Using AI for Medical Report Analysis

Using AI for Medical Report Analysis

I used Google Gemini, Langchain and Llama 3 to turn a medical report into an action plan for a patient. I couldn’t believe how good of a job they did, and it only took me one hour. Here’s how I did it:

Gemini

I started with a picture of a medical report (it’s a sample from Quest Diagnostics, not mine).

This is a lipid panel (used to test for heart disease) and covers other heart-related risks.

I then gave that report to Google Gemini and asked it to evaluate it.

I understood about zero of it, but I assume it was great.

It gave a great rundown of the report. It talked to me like I was a doctor, which was pretty helpful for the next part.

Llama 3

I then asked it for potential risk-factors based on the report and had it summarized everything in one list. I took the summary and used Langchain and Ollama to prompt Llama 3 with it.

This step was pretty important because Gemini and other cloud-based models refuse to give medical advice. They’re worried about pesky things like “liability” and want you to “ask a medical professional.”

But with Llama 3, I could just ask the AI for potential risk factors and course of action.

Results

This did an amazing job!

I did not expect it to be this good at interpreting the data.

This is great because it gives you someone to help you with the basics of interpreting health data at your beck and call.

Of course we’d all rather discuss things with medical professionals. But sometimes, I’d like to have some questions answered in seconds rather than days of phone-tag or scheduling.

Follow Xavier Gray on Twitter and LinkedIn for more demos like this!

Recent Posts