ai medical imaging agent is not able to produce analysis report for images #29

Closed
opened 2025-11-06 14:31:35 -06:00 by GiteaMirror · 1 comment
Owner

Originally created by @abhinav-vimal13 on GitHub (Jan 3, 2025).

https://github.com/Shubhamsaboo/awesome-llm-apps/blob/main/ai_agent_tutorials/ai_medical_imaging_agent/ai_medical_imaging.py

I tried to run above agent for some medical images related to x-ray and ultrasound, but it is not giving any results instead of that it is just giving output: '''Okay, I understand. I'm ready to analyze the medical image you provide. Please share the image with me, and I will follow the structure you've outlined to provide a comprehensive analysis.''' although I am passing image.

please look into this

Originally created by @abhinav-vimal13 on GitHub (Jan 3, 2025). https://github.com/Shubhamsaboo/awesome-llm-apps/blob/main/ai_agent_tutorials/ai_medical_imaging_agent/ai_medical_imaging.py I tried to run above agent for some medical images related to x-ray and ultrasound, but it is not giving any results instead of that it is just giving output: '''Okay, I understand. I'm ready to analyze the medical image you provide. Please share the image with me, and I will follow the structure you've outlined to provide a comprehensive analysis.''' although I am passing image. please look into this
Author
Owner

@Madhuvod commented on GitHub (Jan 19, 2025):

hey @abhinav-vimal13, I would ask you to try again. We used gemini 2.0 flash experimental model here in this demo which recently released on dec 24, since it is an experimental model, the error rate is a little high so it sometimes hallucinates saying that there's no input. here's a note about experimental models from google's docs:

Support for individual experimental models is not permanent or long-term, and can be turned off without prior warning. Experimental models cannot be used in production code.

So yes, try again, it would def work since it did work for me right now:

https://github.com/user-attachments/assets/c65bb04d-5aed-49ae-ad3f-e484b744ee5c

@Madhuvod commented on GitHub (Jan 19, 2025): hey @abhinav-vimal13, I would ask you to try again. We used gemini 2.0 flash experimental model here in this demo which recently released on dec 24, since it is an experimental model, the error rate is a little high so it sometimes hallucinates saying that there's no input. here's a note about experimental models from google's docs: > Support for individual experimental models is not permanent or long-term, and can be turned off without prior warning. Experimental models cannot be used in production code. So yes, try again, it would def work since it did work for me right now: https://github.com/user-attachments/assets/c65bb04d-5aed-49ae-ad3f-e484b744ee5c
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/awesome-llm-apps#29