AI Doctors + Medical Advisors
based on the non-invasive detection of biomarkers in human skin, blood & cellular fluid

SIATLAB is creating AI Doctors and Medical Advisors for smart patients, medical offices, clinical centers, insurance companies and public health institutions. We are developing deep-tech hardware and software solutions for the non-invasive detection of biomarkers in human skin, blood and cellular fluid, that can be integrated into a wide range of devices and products. The measurement data is combined with existing historical clinical data and (NGS) Next Generation Genome Sequencing to create data products and large scale prediction models for public and private customers. SIATLAB has developed a sensor for the non-invasive monitoring of blood glucose with a smartphone and built profound know-how in the fields of data analysis and artificial intelligence. We created AI-based algorithms to identify key parameters of human physiology and developed multispectral analysers leading into a deep understanding of human pathophysiology. A first clinical study confirmed proper accuracy, a second study on a much larger patient pool has already started.

Key Elements

We are using multispectral optical sensors with light radiation and detection that can be implemented in any kind of measuring device. The measurement takes about one minute and 20 seconds, the signal processing and sugar level prediction from the artificial intelligence cloud service algorithm is performed in only 10 seconds. Our application on the device guides the user through the whole process. 


Multispectral Optical Sensor with light radiation and detection


or any other kind of device with our GUI application

Artificial Intelligence 

AI Cloud Service Algorithm, signal processing and value calculation

Clinical Study

The research was done to assess the reliability of non-invasive glucose measurements compared to established standard methods. A standardized analysis of the accuracy of blood glucose measurements is generally shown by the Clark-Error-Grid diagram. It shows a visual statistical estimation of accuracy based on the distribution of test results by individual zones: A - E. It is important that most measurements are in the lowest (hence A) zone.

  • OGTT tests performed on 40 persons of different gender and age according to standard clinical practice
  • Clinical study conducted at the Faculty of Medicine in Maribor confirms proper accuracy
  • Clinical research on a large patient pool at Maribor University Clinical Center

Development Goal

After glucose, SIATLAB is now working on the non-invasive detection of other key blood parameters that can be measured based on their molecular structure.


SIATLAB GmbH (FN 530760 f) is based in Graz/Austria and is bringing together Slovenian and Austrian (SI/AT) researchers in the fields of electronics, medicine and artificial intelligence.