DIARETDB1 - Standard Diabetic Retinopathy Database

Calibration level 1

Description

This is a public database for benchmarking diabetic retinopathy detection from digital images. The main objective of the design has been to unambiguously define a database and a testing protocol which can be used to benchmark diabetic retinopathy detection methods. By using this database and the defined testing protocol, the results between different methods can be compared. For more information refer to the documentation.

Authors

Tomi Kauppi
Valentina Kalesnykiene
Joni-Kristian Kamarainen
Lasse Lensu
Iiris Sorri
Asta Raninen
Raija Voutilainen
Juhani Pietilä
Heikki Kälviäinen
Hannu Uusitalo

Copyright

DIARETDB1 is Copyright © 2007 by Tomi Kauppi, Valentina Kalesnykiene, Joni-Kristian Kamarainen, Lasse Lensu, Iiris Sorri, Asta Raninen, Raija Voutilainen, Juhani Pietilä, Heikki Kälviäinen, and Hannu Uusitalo.

Usage

The database can be freely downloaded and used for scientific research purposes. You are not allowed redistribute or modify it. Proper citing of this resource is expected if the database is used in research or other reporting.

The database is meant to be useful, but it is distributed WITHOUT ANY WARRANTY, and without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

Inclusion of this database or even parts of it in a proprietary program is not allowed without a written permission from the owners of the copyright. If you wish to obtain such a permission, you should contact

    Machine Vision and Pattern Recognition Research Group
    Laboratory of Information Processing
    Lappeenranta University of Technology
    PO Box 20
    FI-53851 Lappeenranta
    FINLAND
or owner-diaretdb<at>lists.lut.fi (in the address, replace the string <at> with @, please).

All communication related to the database updates and announcements of future releases will be carried out through the diaretdb mailing list. You can subscribe the list by sending a message to the address diaretdb-on<at>lists.lut.fi by using your default e-mail system. You will get a confirmation with further instructions after your subscription.

If you find any errors or bugs, or have comments related to the database contents, please contact diaretdb<at>lists.lut.fi.

Data

The database consists of 89 colour fundus images of which 84 contain at least mild non-proliferative signs (Microaneurysms) of the diabetic retinopathy, and 5 are considered as normal which do not contain any signs of the diabetic retinopathy according to all experts who participated in the evaluation. Images were captured using the same 50 degree field-of-view digital fundus camera with varying imaging settings. The data correspond to a good (not necessarily typical) practical situation, where the images are comparable, and can be used to evaluate the general performance of diagnostic methods. This data set is referred to as "calibration level 1 fundus images".

Example entry with ground truth:

image008 image010 image015 image020 image025 image026

Ground truth

Independent markings from 4 medical experts were collected by using a software tool provided for image annotation. The computer displays used in the collection process were not calibrated. A person with medical education and solid experience in ophthalmology was considered as an expert. The experts were asked to mark the areas related to the microaneurysms, hemorrhages, and hard and soft exudates. The experts were instructed to avoid marking the findings so that the borders of the marked areas contain any pixels belonging to the finding. The experts were further instructed to report their confidence and especially annotate the single most representative point for each finding. The ground truth confidence levels, {<50%,~>50%,~100%}, represented the certainty of the decision that a marked finding is correct.

The uninstructed collection process caused significant differences between the medical experts. Therefore, the original expert knowledge was fused for a better spatial accuracy and suppression of outliers.

Example visual expert markings:

image015 expert 1 expert 2 expert 3 expert 4

Documentation

  1. Kauppi, T., Kalesnykiene, V., Kamarainen, J.-K., Lensu, L., Sorri, I., Raninen A., Voutilainen R., Uusitalo, H., Kälviäinen, H., Pietilš, J., DIARETDB1 diabetic retinopathy database and evaluation protocol, Technical report (PDF).
  2. Kauppi, T., Kalesnykiene, V., Kamarainen, J.-K., Lensu, L., Sorri, I., Raninen A., Voutilainen R., Uusitalo, H., Kälviäinen, H., Pietilš, J., DIARETDB1 diabetic retinopathy database and evaluation protocol, In Proc of the 11th Conf. on Medical Image Understanding and Analysis (Aberystwyth, Wales, 2007). Accepted for publication

Download

If you have problems or suggestions concerning this software package, do not hesitate to contact owner-diaretdb<at>lists.lut.fi, or a member of the ImageRet project.

Links to other resources