Kontakt

No. Savana works on a license model. One of our core fundamental values is to reject any model reliant on the sale of personal data to third parties. Savana only helps users manage their own data.

Savana’s algorithm ensures anonymization; therefore, no data protection regulation applies.

Additionally, clinical records are never extracted from the institution; only aggregated clinical information. Individual data is never used to train the algorithms.

Any clinical document: reports, test results, prescriptions, etc.

No, just on electronic clinical records.

Yes, electronic clinical records are rapidly penetrating the healthcare system, and the majority of developed countries have a growing number of institutions working with them.

Because Savana reads plain text, it can work in any system.

In every case, Savana performs a quality analysis of the documents, selecting those who meet the required standards for extraction and reporting the result to our users.

Savana is a medical company and the first which has been able to introduce deep learning in the field, thus able to train millions of clinical texts in the shortest time. Furthermore, it was founded by an experienced clinical team, able to develop useful functionalities for the real clinical environments where fast multiple decision making is both mandatory and common practice.

Proprietary EHRead technology combines several techniques in order to detect concepts and relationships among them. Supported by linguistic tools, statistics, databases and medical knowledge, the platform is able to process the particularities of human language: typos, multiple ways of expressing negation or speculation, acronyms, subjectivity, subordinated structures, etc., in order to determine which unique concepts correspond to each clinical concept.

Savana’s tools are designed to deliver the most complete information in an agile, intuitive and smart way, without a user guide.

A support team of IT specialists and practicing physicians is available at all times.