Comparing Forward-Forward and Backpropagation in U-Net for Melanoma Image Classification

In recent years, deep neural networks have become essential in medical imaging, especially for precise diagnostic applications. This paper compares two main learning meth-ods for neural networks-Forward-Forward and Backpropagation-focused on the U - N et architecture for classifying derma-tological images, specifically melanomas. The Forward-Forward approach, which sidesteps traditional gradient-based Backpropagation in favor of a simpler, unidirectional process, offers a more computationally efficient alternative. In contrast, Backpropagation is a well-established method for optimizing network weights, especially for complex tasks where high accuracy is crucial. We trained U-Net models on a dataset of melanoma images, evaluating both their computational and diagnostic performance. The findings show that while Backpropagation achieves higher accuracy and precision, the Forward-Forward method stands out in computational efficiency, making it valuable in resource-limited settings. This study highlights the balance between computational speed and diagnostic accuracy, suggesting potential ways to optimize neural networks for medical diagnostics.