Kidney Stone and cyst Detection using Deep learning, CNN and AutoKeras

Kidney Stone Detection Using CNN and Autokeras

Millions globally suffer from kidney stones, emphasizing the need for early identification. Convolutional neural networks (CNNs) have shown promise in medical imaging, facilitated by AutoKeras. This work proposes a CNN and AutoKeras-based kidney stone detection model. It uses medical images for training and testing, achieving superior accuracy and precision. Kidney stones, solid deposits causing pain and complications, can be diagnosed through medical history, examinations, and imaging tests like ultrasound, CT scans, or X-rays. Untreated kidney stones lead to infections, kidney damage, or failure, sometimes requiring surgery.

Project Overview: The project aims to develop a deep learning model for kidney stone classification using CNN and Autokeras. Ultrasound images are categorized into normal, cyst, tumor, and stone.

Project Aim: A CNN model is trained with an extensive ultrasound dataset, enabling it to recognize patterns linked to each category. Autokeras optimizes model hyperparameters, ensuring optimal kidney stone detection and classification.

Problem Definition: The problem is to have the CNN model categorize kidney stones into Normal, Cyst, Tumor, or Stone, using the given dataset.

Problem Statement: Using CT Kidney Dataset images, the study aims to create a model that categorizes kidney stones into these four categories. An interface in the form of a web application enables users to utilize the model's capabilities. The web app takes a user's CT scan, predicts the condition represented by the scan, and displays results to the user.

Project comes with Code,report,data,PPT and web app