Medical decision support system

Software for deciphering blood test results and diagnosing assistance


  • Information technology
  • Healthcare


  • Artificial intelligence


  • Keras
  • Numpy
  • TensorFlow
  • Python
  • pandas
  • Scikit-learn

When it comes to human life, every second counts. Rubius has developed an AI-based system that helps doctors instantly diagnose blood clotting disorders at the patient's bedside.

The system was ordered by Mednorth-Technics, a manufacturer of piezoelectric thromboelastographs. These instruments analyze blood clotting values in real time and provide results in the form of a chart with numerical values. To decipher them, a doctor needs a high qualification and a lot of time. To make the work of medical professionals easier, Mednorth-Technics decided to "teach" the device to issue conclusions based on thromboelastography results.

The Rubius team analyzed 1,300 thromboelastograph studies – 20 blood clotting parameters in each record and their corresponding diagnosis. Using this information, we developed a machine learning algorithm for classifying disorders and created a medical decision support system. It analyzes blood clotting parameters recorded by the thromboelastograph, determines what kind of disorder the patient has, and provides the doctor with a clear conclusion. It takes a couple of seconds to get the result, and its accuracy is 98.4%.

Let’s discuss your project

Tell us about your requirements and we’ll get back with a possible technical solution

By clicking «Submit», you consent to the processing of your personal data