Open datasets
Raman spectral sets with acquisition metadata, a data dictionary, licence terms and a citable DOI.
Browse datasets →Data Hub
BIRALAB open Raman spectral data, preprocessing standard and open-source tools — released with complete acquisition metadata so third parties can reproduce the results.
What the Data Hub is for
The Data Hub is where BIRALAB publishes Raman spectral data, its preprocessing standard and open-source tools. The goal is not storage but becoming the standardisation layer of Vietnam's Raman ecosystem: when other groups use our data and tools as a reference, a shared basis for comparison emerges and domestic research results become reproducible.
Everything here is released under an open licence, with full acquisition conditions so users can judge fitness for purpose before downloading. We publish no data containing patient-identifying information; clinical data is shared only after full de-identification and ethics committee approval.
When other groups, at home and abroad, use your data and tools as their reference standard, a leading position is established objectively, independently of publication counts.
Three ways into the Data Hub
Open-source tools
Preprocessing libraries, classification models and edge-deployment code, each with usage documentation.
Browse tools →Preprocessing standard
The mandatory metadata fields for every measurement and the recommended spectral preprocessing pipeline, step by step.
Read the standard →
Datasets
Each dataset states its spectrum count, instrument, version and usage licence.
- Licence: CC BY 4.0
In Vivo Raman Skin-Cancer Seed Spectra and Physically Constrained Augmented Cohort (2026)
Open Raman spectral dataset released with the paper "Physically Constrained Data Augmentation and Interpretable Machine Learning for Class-Imbalanced Raman Diagnosis of Skin Cancer" (ISDS-2026, paper 6661; Can Tho Unive…
- Spectra
- 1,254 spectra
- Instrument
- Portable Raman + autofluorescence spectroscopic system, Samara National Research University (Khristoforova et al., J. Biophotonics 12, e201800400, 2019)
- Version
- 1.0
- Released
- 2026
- Licence: CC BY 4.0
Technology Acceptance Survey for Raman Spectroscopy & AI in Biomedical Diagnostics (2026)
A conjoint analysis / discrete choice experiment surveying 1,890 respondents on their acceptance of Raman spectroscopy combined with AI for biomedical diagnostics. Includes 14 choice tasks, demographic profiles, and att…
- Spectra
- 1,890 spectra
- Instrument
- Online survey (conjoint analysis / DCE design, 20 blocks)
- Version
- 1.0
- Released
- 2026
Tools
Open source that runs as published, frozen at the version tag cited in the corresponding paper.
- Language: PythonLicence: MIT
raman-skincancer-augmentation
Open-source implementation of physically constrained data augmentation and interpretable machine learning for class-imbalanced Raman diagnosis of skin cancer. Full preprocessing chain, augmentation pipeline, PCA-30 + logistic regression classifier, and leakage-safe cross-validation. Regenerates every figure and metric of the paper.