Field of study
Most of my projects are inspired by healthcare problems and are designed to help the diagnose and monitoring of different conditions in different areas, such as cardiology, radiology and production of healthcare equipment. Other projects showed aim to assist companies to solve issues with bottlenecks in their processes or as generic models to accelerate prototyping and developemnt.
ECG signal classification
This project is my final paper at college and utilizes a convolutional neural network 1D made from scratch to classify different timeseries (ECG raw signal from "PTB-XL" database) between 5 categories, achieving a performace of 84% of accuracy and recall across different architectures, such as Resnet-50 and ensemble models trained with the same method.
LLM4Trademark - Brand and Product Classification for the automation of trademark registration
This project was made as a solution for the Technology Innovation Office (TIO or Technology Transfer Office (TTO)) of Biopark Educação (Brazil) to adress the current bottleneck in the process of brand and product patent depositing and trademark registering. The usual process involves searching in the Nice classification and Viena classification documents distributed by INPI (National Institute of Research and innovation) which consumes precious time. With a description of the product or brand the solution implemented can rapidly search through the documents and answer with the most probable categories fitted for the brand or product, saving precious time for the TIO team at the university.
Biossensor quality inspection
This project is a classifier of non-conformities for test strips for the company Biosens. Using an algorithm based on Inception V3, the model had the task to classify the images among 5 different categories and achieved a result of 88% of accuracy in the test with 16 images never seen before (14 correct out of 16). All sensors that were suitable for prodution followed to the next propduciton steps. The 2 missclassified images were still classified as errors and did not interfere with the production.
Chest X-ray classification
Using a convolutional neural network to classify 4 chest X-ray conditions of the dataset "COVID-19 Radiography Database" from Kaggle, The dataset was trained using different configurations of Resnet-50 and A CNN made from scratch and achieved an acuracy above 90%.
Other projects
Here you can find my other projects that can be used for study or as base for different problems and projects in a fast way to prototipe solutions and test theories.
Personal Summary
I'm a Biomedical engineer and an Applied Computing Master's student. I have 2 years of experience in research at the Startup Biosens Tech, working with manufacturing processses of microfluidic channels of rapid tests for several conditions and a data science project for inspection of production materials. Currently, I'm working at Biopark Educação as an R&D analyst, responsible for R&D KPIs, report automation (involving ETL, data analysis, python developing) and maintenence for current software used in the sector. Academically I have projects in the biomedical image and signal processing and many small projects that can be used as basic generic models for diverse problems for fast prototiping and testing.