Scientific Research

Solution

Easily enhance your research with wearables data-driven deep learning approach

Using our data standardisation you can easily share your code, increase your work visibility and citation rate.

Mastering wearables data formats, deep learning techniques for time series and circadian rhythms analysis has never been easier.

We want to help you remove the intimidating confusion of what device, data format, dataset to use in your research connected to circadian rhythms.

Circadian rhythms are individual and research groups struggle to better understand how synchronization of inner body clock with lifestyle habits impacts health and performance. We have a team of experts in academic research and we understand the needs of scientific community well. We want to help researchers to easily adopt consumer-grade wearables - a rich source of data on circadian rhythms, their stability and sensitivity to lifestyle and health factors. To do that we developed a data processing and feature extraction algorithms suitable to incorporate various data formats, data sources and quality:

Deep learning

Easily employ deep-learning neural network models in your work.

Batch effect

Remove batch effects by adjusting for device manufacturer/ os version.

Missing data

Impute short periods missing data.

Predict

Build predictive models, do statistical analysis.

Validate

Evaluate your results versus biobank references.

Stratification

Adjust data/biomarkers for demographic, lifestyle, health, and socio-economic characteristics.

Label data

Detect anomaly events (time zone, lifestyle, etc.)

Recovery rates

Estimate individual sample's stability and recovery time.