow", "display" : "Clinical Documentation and Workflow", "definition" : "Creates, summarizes, extracts, codes, routes, or quality-checks clinical and administrative information to support healthcare workflows." }, { "code" : "Drug-Discovery-and-Development", "display" : "Drug Discovery and Development", "definition" : "AI in this category accelerates the screening of candidate drugs and optimizes the design of clinical trials." }, { "code" : "Medical-Quality-Control", "display" : "Medical Quality Control", "definition" : "These AI tools are used to generate standardized medical document templates and detect defects in medical documents and images." }, { "code" : "Patient-Services", "display" : "Patient Services", "definition" : "AI systems here provide patients with services such as intelligent medical guidance, symptom self-assessment, and medical consultation." }, { "code" : "Population-Health-and-Operations", "display" : "Population Health and Healthcare Operations", "definition" : "Supports population stratification, public health, resource planning, scheduling, logistics, or other healthcare operational activities." }] }, { "code" : "AI-By-DataType", "display" : "Classification by Processed Data Type", "definition" : "This category classifies AI systems based on the types of medical data they primarily process.", "property" : [{ "code" : "abstract", "valueBoolean" : true }], "concept" : [{ "code" : "AI-for-Medical-Imaging-Data", "display" : "AI for Medical Imaging Data", "definition" : "It mainly processes medical imaging data such as X-rays, MRIs, and CT scans." }, { "code" : "AI-for-Physiological-Signal-Data", "display" : "AI for Physiological Signal Data", "definition" : "This type of AI deals with physiological signal data like electrocardiograms (ECG) and electroencephalograms (EEG)." }, { "code" : "AI-for-Medical-Text-Data", "display" : "AI for Medical Text Data", "definition" : "It processes text data such as electronic health records (EHRs) and medical abstracts." }, { "code" : "AI-for-Structured-Clinical-Data", "display" : "Structured Clinical Data", "definition" : "Processes coded, tabular, or relational health data such as diagnoses, medications, laboratory results, observations, claims, or FHIR resources." }, { "code" : "AI-for-Genomic-and-Omics-Data", "display" : "Genomic and Omics Data", "definition" : "Processes genomic, transcriptomic, proteomic, metabolomic, or related molecular data." }, { "code" : "AI-for-Audio-Data", "display" : "Audio Data", "definition" : "Processes speech, heart sounds, respiratory sounds, or other clinically relevant audio." }, { "code" : "AI-for-Video-Data", "display" : "Video Data", "definition" : "Processes temporal image sequences such as endoscopy, ultrasound cine loops, gait recordings, or procedure video." }, { "code" : "AI-for-Multimodal-Data", "display" : "Multimodal Data", "definition" : "Jointly processes two or more data modalities, such as images and text or physiological signals and structured clinical data." }] }] }] }