Dated: 2024-2025

Project Title: Gesture Recognition System For Robotic Hand Manipulation  

Description:

This project explores the use of gesture recognition and robotic manipulation in biomedical applications, demonstrating how computer vision and AI can bridge the gap between human intent and robotic control. Inspired by advancements in machine learning and human-computer interaction, the system uses MediaPipe Hands to detect hand gestures and translate them into precise robotic movements.

Through a real-time processing pipeline, the robotic hand mimics natural finger movements, showcasing its potential in prosthetics, rehabilitation, and assistive robotics. This study highlights the impact of intuitive robotic interfaces in medical and industrial applications, offering an innovative approach to gesture-based control.

Key Features:

• AI-powered gesture recognition using MediaPipe and OpenCV

• Real-time robotic hand control via Arduino and servo motors

• Potential applications in prosthetics, rehabilitation, and assistive technology

• Demonstration at MedFest 2025, showcasing hands-on interaction with robotics

Date: 2025–2026

Project Title: Hand Gesture Classification Using Surface EMG Signals

Description: 

In this project, I focused on the development of an intelligent system for recognizing hand gestures based on surface electromyographic (sEMG) signals. The study addressed the challenge of accurately interpreting muscular electrical activity in order to classify hand movements in a robust and subject-independent manner.

Key Features:

• Biomedical Signal Processing: Implementation of sliding window segmentation, signal rectification, and low-pass Butterworth filtering to extract the muscle activation envelope and reduce noise.

• Time–Frequency Representation: Generation of spectrograms using Short-Time Fourier Transform (STFT), enabling transformation of multichannel sEMG signals into structured time–frequency tensors suitable for convolutional processing.

• Deep Learning-Based Feature Extraction: Design and implementation of a CNN architecture with parallel convolutional blocks and residual connections to automatically extract relevant spatial features from spectrogram representations.

• Temporal Modeling with LSTM: Integration of a Long Short-Term Memory (LSTM) layer to capture temporal dependencies between consecutive signal windows, improving gesture discrimination in dynamic conditions.

• Post-Processing for Prediction Stabilization: Application of a temporal smoothing algorithm to eliminate isolated misclassifications and increase recognition consistency in real-time scenarios.

• Dataset Utilization: Evaluation performed on the NinaPro DB5 database using multichannel sEMG signals collected with dual Myo armbands, focusing on a representative subset of gestures to ensure computational feasibility while maintaining functional relevance.

• Performance Evaluation: Analysis based on both classification accuracy and temporal recognition accuracy, highlighting the importance of stability in real-time gesture-based control systems.

Date: 2025–2026

Project Title: Monitoring the Clinical Evolution of Post-COVID Patients Using Wearable Devices

Description: 

In this project, I focused on analyzing the clinical evolution of post-COVID patients through physiological data collected by wearable devices. The study addressed the challenge of continuously monitoring autonomic and cardiovascular function outside the hospital environment, using real-world data acquired via photoplethysmography (PPG) and consumer-grade wearables.

Key Features:

• Analysis of the HRV-COVID19 dataset (multimodal wearable data).

• Evaluation of time- and frequency-domain HRV parameters (SDNN, RMSSD, LF/HF).

• Longitudinal cohort selection (85 participants, 1723 daily recordings).

• Implementation of K-Means clustering to identify physiological subgroups.

• Assessment of wearables as a non-invasive remote patient monitoring tool.

Date: 2025–2026

Project Title: Seismocardiographic Signal Processing for Cardiac Pulse Estimation

Description: 

This project focused on the processing and analysis of seismocardiographic (SCG) signals in order to estimate heart rate from thoracic acceleration measurements. The study addressed the challenge of extracting meaningful biomechanical cardiac vibrations from noisy acceleration data acquired using a MEMS sensor.

Key Features:

• Designed 5–45 Hz band-pass filter to isolate cardiac components.

• Implemented full signal processing pipeline in LabVIEW.

• Developed peak detection and temporal filtering algorithm.

• Computed inter-beat intervals and BPM (~120 post-exercise).

• Validated SCG as a non-invasive cardiac monitoring method.

Date: 2025–2026

Project Title: Design of an Educational Platform for the Study of Optical Lenses

Description: 

In this project, I designed a didactic platform for the experimental study of geometrical optics, supporting convergent and divergent lens experiments and validation of the thin lens equation

Key Features:

• Applied Snell’s law, thin lens equation, and magnification analysis.

• Enabled real/virtual image observation and focal distance measurement.

• Developed a mechanical concept for stable and precise optical alignment.

• Integrated modern educational approaches (simulation tools, digital concepts).

Date: 2024-2025 

Project Title: Designing A Myoelectric Hand Prosthesis 

Description: 

In this project, I focused on the design and development of a myoelectric hand prosthesis, specifically a right-hand prosthesis for a patient with unilateral amputation. The study addressed the complex challenge of technically substituting the human hand by integrating principles from functional anatomy, biomechanics, mechanical engineering, and electronics.

Key Features:

• Performed anatomical and anthropometric analysis for functional sizing.

• Designed an antiparallelogram mechanism for a synchronized three-jaw grip.

• Calculated kinematics and torque requirements for finger motion.

• Selected DC motor (Portescap 16G88) with planetary and worm gear reducers for compact, self-locking actuation.

• Optimized component integration for ergonomic and lightweight design.

Date: 2024–2025

Project Title: Identifying Pulmonary Lesions from Thoracic CT Scans Using Classical Image Processing

Description:

This project focused on detecting lung lesions from axial CT scans by implementing classical image processing techniques on a publicly available Kaggle dataset containing of thoracic CT images (normal and with different types of lung cancer). The project aimed to isolate and visualize pulmonary nodules through contrast enhancement, binarization, and morphological operations.

