Welcome to the Leaspy ML Tool in Neugrid!

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 666992.


This tool predicts the evolution of cognitive scores for an Alzheimer’s Disease patient given baseline and optional follow-up information. The tool is based on a Leaspy model trained on cognitive scores of ADNI amyloid-positive (PET or CSF Ab42) subjects. The Leaspy model used on neugrid2 is a limited version, not taking advantage of many other biomarkers (MRI, PET, CSF) that can be modelled as well to improve predictions. Please see the section below for more information about Leaspy.

Outputs of this tool are for research use only and come without warranty of any kind!

How to run the analysis:

You have to provide the following information:

  • The email you want the report delivered to
  • baseline Age of the subject
  • at least one between baseline Mini Mental State Examination (MMSE) score and baseline Alzheimer’s Disease Assesment Score-Cognitive- 13 items (ADAS-Cog-13)
  • optional follow-up information (age, and at least one cognitive score must be provided for each follow-up you wish to enter)

Please note that the model expects a decreasing trend for MMSE and an increasing trend for ADAS-Cog-13. In case of input of longitudinal data with a trend opposing the expected one, the prediction may be inaccurate.

Enter your e-mail:

  Age MMSE ADAS-Cog-13
Baseline (mandatory)
First follow-up (optional)
Second follow-up (optional)
Third follow-up (optional)

Description of Leaspy (LEArning Spatiotemporal Patterns in Python)

Leaspy is a software package for the statistical analysis of longitudinal data, particularly medical data that comes in a form of repeated observations of patients at different time-points.

Considering these series of short-term data, the software aims at :

  • Recombining them to reconstruct the long-term spatio-temporal trajectory of evolution

  • Positioning each patient observations relatively to the group-average timeline, in term of both temporal differences (time shift and acceleration factor) and spatial differences (different sequences of events, spatial pattern of progression, …)

  • Quantifying impact of co-factors (gender, genetic mutation, environmental factors, …) on the evolution of the signal

  • Imputing missing values

  • Predicting future observations

  • Simulating virtual patients to unbias the initial cohort or mimic its characteristics

The software package can be used with scalar multivariate data whose progression can be modeled by a logistic shape, an exponential decay or a linear progression. The simplest type of data handled by the software are scalar data: they correspond to one (univariate) or multiple (multivariate) measurement(s) per patient observation. This includes, for instance, clinical scores, cognitive assessments, physiological measurements (e.g. blood markers, radioactive markers) but also imaging-derived data that are rescaled, for instance, between 0 and 1 to describe a logistic progression.

Leaspy has exstensively been used for scientific production (Schiratti et al. 2017, Maheux et al. 2020, Koval et al. 2021) and it is available to researchers and developers as a python package (https://gitlab.com/icm-institute/aramislab/leaspy