Resource

Data

A Database on Parkinson’s Disease

A Parkinson’s database comprising MRI and DaT Scan data from 78 subjects, 55 patients with Parkinson’s and 23 non-patients, has been recently released, in a collaboration with the Department of Neurology of the Georgios Gennimatas General Hospital in Athens and the National Technical University of Athens, Greece (please refer to [1]). It includes, in total, 41528 MRI data (31147 from patients and 10381 from non-patients) and 925 DaT Scans (595 and 330 respectively).

Our target has been to train Deep Convolutional and Convolutional-Recurrent Neural Networks with this database for Parkinson’s diagnosis, viewed as a classification problem in two categories (patients, non-patients). In this framework, we form triplets of MRI consecutive frames (per subject), so as to target estimation of the volume of structures in the brain.
We then split the database in a training/validation dataset and in a test dataset, performing an augmentation procedure in the train dataset, as follows. After forming the triplets of consecutive MRI frames, we generate combinations of these image triplets with each one of the DaT Scans in each category (patients, non-patients). Consequently, we created a dataset of 66,176 training inputs, each of which consists of 3 (gray-scale) MRI and 1 (colour) DaT Scan images. In the test dataset, which refers to different subjects than in the training dataset, we make this combination per subject; this created 1130 test inputs.
Both the training and test datasets, each being split in two folders, patients and non-patients, are available upon request. Please contact Professor Stefanos Kollias at skollias@lincoln.ac.uk.

[1] D. Kollias, A. Tagaris, A. Stafylopatis, S. Kollias, and G. Tagaris, “Deep neural architectures for prediction in healthcare”, Complex & Intelligent Systems, vol. 4, no. 2, pp. 119–131, 2018.


Single Slot Display Advertising Datasets

This data description contains two parts:

  • Overview of the datasets
  • Design of data crawler

For details, please visit https://github.com/boweichen/SingleSlotRTBTradeoffsMultimediaDataset

Multi-Slot Display Advertising Datasets

This data description contains two parts:

  • Overview of the datasets
  • Design of data crawler

For details, please visit https://github.com/boweichen/MultiSlotRTBMultimediaDataset


Leave a Reply

Your email address will not be published.