This page is currently under development. Content may be incomplete or contain inaccuracies. If you notice any errors or have suggestions, please contact us.
Medical & Clinical Data Glossary
This glossary provides definitions for common terms used in medical data science, clinical research, and health informatics.
Clinical Research
Clinical Trial
A prospective research study that assigns participants to interventions (treatments, procedures, devices) to evaluate their effects on health outcomes.
Randomized Controlled Trial (RCT)
A clinical trial in which participants are randomly assigned to treatment or control groups, considered the gold standard for evaluating interventions.
Cohort Study
An observational study that follows a group of individuals over time to identify factors associated with outcomes.
Case-Control Study
An observational study that compares individuals with a condition (cases) to similar individuals without it (controls) to identify risk factors.
Blinding (Single / Double)
A method to prevent bias by concealing group assignments. Single-blind: participants don't know their group. Double-blind: neither participants nor investigators know.
Informed Consent
The process by which a participant voluntarily confirms their willingness to participate in a study, after being informed of all relevant aspects.
GCP (Good Clinical Practice)
An international ethical and scientific quality standard for designing, conducting, and reporting clinical trials (ICH E6).
IRB / Ethics Committee
An institutional review board or ethics committee that reviews and approves research involving human subjects.
Data Standards and Formats
DICOM
Digital Imaging and Communications in Medicine — the universal standard for medical images, containing both pixel data and extensive metadata.
NIfTI
Neuroimaging Informatics Technology Initiative — a simplified imaging format used in brain imaging research after anonymization from DICOM.
HL7 FHIR
Fast Healthcare Interoperability Resources — a modern standard for exchanging electronic health records using RESTful APIs and JSON/XML resources.
OMOP CDM
Observational Medical Outcomes Partnership Common Data Model — a standardized relational model for harmonizing clinical data from heterogeneous sources.
CDISC
Clinical Data Interchange Standards Consortium — a family of standards for clinical trial data (CDASH, SDTM, ADaM).
SDTM
Study Data Tabulation Model — a CDISC standard for organizing clinical trial data into standardized domains (DM, AE, LB, VS, etc.).
ADaM
Analysis Data Model — a CDISC standard for analysis-ready datasets derived from SDTM data.
BIDS
Brain Imaging Data Structure — a standard for organizing neuroimaging data in a consistent directory structure.
Medical Terminology Systems
ICD (International Classification of Diseases)
The WHO standard for coding diagnoses and health conditions. Current versions: ICD-10 (widely used) and ICD-11 (latest).
SNOMED CT
Systematized Nomenclature of Medicine — Clinical Terms. The most comprehensive clinical terminology covering diseases, procedures, and substances.
LOINC
Logical Observation Identifiers Names and Codes — a universal coding system for laboratory tests and clinical observations.
MedDRA
Medical Dictionary for Regulatory Activities — a standardized terminology for coding adverse events and medical history in clinical trials.
MeSH
Medical Subject Headings — a controlled vocabulary maintained by the NLM for indexing biomedical literature.
ATC
Anatomical Therapeutic Chemical classification — a WHO system for classifying drugs by their therapeutic use and chemical properties.
HPO
Human Phenotype Ontology — a standardized vocabulary for describing phenotypic abnormalities, used in rare disease research.
Privacy and Ethics
GDPR
General Data Protection Regulation — the EU regulation governing the processing of personal data, including health data.
Anonymization
The irreversible process of removing all identifying information so that an individual cannot be re-identified from the data.
Pseudonymization
Replacing identifying information with artificial identifiers (pseudonyms), such that re-identification is possible only with a separate key.
De-identification
The general process of removing or obscuring personal identifiers from data. Encompasses both anonymization and pseudonymization.
Data Use Agreement (DUA)
A legal agreement specifying the conditions under which a dataset may be accessed and used for research.
Federated Learning
A machine learning approach that trains models across multiple institutions without sharing raw patient data — the algorithm travels to the data, not the other way around.
Biostatistics
Primary Endpoint
The main outcome measure of a clinical trial, pre-specified in the protocol.
Power Analysis
A statistical calculation to determine the minimum sample size needed to detect a clinically meaningful effect with sufficient probability.
Kaplan-Meier Estimator
A non-parametric method for estimating survival probabilities from time-to-event data, accounting for censored observations.
Hazard Ratio
A measure of the relative risk of an event (e.g., death, recurrence) in the treatment group compared to the control group over time.
Intention-to-Treat (ITT)
An analysis that includes all randomized participants regardless of whether they completed the study protocol, preserving the benefits of randomization.
Per-Protocol Analysis
An analysis restricted to participants who completed the study according to the protocol. Less conservative than ITT.
Multiple Testing Correction
Statistical adjustment for performing many simultaneous tests to control the false discovery rate. Common methods: Bonferroni, Benjamini-Hochberg.
Censoring
In survival analysis, an observation is censored when the event of interest has not occurred by the end of the study period or the participant is lost to follow-up.
Medical Imaging
CT (Computed Tomography)
An imaging technique that uses X-rays to create cross-sectional images of the body. Measured in Hounsfield Units (HU).
MRI (Magnetic Resonance Imaging)
An imaging technique using strong magnetic fields and radio waves to produce detailed images of soft tissues.
Hounsfield Unit (HU)
The unit of measurement for CT image intensity. Water = 0 HU, air = -1000 HU, bone = +400 to +1000 HU.
Segmentation
The process of delineating anatomical structures or regions of interest in a medical image.
PACS
Picture Archiving and Communication System — the hospital system for storing, retrieving, and distributing medical images.
Radiomics
The extraction of quantitative features (shape, texture, intensity) from medical images for use in predictive modeling.
Need clarification on a term? Contact us at rdhub@mdsi.tum.de or explore our Use Cases for practical examples.