The goal was to evaluate the performance of non-AI-based methods for medical image segmentation, particularly in conditions where lesions were clearly delimited from surrounding anatomical structures. The study highlighted both the potential and the limitations of classical approaches, showing their reliance on the clarity of lesion boundaries and local contrast.

Key Features:

• Stepwise image preprocessing: piecewise linear contrast adjustment, contrast stretching, clipping, binarization

• Morphological processing: erosion, dilation, interior contour extraction, and morphological gradient

• Identification and segmentation of nodules using Python and OpenCV-based techniques

• Comparison of segmentation effectiveness across different preprocessing strategies

Date: 2024-2025

Project Title: Glycemic Analysis and Wearable Sensor Data Integration for Early Prediabetes Detection

Description:

This project investigated the physiological and demographic correlations related to postprandial glycemia using real-world data from wearable sensors, sourced from the BIG IDEAs Lab (PhysioNet). The aim was to explore whether patterns in glucose responses and nutritional input could serve as early indicators of prediabetes risk.

A statistical framework was applied to compare HbA1c values by sex and to assess postprandial glucose responses in relation to carbohydrate and caloric intake. The project used Python for preprocessing, correlation analysis, and visualization, offering a data-driven perspective on metabolic variability and the potential of wearables for passive health monitoring.

Key Features:

• Analysis of HbA1c variation by sex and its statistical significance (p = 0.84)

• Pearson correlation between carbohydrate intake and 2h postprandial glucose levels (r = 0.35, p = 0.002)

• No significant correlation found between caloric intake and postprandial glucose (r = 0.14, p = 0.23)

Dated: 2023-2024

Project Title: Management System for a Medical Equipment Company

Description:

This project involves the development of a comprehensive database management system for a company that sells medical equipment. The system is designed to facilitate the management of orders, product inventory, client information, and employee records. It includes the creation of a relational database using SQLite, with entities such as products, clients, orders, departments, and employees. The project also implements automated features like stock updates after each order and the recalculation of average departmental salaries upon hiring new employees.

Key Features:

• Creation and management of a relational database using SQLite

• Automated stock level updates and salary adjustments

• Custom SQL queries to analyze business performance, including monthly revenue and client purchase behaviors

• Implementation of data integrity checks through SQL triggers to ensure accurate and reliable data entry

Dated: 2023-2024

Project Title: Exploring the Role of Robots in Advancing Psychotherapy

Description:

This project delves into the innovative use of robots in the field of psychotherapy, examining how robotic systems can enhance therapeutic practices. The study explores various applications of robots, such as the AI-driven platform Woebot and the QTrobot, particularly in addressing psychological conditions like anxiety, autism, and dementia. Through a detailed case study, we evaluated the effectiveness of a social assistance robot named "Blossom," powered by a large language model (LLM), in delivering cognitive behavioral therapy (CBT) to university students.

Key Features:

• Investigation into AI and robotics in mental health care

• Case study on the use of the Blossom robot for CBT

• Applications of robotic therapy for conditions such as autism and dementia

Dated: 2023-2024

Project Title: Finite State Machine for Patient Triage at Home

Description:

This project explores the development of a Finite State Machine (FSM) designed to assist in the home-based triage of patients. The FSM guides patients from their initial state of illness to receiving appropriate treatment, using a combination of accessible devices and an AI-powered mobile application. The system evaluates vital signs and overall health to direct patients to either a general practitioner or emergency services, depending on the urgency of their condition.

Key components of the project include the design of a deterministic FSM, the integration of an AI algorithm for initial patient assessment, and the implementation of the system using OrCAD for detailed simulation and analysis. This project demonstrates the potential of combining FSMs and AI to improve access to medical care and optimize response times in critical situations.

Key Features:

• Development of a deterministic Finite State Machine for patient triage

• AI integration for initial health assessment and decision-making

• Implementation and simulation in OrCAD for system optimization

• Potential applications in improving healthcare access and response times

Date: 2023-2024

Project Title: Anthropometric and Kinematic Analysis of the Lower Limb

Description:

In this project, I explored the biomechanical analysis of the lower limb, focusing on understanding the mechanics of human movement. By combining tools like GeoGebra Classic 6 with traditional mathematical methods, I analyzed key factors such as the center of gravity and kinematic angles at the knee and ankle. This study has practical applications in sports medicine and rehabilitation, offering insights into improving performance and developing personalized rehabilitation strategies.

Key Features:

• In-depth analysis of lower limb biomechanics

• Use of advanced software for data modeling and visualization

• Practical applications in sports and medical fields

Dated: 2023-2024

Project Title: Cerebrospinal Fluid: Properties and Flow Dynamics

Description:

This project focuses on the cerebrospinal fluid (CSF), an essential component of the central nervous system. CSF plays a critical role in protecting the brain and spinal cord, regulating intracranial pressure, and maintaining chemical homeostasis. The study delves into the physical and rheological properties of CSF, examining its behavior as a Newtonian, incompressible fluid. By applying the Navier-Stokes equation, the project models the flow of CSF through the cerebral aqueduct, providing insights into the laminar flow characteristics and pressure dynamics within this vital system.

Key Features:

• Analysis of CSF's physical and rheological properties

• Application of the Navier-Stokes equation to model CSF flow

• Exploration of pressure and flow dynamics in the cerebral aqueduct

• Use of dimensional analysis and Reynolds number to characterize the flow regime