The platform, powered by SAP HANA, has already generated powerful data reports and insights in specific diseases for physicians and researchers
Danny Platform is an Analytics Healthcare platform able to evaluate massive amounts of real-world data in the oncology and cardiology fields. The solution provides real-time aggregated analyses of structured and unstructured health data with the integration of registries, external clinical and drug databases, etc.
It is no surprise that the analysis of big data has turned out to be one of the biggest undertakings in recent years for the healthcare industry. There are many benefits of turning data assets into insights. But there are also many challenges that need to be overcome before achieving data-driven clinical and financial goals.
In order to ensure high levels of accuracy and integrity in healthcare datasets, Danny utilizes proprietary algorithms and methodologies for data quality, data validation and predictive insights based on de-identified patient records.
The following add-on tools are aimed at improving the quality and cleanliness of data in Danny to ensure that datasets are accurate, correct, consistent and relevant.
- Data source validation and mapping
For each new data source, the system uses internal data structure mapping templates to list and examine all available parameters in the data source and their possible mapping to Danny data structures. This process describes any potential problems observed for each parameter. Each problematic parameter is further clarified with the hospital physician – in order to define the best strategy for parsing and normalization of such parameters.
Once each parameter is clarified, data structure mapping (from the data source to Danny) is performed, per template. Wherever applicable, additional techniques (such as regular expression, machine learning/natural language processing (NLP), etc.) for data extraction are further incorporated per parameter in the Danny system.
- Data parsing and validation
Draft parsing of the data is performed and loaded onto the Danny development system. Data parsing and insertion are validated manually by comparing the raw data and the data in our system, both in the backend (via querying of the database) and frontend (via visualizing the data through Danny applications).
- Data normalization and QA
Using the internal Danny QA system, the following checks are further performed
- Through import history: ensure all files are imported without error
- With categorical data normalization: ensure each relevant categorical parameter is normalized. Any ambiguous cases are further consulted with our contacts (consultants/physicians/IT) accordingly. Any normalizations that need additional feature building are to be discussed with the team.
- Using Quantitative data validation: ensure each relevant numerical parameter is normalized. Any ambiguous cases are further consulted with our contacts
Danny ensures high-quality and consistency in the data using custom-built data source integration, harmonization, and ML-based data quality controls pipeline.
Once the data from different sources is cleaned and normalized, Danny analytics are used to answer pertinent questions regarding the target patient population of interest, to generate summary reports, and to make timely decisions. The platform can, for example, summarize epidemiological statistics, analyze how specific factors influence and improve patient treatment, and compare the effectiveness of different regimens to achieve better patient care.
Sqilline is a leading software company for Big Data and AI technologies in the areas of Precision Medicine, Life Science and Healthcare.
“The improved new version of Danny can now provide better statistical insights and accurate analytics reports to all our customers. This type of technology brings critical and novel insights to guide the use of new therapies and improve the lives of everyday patients with cancer or cardiovascular issues”, added Desislava Mihaylova, Managing Director of Sqilline.
