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 Types & File Formats
Medical and clinical research relies on a broad range of data types — from imaging data and structured health records to clinical trial databases and wearable sensor streams. These formats are shaped by strict interoperability requirements and privacy regulations.
Medical Imaging Formats
DICOM (.dcm)
Digital Imaging and Communications in Medicine — the universal standard for medical images. A DICOM file contains both the image data and extensive metadata (patient info, acquisition parameters, equipment details).
- Use: Radiology (CT, MRI, X-ray, PET), ultrasound, nuclear medicine
- Tools: 3D Slicer, OsiriX/Horos, OHIF Viewer, pydicom (Python)
- Spec: DICOM Standard
- Note: DICOM files contain patient-identifiable information and must be anonymized before sharing
NIfTI (.nii, .nii.gz)
Neuroimaging Informatics Technology Initiative — a simplified imaging format widely used in neuroimaging research after anonymization from DICOM.
- Use: Brain MRI analysis, fMRI, DTI, research image sharing
- Tools: FSL, FreeSurfer, ANTs, nibabel (Python), 3D Slicer
- Spec: NIfTI-1 Format
NRRD (.nrrd)
Nearly Raw Raster Data — a flexible format for multi-dimensional raster data, commonly used in medical image analysis research.
- Use: Segmentation results, image analysis pipelines
- Tools: 3D Slicer, ITK, pynrrd (Python)
- Spec: NRRD Format
Whole Slide Images (WSI) — SVS, NDPI, MRXS
Vendor-specific formats for digitized histopathology slides. These are typically multi-resolution pyramid images reaching several gigabytes per slide.
- Use: Digital pathology, computational pathology, AI-based diagnosis
- Tools: OpenSlide, QuPath, ASAP
- Note: OpenSlide provides a unified API for reading various WSI formats
Clinical and Health Record Formats
HL7 FHIR (Fast Healthcare Interoperability Resources)
A modern RESTful standard for exchanging electronic health records. Data is organized as discrete resources (Patient, Observation, Condition, etc.) in JSON or XML.
- Use: EHR data exchange, clinical data integration, mobile health apps
- Tools: HAPI FHIR (Java), fhirpy (Python), SMART on FHIR
- Spec: HL7 FHIR
HL7 v2 / CDA
Older health information exchange standards still widely used in hospital systems. HL7v2 uses pipe-delimited messages; CDA uses XML clinical documents.
- Use: Lab results, admission/discharge messages, clinical summaries
- Note: Being gradually replaced by FHIR in new implementations
OMOP CDM (Observational Medical Outcomes Partnership Common Data Model)
A standardized relational data model that maps heterogeneous clinical data into a common structure, enabling cross-institutional observational research.
- Use: Observational studies, pharmacovigilance, real-world evidence
- Tools: OHDSI tools (ATLAS, Achilles, HADES)
- Spec: OMOP CDM
CDISC Standards (CDASH, SDTM, ADaM)
A family of standards for clinical trial data:
- CDASH — Clinical Data Acquisition Standards Harmonization (data collection)
- SDTM — Study Data Tabulation Model (regulatory submission)
- ADaM — Analysis Data Model (statistical analysis)
- Use: Clinical trial data management and regulatory submissions (FDA, EMA)
- Spec: CDISC
REDCap Export Formats (.csv, .r, .sps)
REDCap (Research Electronic Data Capture) exports data in CSV along with statistical software syntax files for labeling variables.
- Use: Survey data, clinical study databases, registries
- Tools: REDCap, R (REDCapR), Python (PyCap)
Biosignal and Wearable Data
EDF / EDF+ (.edf)
European Data Format — a standard for multi-channel biosignal recordings.
- Use: EEG, EMG, ECG, polysomnography, sleep studies
- Tools: MNE-Python, EDFbrowser, EEGLAB
- Spec: EDF Spec
WFDB (.hea, .dat)
WaveForm DataBase format — used by PhysioNet for physiological signal data.
- Use: ECG, heart rate variability, physiological signal databases
- Tools: WFDB (Python/MATLAB), PhysioNet
- Spec: PhysioNet WFDB
GDT / Accelerometer CSV
Various proprietary and open formats for wearable device data (accelerometers, gyroscopes, heart rate monitors).
- Use: Activity recognition, remote patient monitoring
- Note: No universal standard; most devices export CSV or JSON
Genomic and Molecular Data in Clinical Context
Clinical genomics uses many formats shared with bioinformatics (VCF, FASTQ, BAM). Additionally:
Phenopackets (.json)
A GA4GH standard for sharing disease and phenotype information associated with a patient or sample.
- Use: Rare disease diagnosis, clinical genomics, phenotype-genotype association
- Spec: GA4GH Phenopackets
Format Selection Guide
| Use Case | Recommended Format |
|---|---|
| Medical imaging (clinical) | DICOM |
| Neuroimaging (research) | NIfTI |
| Digital pathology | SVS / OpenSlide-compatible |
| Health record exchange | HL7 FHIR |
| Observational research | OMOP CDM |
| Clinical trial data | CDISC (SDTM/ADaM) |
| Study databases | REDCap |
| EEG/ECG signals | EDF+ |
| Clinical genomics phenotypes | Phenopackets |