The paper uses GAN-based technique as an innovational method toward enhancing a joint tumor detection brain by itself through the creation of added MRI images under environmental noisy settings like Gaussian noise, salt-and-pepper, and speckle for MRI. The approach addresses the key limitations observed in prior methods. Conventional tumor detection models often struggled with handling noisy MRI data, which compromised both the clarity of images and the accuracy of diagnosis. The inability to simultaneously enhance and detect tumors in a unified pipeline posed significant challenges in clinical workflows. By leveraging GANs, our model overcomes these issues, demonstrating robustness against various noise types while maintaining high diagnostic accuracy. These developments contribute to more reliable and accurate diagnostic results in real-world, noisy clinical imaging scenarios. The noise-injection dataset was used to denoise the GAN, utilizing its strong generative capability to reconstruct lost clarity in an image; pre-processing thus strengthens the robustness of our model by making it capable to handle noisy data, as common in clinical imaging scenarios where ideal conditions cannot always be available. After denoising, the GAN also assisted in tumor detection with an accuracy of 96.75%. A primary strength of our approach is that both the enhancing and the detection functionalities can be considered in one pipeline. This paper presents a widely promising result toward improved diagnostic accuracy in the case of medical imaging workflows by exploiting the advantage of GANs in reconstructing high-quality MRI images while identifying noise affected brain tumors accurately, which is a significant advancement in medical imaging for early and precise diagnosis of tumors.