divya196 / radiologist_expert_ai_application Goto Github PK

###Radiologist Expert: Gemini Pro Vision LLM

Project Overview

Recently my mom met with an accident,we took CT Scan for her and waiting for long time to meet the doctor,untill that i was searching google for each terminologies mentioned in report as I'm a biology student in my 10th standard but pursed information technology as my bachelors,so my family members will share report and ask me the second opinion as like i'm an expert.....So i took this as a idea to my project, I created an application using Google generative AI [GEMINI PRO VISION] and built an application . Usually We'll get the scan report from the centre and consult doctor accordingly based on their availability... So to make ease of this process, I built an application called Radiology Expert which answer about the scan image .

Objectives

To provide an intuitive tool for diagnosis.

To leverage advanced AI technology for analyzing and providing feedback on medical imaging.

To offer a user-friendly interface that simplifies the process.

Features

Scan Image Upload: Users can upload their medical imaging in DICOM,JPEG,JPG,PNG format.

Any Questions: A text input field allows users to ask questions, you will have your Radiologist expert right in front of you.

AI-Powered Analysis: Utilizing Gemini AI, the application provides a detailed analysis of the scan image in context with the questions asked.

Technologies Used

Streamlit: For creating the web application interface.

Google Generative AI (Gemini Pro Vision): For processing and analyzing the image content.

Python: The primary programming language used for backend development.

PDF2Image & PIL: For handling PDF file conversions and image processing.

Challenges Faced

Integration with Gemini AI: Ensuring seamless communication between the Streamlit interface and Gemini AI model.

User Experience Optimization: Creating an intuitive and responsive UI.

Future Enhancements

Support for Multiple scan images: Extend the functionality to handle multi-images.

Customizable Feedback Categories: Allow users to choose specific areas for feedback.

Interactive conversation: Integrate a feature to interact directly based on the AI's suggestions.

Enhanced Error Handling: Improve the system's robustness in handling various file formats and user inputs.

Conclusion The Radiologist Expert Streamlit application stands as a significant tool in bridging the gap. By harnessing the power of AI, it provides valuable insights and recommendations, making it a pivotal step in enhancing the understanding process.