Generative Fabrication of Medical Images for Machine Learning Training

• Andres G. Calzada-Jasso CICESE Research Center
• Andrei Tchernykh CICESE Research Center / Institute for System Programming
• Ixchel D. Avendaño-Pacheco CICESE Research Center
• Jorge M. Cortés-Mendoza National College of Ireland
• Bernardo Pulido-Gaytan National College of Ireland
• Mikhail Babenko North-Caucasus Federal University
• Alfredo Goldman USP
• Horacio González-Vélez National College of Ireland

Resumo

Training in supervised machine learning is based on the availability of datasets; however, medical datasets must comply with stringent privacy regulations. Generative Adversarial Networks (GANs) are a relevant alternative to solve the limitation of small medical datasets due to their ability to generate additional data with desired features. A significant drawback of these models is that they may produce unrealistic, blurred, or insufficiently diverse images. This paper proposes a data augmentation technique using GANs to create synthetic Magnetic Resonance Imaging (MRI) of four stages of Alzheimer's Disease (AD): non-demented, very mild demented, mild demented, and moderate demented. We designed a GAN based on the Pix2Pix model, which learns the features of each AD stage. Generated images are evaluated by multistage Convolutional Neural Network (CNN) models, greyscale histograms of the distribution of pixel intensities, and brain mass measurements on binarized images. The results indicate that AD synthetic MRI effectively captures disease patterns, demonstrating the potential of GANs to improve training and diagnosis of neurodegenerative diseases.

Palavras-chave: Training, Histograms, Magnetic resonance imaging, Machine learning, Generative adversarial networks, Data augmentation, Brain modeling, Convolutional neural networks, Alzheimer's disease, Medical diagnostic imaging, Alzheimer's Disease, Binarization, Convolutional Neural Network, Data Augmentation, Generative Adversarial Networks