Indholdsfortegnelse: :: Library Catalog

Saved in:

Bibliografiske detaljer
Main Authors:	Mr.Gulve Rushikesh Somnath, Mr.Hase Onkar Balasaheb, Mr.Jadhav Pranav Prashant, Mr.Warghude Rushikesh Sudhakar, Ms.K. T. Bhandwalkar
Format:	Recurso digital
Sprog:	engelsk
Udgivet:	Zenodo 2026
Fag:	brain tumor detection, Alzheimer's disease classification, Modified CNN, EfficientNetB0, transfer learning, MRI analysis, Flask, Django, deep learning, clinical decision support, neurodiagnostics
Online adgang:	https://doi.org/10.5281/zenodo.20201511
Tags:	Tilføj Tag Ingen Tags, Vær først til at tagge denne postø!

Indholdsfortegnelse:

The timely and precise identification of neurological conditions such as brain tumors and Alzheimer’s disease carries profound implications for patient survival, treatment efficacy, and long-term quality of life. This paper introduces NeuroDetect AI, a web-deployable Intelligent Neurodiagnostic Platform that automates MRI-based brain scan classification across brain-tumor-positive, Alzheimer’s-positive, and neurologically normal categories. The system adopts a dual deep learning strategy: a custom Modified Convolutional Neural Network for brain tumor classification and an EfficientNetB0 transfer-learning model for Alzheimer’s detection. A standardized preprocessing pipeline consisting of grayscale conversion, CLAHE, intensity normalization, and augmentation feeds both models. The platform uses a Flask REST API for inference and real-time doctor-patient communication, while a Django-backed module manages authentication, patient records, appointment scheduling, and role-based access control. Evaluation on 6,500 combined MRI scans from the Kaggle Brain Tumor MRI Dataset and ADNI yielded 96.9% accuracy for tumor detection and 95.8% for Alzheimer’s staging, with an average inference latency of 1.3 seconds per scan. The platform integrates confidence-based clinical triage, physician override, PDF report generation, and real-time consultation within a single deployable web application.

Lignende værker