gitmyhub

Smart-AI-Healthcare

Python ★ 54 updated 26d ago

A real-time AI and Machine Learning based healthcare application that predicts diseases from user symptoms using text, voice, and image inputs. The system supports multilingual communication, severity analysis, diet recommendations, PDF report generation, and nearby hospital navigation for smart medical assistance.

A Python web app that predicts possible diseases from typed, spoken, or image-based symptoms using machine learning, and returns a severity level, dietary advice, precautions, a PDF report, and a map of nearby hospitals.

PythonFastAPIscikit-learnReportLabJavaScriptHTMLCSSsetup: moderatecomplexity 3/5

This is a healthcare web application that tries to predict possible diseases based on symptoms you describe. You can type your symptoms, speak them aloud, or upload an image, and the system uses machine learning models to suggest what condition you might have. The README does not describe the specific models used or how they were trained, only that they rely on symptom and disease datasets.

Alongside the disease prediction, the application provides additional outputs: a severity level for the predicted condition, dietary suggestions, precaution advice, and a PDF report summarizing the results. It also integrates with Google Maps to show nearby hospitals based on your location.

The application supports multiple languages, which the README lists as a feature but does not detail further. Voice input is handled through the browser, and image-based input is mentioned as a way to describe symptoms visually, though the README does not explain how image analysis works internally.

The backend is built with FastAPI, a Python web framework, and the machine learning components use scikit-learn, a common Python library for classification and prediction tasks. The frontend is plain HTML, CSS, and JavaScript. PDF generation uses a Python library called ReportLab.

To run it locally, you clone the repository, install Python dependencies, start the FastAPI server, and open the frontend HTML file in a browser. The README is brief and does not cover model accuracy, data sources in detail, or any limitations of the predictions. This appears to be a student or portfolio project rather than a clinical tool.

Where it fits