Muainishi Wa Saratani is an advanced, AI-powered web application designed for bio-informaticists and healthcare practitioners. It classifies oncological samples by analyzing a combination of genomic data, clinical notes, patient history, and medical imaging reports. The tool provides detailed analyses, data visualizations, and web-grounded information to support and accelerate clinical decision-making.
- Multi-Modal Data Analysis: Accepts a variety of inputs including clinical text, patient history, and file uploads for gene expression data (.csv, .tsv, .txt) and imaging reports.
- AI-Powered Classification: Utilizes the Gemini API to deliver a primary cancer classification, complete with a confidence score and a detailed AI-generated summary.
- Intelligent File Parsing: Automatically extracts patient demographics and vitals from various hospital system file formats, including JSON, PDF, DOCX, CSV, and XLSX.
- In-Depth Analytics:
- Displays a list of the top potential cancer classifications with their respective probabilities.
- Identifies and visualizes the most influential genetic and clinical markers that contributed to the diagnosis.
- Rational Treatment Options: Generates a list of evidence-based treatment options (Standard of Care, Emerging, Clinical Trial) tailored to the patient's prognosis.
- Web-Grounded Information: Fetches up-to-date information, recent studies, and treatment protocols for the classified cancer type using Google Search grounding, with citations.
- Interactive Visualizations: Presents complex data through clear, easy-to-understand charts for probabilities and marker importance.
- Responsive & Modern UI: Built with a clean, responsive interface using Tailwind CSS for a seamless experience on any device.
- Frontend: React, TypeScript
- AI Model: Google Gemini API (
gemini-2.5-flash) - Styling: Tailwind CSS
- Charting: Recharts
- Bundling/Dependencies: No build step required; uses modern browser features like ES modules and
importmap.
This project is a pure client-side application and does not require a build step or a local server to run.
- A modern web browser (e.g., Chrome, Firefox, Safari).
- A valid Google Gemini API Key.
The application requires a Google Gemini API key to function. This key must be available as an environment variable process.env.API_KEY in the execution context.
For local development or testing without a build system, you can set this up by adding a script tag to index.html before the main application script is imported.
-
Open
index.html. -
In the
<head>section, add the following script tag, replacing"YOUR_API_KEY_HERE"with your actual key:<script> // WARNING: For development or restricted environments only. // Do not expose your API key publicly. window.process = { env: { API_KEY: 'YOUR_API_KEY_HERE' } }; </script>
Security Note: This method exposes your API key in the client-side code. Ensure your API key has appropriate restrictions and is never committed to a public repository. For production, use a secure method like a backend proxy or a build system to manage secrets.
Simply open the index.html file in your web browser.
(Please replace the placeholder images in the /screenshots directory with actual screenshots of your application.)
The user-friendly form allows for inputting patient history and uploading various file types. Note the uploader for demographics, which accepts multiple formats.
After analysis, the app displays the primary classification, confidence score, and a detailed summary.
Interactive charts show the top classification probabilities and the importance of various biological markers.
Evidence-based treatment options are presented with clear visual cues for their evidence level (Standard of Care, Emerging, etc.).
Users can fetch up-to-date, web-grounded information about the diagnosis, complete with source links.
This application is intended for research and informational purposes only. It is not a substitute for professional medical advice, diagnosis, or treatment. All analyses and suggestions generated by this tool must be verified and interpreted by a qualified healthcare professional.