The Importance and Scientific Outcomes of Data Augmentation Methods in Small-Scale Medical Datasets

One of the major challenges in developing intelligent medical systems is the lack of high-quality and sufficiently large datasets. This paper provides an in-depth analysis of the effectiveness of Data Augmentation (DA) techniques when applied to small-scale medical datasets. During the study, classical geometric transformations and Generative Adversarial Network (GAN)-based methods were compared, and their impact on model performance metrics such as Accuracy and F1-Score was mathematically and experimentally justified. The results demonstrate that data augmentation significantly improves model generalization and reduces overfitting in data-scarce medical applications [1][2].