International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 13 Issue: 02 | Feb 2026
p-ISSN: 2395-0072
www.irjet.net
AI-DRIVEN BRAIN TUMOR DETECTION, SEGMENTATION AND PERSONAL GUIDANCE USING LLM Mrs.D.Kavitha1, D.Chandana2, K.Ganesh3, Ch.Gyanadeep4, J.Abhishek reddy5 1Asst.Professor in Department of IT, TKR College of Engineering and Technology, Telangana, India 2345BTECH Students in Department of IT, TKR College of Engineering and Technology, Telangana, India
----------------------------------------------------------------------***---------------------------------------------------------------------------
Abstract - Brain tumour detection is a critical task in
accurate classification of tumour and non-tumour regions [3], [8], [10].
medical diagnostics, requiring accurate and timely identification to ensure effective treatment. This project presents an automated system for detecting and segmenting brain tumours from MRI images using deep learning and image processing techniques. A pre-trained convolutional neural network model is employed to classify MRI scans as tumour-positive or tumour-negative. Tumor regions are then segmented using morphological operations and the watershed algorithm for precise visualization. Additionally, the system integrates a Generative AI assistant (Google Gemini via Lang Chain) to provide patients with personalized medical advice, treatment suggestions, and lifestyle precautions based on the detection results. The system also includes a webbased platform for user registration, secure login, and admincontrolled account activation, ensuring controlled access.Experimental results demonstrate accurate tumour detection and effective integration of AI- driven guidance, highlighting the potential of combining deep learning and generative AI for enhanced medical diagnostics.
In addition to classification, tumour localization and segmentation play a vital role in understanding tumour shape, size, and spread. Image processing techniques such as morphological operations and watershed algorithms are widely used for precise tumour region segmentation and visualization [2], [9]. Furthermore, the integration of Generative Artificial Intelligence (AI) models enables intelligent interaction with patients by providing personalized medical guidance, treatment suggestions, and precautionary recommendations based on detected tumour conditions [6], [10]. By combining MRI-based tumour detection using CNNs, automated segmentation techniques, and AI-assisted recommendation systems into a web-based platform, the proposed system offers a comprehensive solution for early diagnosis, enhanced clinical decision support, and improved patient awareness while maintaining administrative control over user management [1], [5], [9].
KEYWORDS: Brain Tumour Detection, MRI, Deep Learning, Image Segmentation, Generative AI, Google Gemini, Lang Chain, Web-Based Medical System.
1.1 MRI-Based Brain Tumour Detection and Automated Segmentation
1. INTRODUCTION
Magnetic Resonance Imaging (MRI) plays a vital role in the detection of brain tumours due to its high spatial resolution and excellent soft-tissue contrast. In the proposed system, MRI scans uploaded by users are first pre-processed to remove noise, normalize intensity levels, and enhance image quality. These preprocessing steps ensure that the relevant tumour features are preserved for accurate analysis.
Brain tumours are abnormal growths of cells within the brain that can lead to severe health complications if not detected early. Accurate diagnosis of brain tumours is crucial for effective treatment planning and improving patient survival rates [5], [7]. Magnetic Resonance Imaging (MRI) is a widely used noninvasive imaging technique that provides high-resolution visualization of brain tissues, making it the preferred modality for brain tumour detection [2], [5]. However, manual examination of MRI scans is time-consuming, highly dependent on expert radiologists, and susceptible to human error, particularly in early-stage tumour identification [4], [7].
A Convolutional Neural Network (CNN) is employed to automatically analyse MRI images and classify them as tumour-positive or tumour-negative. CNNs are well-suited for this task as they can automatically learn hierarchical features such as edges, textures, and complex tumour patterns directly from the image data. Once a tumour is detected, image segmentation techniques are applied to precisely identify the tumour region.
Recent advancements in deep learning and medical image processing have enabled automated brain tumour detection and segmentation, significantly improving diagnostic accuracy and efficiency [1], [6], [9]. Convolutional Neural Networks (CNNs) have demonstrated strong capability in learning complex spatial features from MRI images, allowing
© 2026, IRJET
|
Impact Factor value: 8.315
Morphological operations combined with the watershed algorithm are used to segment the tumour area from surrounding brain tissues. This segmentation process highlights the size, shape, and location of the tumour,
|
ISO 9001:2008 Certified Journal
|
Page 199