Download results!

Download your results with the Job Id and password received by email.

What is GP Progression Model?

GP Progression Model estimates the long term trajectory underlying the ensemble of observed short term data. When applied to clinical data, it provides a data-driven description of the natural evolution of the pathology.
GP Progression Model in action
GP Progression Model: from spaghetti plot (short term data) to long-term trajectories
Features:
  • Explicit description of the biomarker transition from normal to pathological stages along the estimated disease time axis.
  • Quantification of the variability and the diagnostic value of biomarkers across disease stages.
  • Disease severity quantification with respect to missing measurements, biomarkers, and follow-up information.
  • The estimated trajectory provides a statistical reference for the accurate probabilistic assessment of the pathological stage in testing individuals

Website content

The authors take no responsibility for the accuracy, completeness or quality of the information provided. The authors are in no event liable for damages of any kind incurred or suffered as a result of the use or non-use of the information presented on this website or the use of defective or incomplete information unless the authors have been acting deliberately or in a wantonly negligent manner. The contents of this website are subject to confirmation and not binding. The authors expressly reserves the right to alter, amend or remove pages, whole and in part, without prior notice or to discontinue publication for a period of time or even completely.

Data Protection

The user has the possibility to enter clinical or personal data. The disclosure of this data is voluntary. This service can be used without disclosing personal information or by use of anonymised data or aliases.

This site uses SSL or TLS encryption for security reasons and for the protection of the transmission of confidential content, such as the inquiries you send to us as the site operator. You can recognize an encrypted connection in your browser's address line when it changes from "http://" to "https://" and the lock icon is displayed in your browser's address bar. If SSL or TLS encryption is activated, the data you transfer to us cannot be read by third parties.

Every data transmitted to this web server will be stored as a protocol file for a maximum of 7 days exclusively for the purpose of data security and information retrieval. The protocol file includes the following data: transmitted data file, derived data and information obtained by running the software, date and time of access, volume of data transferred, user identifier. The data saved are not used for statistical purposes, and will not be disclosed to third parties for commercial or non-commercial purposes.

A simple front-end to GP Progression Model


Try it now



    Instructions:

    • Data should be in .csv format (comma separated)
    • After loading the data, the user can select the variables by cheking the respective left boxes
    • Three special data fields must be initially indicated: Subject identifier, Time, and Group
    • The user can further select the fields to be analyzed by GP Progression Model. By clicking on the right selection tool, the user should specify whether the progression of the field is expected to be monotonically decreasing (-) or increasing (+). If no apriori behavior is known, the user can choose (0).
    • When GP Progression Model completes the estimation the user will receive a notification with a link for downloading the results.


    Acknowledgments

    If you found GP Progression Model useful for your work, please cite the following papers:

    • Marco Lorenzi, Maurizio Filippone, Giovanni B. Frisoni, Daniel C. Alexander, Sebastien Ourselin. Probabilistic disease progression modeling to characterize diagnostic uncertainty: application to staging and prediction in Alzheimer’s disease. NeuroImage, S1053-8119(17)30706-1, 2017.
    • Marco Lorenzi and Maurizio Filippone. Constraining the Dynamics of Deep Probabilistic Models. Proceedings of the 35th International Conference on Machine Learning (ICML), PMLR 80:3233-3242, 2018.