Intensive research on basic scientific questions with high potential impact on society.
What we do
Our gained knowledge to develop technological solutions for medical problems.
Medical solutions with clinical trials, translating technology to the real world.
Who we are
- Daniela Andres, MD, PhD.
- Gianfranco Bianchi, MSc Biom. Eng.
- Oscar Filevich, PhD.
We are a group of professionals with passion for medical technology and translational research. Our group was born at UNSAM. We work in collaboration with the LabNIng.
We are working on
Quantitative diagnosis of Parkinson’s disease
Parkinson’s disease is a complex disease with motor and non-motor symptoms that affects 2% of the world’s population over 65 years. Some of our patients are being diagnosed with Parkinson’s disease and some have already been, but are now in the process of adjusting the treatment. While patients have difficulties accessing specialized neurologists, medics struggle with derivation decisions and follow-up difficulties.
- For patients, we solve accessibility to high quality medical diagnosis of Parkinson’s disease. We developed a system that draws on the knowledge of experienced neurologists, allowing more people to access medical treatment. Our system offers an accurate evaluation of the motor tasks that are impaired in Parkinson’s in only 10 minutes.
- For specialists, the problem that we solve is the inefficient follow-up of patients, which complicates medication selection and dosing adjustments.
- In the case of general practitioners, they can use this system to optimize the selection of patients who need to be derived to a specialist.
Quantitative diagnosis in Cardiology
Automatic ischemia detection
Myocardial ischemia is the pathology due to lack of oxygen in the heart tissue, whose most serious expression is myocardial infarction. Using the public Physionet database, we reviewed and classified electrocardiograms to obtain representative isolated cardiac cycles from healthy and ischemic cases. We apply our own cardiac cycle detection algorithm, and use them to train a neural network capable of detecting ischemia from a single ECG cycle. We obtained an accuracy of 86%. This type of analysis can be applied to the early detection of ischemia, in stages in which myocardial infarction can be prevented or treated without leaving sequelae.
Cuantificación de obstrucción carotidea
Aortic valve replacement accounts for approximately 47% of the 182,000 heart surgeries performed annually in the United States, a market valued at $7.2 billion. We analyzed color Doppler images of patients with and without carotid artery obstruction. From the image provided by the team, the turbulent zones were detected and a box-counting method was implemented to measure their fractal dimension frame by frame, that is, to measure the coverage of the plane by turbulent zones. Our analysis allowed us to obtain images with turbulent zones with dimension D in about 400 frames per study. D fluctuates between 0 and 1, oscillating smoothly with the cardiac cycle, with higher values in patients with obstructive atherosclerotic plaque.