Signal processing of physiological signals

- Data analytics for phone app (Mixpanel, AppsFlyer, Google Analytics GA4) and web (GA4), and its integrations with other platforms (Onesignal, mailchimp, Meta). - Conceptualization and design of a data and artificial intelligence model for the early detection of mental health issues during the perinatal period.- Participation on the creation of a scoring system to measure perinatal well-being and its fluctuations.

Signal processing of physiological signals -- namely Heart Rate (HR) and HR variability, Respiration Rate, Blood Pressure, Oxygen Saturation -- from contact (fingertip) and contact-less (face from selfie video) phone video recordings.

As part of the Laboratory of Computational Neuroimaging, I worked on the development of methods for the analysis of Magnetic Resonance Imaging (MRI) data to study how the brain is connected, physically looking at tissue structure with diffusion MRI and functionally looking at patterns of synchronous activity with functional MRI. In particular, I- Developed software FVT4DWI for the interpretation of structural connectivity from diffusion MRI: www.nitrc.org/projects/conductance - Collaborated with other team members and medical doctors, leading to several publications.- Organized seminar series "Why&How" at the Martinos Center for Biomedical Imaging (2018-2020).- Taught at FreeSurfer software local course and assisted with tutorials.- Participated in the acquisition of different MRI modalities in a 7T scanner, including the recording of physiological parameters during resting state fMRI, visual task fMRI and hypercapnia challenge acquisitions.

Statistical Models for the analysis of Arterial Spin Labeling (ASL) and Blood Oxygen Level Dependent (BOLD) Magnetic Resonance Imaging (MRI) modalities to study brain function and disease:- Worked on Bayesian methods for the analysis of BOLD and ASL functional MRI data -- two modalities that look at blood oxygenation changes that occur in a certain brain region when it is active.- Specifically looked at the changes that are related to certain tasks performed during the scanning.- Developed statistical models that helped us in the detection of brain activity with higher specificity.

Hemodynamically informed parcellation of cerebral fMRI data:- Development of a dedicated spatial clustering method to identify homogeneous territories from functional descriptors extracted from functional MRI

Automated selection of local trajectories in diffusion MRI images:- Processing of diffusion MRI images.- Evaluation of different tractography algorithms.- Selection of locally optimal algorithms with machine learning methods.

Support at Systems, Networks and Communications Group.

New Technologies Department: Helpdesk, web maintenance.