Figure: Example of coloured OCT images. Left column: healthy samples. Right colum: neoplastic samples
Preliminary version (v1)
This dataset consists of Optical Coherence Tomography (OCT) images acquired with a spectral domain commercial system (Thorlabs Callisto) from colon polyps and healthy tissue samples of rat models.
Samples have been acquired from a group of 60 neoplastic rats (clinically validated model Pirc F344/NTac-Apcam1137) and 30 hyperplastic rats (created with an induced colorectal hyperplastic murine model). Healthy samples have been acquired from a group of 10 control animals or non-lesion tissue of hyperplastic rats.
The dataset consists of the following:
healthy tissue (10 samples; 48 C-scans; 12820 B-scans)
hyperplastic polyps (13 samples; 53 C-scans; 13903 B-scans,)
neoplastic polyps (75 samples; 245 C-scans; 67964 B-scans)
*C-scan: 3D volume form of consecutive B-scan images
The dataset is organized in folders considering the target diagnosis as described above. Then, the samples folder name has the following naming conventions:
ORIGIN_ID: Group the specimen belongs to
ANIMAL_ID: specimen identifier within the group
SECTION: Section of the colon analysed (A-ascending, T-transverse, D-descending)
N_SAMPLES_SECTION: Number of samples analysed in the current section
LESION_ID: Identifier of lesion analysed
NUM_LESIONS: Total number of lesion found as part of the sample
SEQUENCE: Number identifying the C-scan. Lesions are usually form by various consecutive C-scan volumes.
To access these images, it is necessary to fill out this form: https://forms.office.com/r/BmpMMihPcj or to contact the Basque Biobank, email@example.com which will inform you of the conditions for access.
Please cite this reference when using this collection:
Saratxaga, C.L.; Bote, J.; Ortega-Morán, J.F.; Picón, A.; Terradillos, E.; del Río, N.A.; Andraka, N.; Garrote, E.; Conde, O.M. Characterization of Optical Coherence Tomography Images for Colon Lesion Differentiation under Deep Learning. Appl. Sci. 2021, 11, 3119. https://doi.org/10.3390/app11073119.
Full version (v2)
NOTE: This collection will be available when the associated publication is published.