Hospital-sourced pathology images, medical imaging, clinical records, molecular data, treatment histories, and longitudinal outcomes — collected and prepared to your project specifications.
REMI Data helps AI developers and life sciences companies evaluate and organize custom oncology data sourcing through healthcare institutions and laboratories in Ukraine.
Six core data modalities organized around buyer-defined oncology requirements. Actual fields, patient counts, linkage and source permissions depend on the specific project.
REMI Data evaluates buyer-defined data sourcing requirements across solid tumors, hematologic malignancies, and skin cancers. Potential sourcing configurations vary by diagnosis, required data modalities, available medical records, treatment histories, and clinical endpoints.
This list represents potential areas for feasibility evaluation, not a catalogue of confirmed datasets.
A useful oncology AI dataset often requires mapping images, diagnosis, treatments, and outcome events to the same patient, specimen, or clinical episode. Patient-level linkage enables multimodal model development, outcome prediction, and clinical validation.
Linkage feasibility, completeness, and missingness are assessed for each requested cohort. Complete multimodal linkage is not guaranteed for every case.
Illustrative cancer-agnostic dataset schema. Actual content is defined by your project specifications.
| Category | Example Fields |
|---|---|
| Identifiers | patient_id, specimen_id, slide_id |
| Pathology | WSI file, histology, tumor grade |
| Clinical | diagnosis, TNM stage, laboratory values |
| Imaging | modality, image_id, acquisition metadata |
| Molecular | biomarkers, mutations, expression data |
| Treatment | treatment_type, treatment_start_date |
| Outcomes | recurrence, progression, survival |
| Follow-up | last_follow_up, censoring_status |
Illustrative data structure only. Actual dataset content, eligible patient counts, completeness, and commercial licensing are determined through feasibility assessment.
Two sourcing models for custom oncology datasets.
Potential assessment of existing hospital archives and records:
Potential organization of buyer-defined protocols:
Long-term survival data generally require existing longitudinal records or future follow-up periods. Immediate delivery of outcomes requiring years of observation should not be assumed.
From requirements to delivery in five steps.
Cancer indication, cohort size, data modalities, clinical endpoints, inclusion criteria, delivery format, geography, and intended AI use.
Evaluate potential Ukrainian hospital and laboratory sources, case availability, completeness, technical specifications, and timelines.
Review data permissions, legal requirements, de-identification, privacy, commercial AI training rights, and contractual restrictions.
Where feasible and authorized, coordinate collection, digitization, clinical abstraction, linkage, normalization, annotation, and quality control.
Validate accepted data against agreed specifications and coordinate secure delivery and documentation.
Data governance and licensing requirements are evaluated and documented for each specific project. De-identification alone does not establish commercial licensing rights.
Applicable consent, legal basis and healthcare source permissions verified for each data source and intended use.
Ukrainian privacy and health-data regulations, HIPAA and GDPR considerations where applicable.
De-identification and privacy safeguards applied according to project requirements and regulatory standards.
Commercial AI-training rights assessed and documented per project. Access to data does not automatically create usage rights.
Secure data transfer methods and data-use agreements structured for each delivery.
Quality assurance, source verification and dataset traceability documented throughout the sourcing process.
Whether you're developing a digital pathology model, multimodal oncology AI system, diagnostic algorithm, or clinical outcome prediction model, REMI Data can review your requirements and evaluate potential sourcing options through Ukrainian medical providers.
Request a Dataset Feasibility Assessment