Marco Antonio Pinto-Orellana

Dissertation: “Time-spectral modeling of biomedical signals.” Advisor: Prof. Hugo Hammer.

Dissertation topic: “Machine learning, signal processing and electrical signals.” Advisor: Prof. Fabio Cerqueira.

Project: “Communication satellites, and attitude and orbit control mathematical modeling.” Advisor: Prof. Zhang Shuhua.

Dissertation: “Wireless communications and brain signals.” Advisor: Prof. Ramiro Fuentes.

• Develop causal cross-frequency models to improve detection of electro-hemodynamic dynamics.
• Redesigned the class CGNS410: Imaging and Neuroscience to integrate state-of-art statistical models into neuroscience interpretation
• Redesigned the class CGNS421: Neuroscience and Artificial Intelligence integrating the design of a mini-brain as part of the course
• Developed content focusing on hands-on experiences to promote experiential learning.

• Developed a high-performance statistical hypothesis test for phase-amplitude coupling that reduced the uncertainty estimation time from 2.5 s to 3.75 ms.
• Developed a novel connectivity method that predicts seizure-generating tissue with a ROC-AUC higher than 80% across several datasets.
• Proposed a model, based on reinforcement learning in game theory, that explains 80% of seizure-related abnormalities.

• Increased brain connectivity prediction accuracy by 10% with large-scale Bayesian models.
• Implemented and optimized signal processing and AI algorithms as statsmodels/sklearn-compatible Python packages.
• Developed 2D/3D visualization tools for brain signals in WebGL (JS, PixiJS, Three.js).

• Developed real-time causality analysis models (16 ms delay) using Kalman filters and coding theory.
• Designed cross-platform applications for complex statistical models using MATLAB and Python.

• Co-led a project funded with a 5 MM NOK research grant from Innovasjon Norge.
• Developed a high-performance discrete optimization engine in C# and Python.
• Integrated large ML modules (>70K CLOC) into a flexible, strongly typed backend system.
• Reduced the execution time of a 2D routing system from 1.5 hours to 2 minutes.

• Improved the UX and business logic of a CRM system built with Spring and Vaadin (Java).

Patient-Centric Engineering in Rehabilitation Project, No. 273599.

9th of 1,742 applicants; youngest selected candidate (21 years old).

1st of 5 selected awards; youngest selected candidate (15 years old).