Biomedical Research – Measurement Group

We conduct cutting-edge research in the biomedical field, our focus spans from advancing medical image processing algorithms also using machine learning and deep learning to the development of chemical sensors and precise position tracking.

In collaboration with prestigious clinics and research hospitals like Casa Sollievo della Sofferenza, we have conducted a comprehensive investigation into the optimization of radiation doses specifically tailored for X-ray and computed tomography applications. Our emphasis lies in maintaining high image quality, ensuring a fruitful partnership between technology and healthcare.

About sensor innovation, we are currently pioneering nanostructured sensors designed for ion and drug detection. We accurately characterize their performance, utilizing machine learning for multi-drug discrimination. This collaborative effort extends to École Polytechnique Fédérale de Lausanne (EPFL).

Our exploration extends to the field of electromagnetic tracking systems for surgical instrument localization in collaboration with Masmec S.p.A. – Biomed division. Our goal is to enhance their operational range, and to achieve this, we have not only analyzed their performance but have also created a simulation tool. This tool predicts tracking system accuracy under various geometric configurations of field sources and operating conditions.

More details about our latest Research activities:

Photoplethysmography (PPG) for telemedicine applications

We are deploying a cuff-less wireless photoplethysmograph to obtain non-invasive measurements of blood pressure. The focus is on the software design and implementation using machine learning and deep learning. Recently an acquisition system, that includes the photoplethysmography sensor and a microcontroller, has been developed.

Ultrasonic arrays for contactless levitation and haptics

Ultrasound phased-arrays (UPAs) generating a pressure pattern enables the levitation of small particles and objects in air, e.g., blood droplets, drugs, and chemical powder. UPAs also allow to produce haptic sensation on the user’s hand, with several impactful biomedical applications like surgeons exploring a CT scan with haptic feedback to feel tumors, training cardiologists to search for a arterial or venous pulse, or in combination with AR/VR technology.

Glaucoma

This research, conducted in collaboration with the University of Maryland, aims to revolutionize diagnostic methodologies for glaucoma, a leading cause of visual impairment. The study unveils a promising avenue for early detection by harnessing the pulsatile corneal deformations evoked by natural eye movements, captured through advanced RGB imaging.

Omics

This research aims to develop clinical machine-learning algorithms for electronic health records (EHR) and -omics data, intending their implementation in a real-world clinical environment. Specifically, machine and deep learning algorithms will be utilized to generate a catalog of gene expression and co-expression predictors associated with various pathologies, with a primary focus on tumors. The study will leverage extensive datasets, encompassing genomics, transcriptomics, radiomics, and clinical data. Special attention is given to the segmentation of medical images related to patients with different types of tumors, enabling the extraction of radiomics data.

ECG segmentation

This activity is focused on the processing of ECG signals acquired by a low-cost single lead heart rate monitor, to improve the segmentation of ECG waves, to give to clinicians accurate temporal information for diagnostic purposes.