Data Analysis

Once your users' data is captured, our system processes the data through four critical steps: data harmonization, data standardization, data cleaning, and normalization. Each step is designed to ensure the delivery of optimal and homogenized data.

Harmonization

Harmonized health data are standardized to be consistent in format, units, and definitions, enhancing the efficiency and accuracy of data collection, sharing, and analysis.

In this phase, we focus on converting data from various sources into a uniform metric system. For instance, distance data received in different units like miles or kilometers are standardized to kilometers to maintain consistency.

Standardization

Standardized health data adhere to established standards, which are sets of rules that define the collection, organization, and usage of health data.

These standards are vital for ensuring compatibility across different health data providers, facilitating easier data collection, sharing, and analysis.

Cleaning

Clean health data are free from errors, inconsistencies, and duplicates, making them crucial for reliable data analysis.

The cleaning process is often intricate, utilizing advanced data cleaning techniques. It addresses issues like multiple entries from different wearable devices or overlapping data from various health data providers and collectors (e.g., Garmin, Polar, Fitbit with Apple Health, Health Connect).

Duplicity

The Duplicates Handling Project aims to streamline data flow from users connected to multiple sources, enhancing our clients' decision-making capabilities. Our architecture incorporates data prioritization, event generation, summary creation, and data cleansing within these summaries. This ensures the delivery of the most pertinent and accurate user data, fostering proactive and effective decision-making. For more details, read the article.

Normalization

Normalized health data have been adjusted to uniform scales and ranges, allowing for effective comparison and analysis across different data sources.

Normalization techniques may include adjusting measurement scales or converting data into a common format. For example, normalizing blood pressure readings ensures comparability, regardless of the measuring device.