Ederson Dorilêo

Interests (Goals): Medical Image Analysis; Computer Vision, Artificial Intelligence, Data Science 

Strengths: Pattern recognition; Image and data analysis: (segmentation, quantification & registration); Machine Learning (Multi-layer perceptron); Computer Vision (detection, tracking, stereo vision); adaptive and predictive modeling; Medical Robots

Tools: OpenCV; Tensorflow; PyTorch; SVM; Kalman (smoothing/prediction); CBIR(Content-based Image Retrieval); 

Programming: Python/Numpy, MySQL, C/C++, Java. Principles: Matlab, Qt, R, OpenGL, VTK, ITK, Prolog, DICOM.

Research Fellow (AI & Radiomics) - MAInLab – Medical Artificial Intelligence Lab– USP-RP, Brazil.

• Research in the domain Artificial Intelligence (AI), Medical Image Analysis and Radiomics (Quantitative Medicine) exploring machine learning models (for Radiology image analysis as a decision support tool) in the context of prognostic prediction modeling and computer-aided diagnosis (CAD) of Idiopathic Lung Diseases.
• Relevance: AI-based radiomic approach using high-resolution CT DICOM images for differential diagnosis of idiopathic pulmonary fibrosis.
• Approach: Supervised SVM (Support Vector Machine) classifier (scientifical journal published); 
• Future Work Motivation: Unsupervised anomaly detection with Generative Adversarial Networks (GANs)
• Keywords: Artificial Intelligence, Radiomics, Medical Image Analysis, (Supervised) SVM, Unsupervised Learning

Doctoral Studies: Medical Robotics, University of Montpellier, Montpellier - France (2012-2015)

Title: “Modeling and Adaptive Needle Path Planning in CT/MRI-compatible Robot-Assisted Interventions”

Master degree: Master in Sciences (Medical Image Analysis & Pattern Recognition)

School of Medicine - University of Sao Paulo (USP-RP), Ribeirao Preto, Brazil (2007-2009)

Title: “Indexing and Content-Based Retrieval in Skin Ulcers Color Images”

Bachelor degree: Biomedical Informatics

University of Sao Paulo (USP-RP), Ribeirao Preto, Brazil (2002-2006)