Industry
Text Link
Healthcare
Technology used
FSL, Python, Bash
Share

Medical research is a very broad and complicated field. It is constantly evolving with new discoveries and techniques emerging that the researchers must stay up to date. They are also dealing with human subjects, which is often regulated, and even regardless of that requires special cautiousness as well as ethical and patient health considerations. Finally, the studied diseases and disorders can have diverse causes, symptoms, and outcomes, with the dependencies between them being very sophisticated. 

All of the mentioned factors and even more are making medical research very challenging and complex – researchers are dealing with a large quantity of multidimensional data that they need to analyze and derive insights from. Even before that, they need to prepare it by actually measuring varied parameters (e.g. computing some structure size on the imaging data) which doing manually is often infeasible. Technology can help here to bridge those needs, but the problem is, that it  often (especially in the case of novel tools) requires technical expertise that medical specialists do not possess. As one of our projects, Datarabbit was helping a research group from one of the top Poland universities (Adam Mickiewicz University in Poznań) to address exactly this issue.

In our collaboration, we developed a toolset tailored to the scientists' needs that is user-friendly and very easy to use, and integrated capabilities of multiple advanced brain MRI data processing (coming from world-renowned research centers like e.g. Oxford) and analytics programs together,  helping the users to:

  • Automatically extract structures and lesions from brain MR imaging and measure their features and sizes.
  • Perform statistical analysis on the combination of patient data from different sources (imaging, forms, etc.) – e.g. looking for correlations, doing regression analysis, and testing varied statistical hypotheses in a streamlined fashion.
  • Generate attractive summary graphs and visualizations.
Visualized (single-patient) brain corpus callosum structure and measurements as generated per developed toolset.
Correlation analysis for a pair of statistical variables on a larger cohort of patients. New tools are performing analysis and generating graphs for a dozens of combinations, and automatically highlight the most relevant ones, greatly offloading researchers.

The analyses and tools developed were part of the broad project conducted by Dr. Natalia Nowaczyk (Manager & Principal Investigator of the project, Assistant Professor at Adam Mickiewicz University). The research project was funded by grant no. 75/44/3.4.3/18/DEA under the Polish National Health Program for 2016–2020, awarded by the State Agency for the Prevention of Alcohol-Related Problems (PARPA) under the auspices of the Polish Ministry of Health. Overall the project was a significant success as it enabled the research group to test multiple hypotheses they weren’t able to evaluate before – which resulted in material for a number of future publications. The following were already published:

Highlighted work

see more case studies

Let's talk about how datarabbit can help your company.

Contact us
We use cookies to improve your experience on our website. By clicking “Accept all’, you agree to the use of all cookies. We may process your cookies outside the European Economic Area.
More information