A new tool for becoming better self-informed

General
1

Greetings, everyone. I just want to place a little spotlight on a tool that can help a patient to be as well-informed as they can be when interacting with a medical practitioner. It’s a tool that many folks are shy about utilizing, however, because it often comes with some degree of hesitancy that is premised on . I speak of the tool of AI.

Here is how I employ AI to help me to receive the maximum benefit from my interactions with my doctors. And they all have expressed an appreciation of my utilization of it in preparation for my appointments.

Most of my appointments with my doctors consist of going over the test results of a routinely scheduled Ct-Scan or MRI or MRA. And as many of us know from personal experience, the test results are always communicated in technical medical terminology. This is where I call upon an AI for assistance

I will ask an AI or two to “translate the technical language of the Ct-Scan/MRI/MRA reported findings into simpler Layman’s English”. I will then form the questions that I will be asking the physician from the AI’s helpful work product. My questions are always highly relevant to the test results, because I am able to understand them in simpler English.

Keep in mind that I am not asking an AI to replace the role of the doctor in this transaction. Instead, I am asking it to translate a foreign language, medicalese, into one that I can understand.

I also make it a practice to share the AI’s translation with the doctor, coupled with the question, “do you agree with the accuracy of the translation, and would you like to expound on or add anything to it?” All of my docs are very appreciative of how substantive and time saving our interactions have become.

Many doctors today are members of a conglomerate and are placed in the position of adhering to certain time constraints when interacting with a patient. I can often feel those constraints as I place my questions on the doctor’s plate. AI has helped me to maximize the limited time that I have with the doctors I trust.

Be happy to answer any questions on this subject that anyone might have.

2 Likes
2

I find AI to be aggravating most times, but I’ve never thought to ask for easy translation. I have corrected (Alexa), gotten into a debate with her and thank her. I did do the update which is what started the whole aggravating thing. I will get out my last set of images and ask her to translate! I see my NP Neurologist tomorrow, so great timing Peter, thanks!

What AIs do you use? I only know of Alexa…

3

Thank you for the question. It may benefit some readers who are curious about this subject.

Firstly, I never put all my eggs in one AI basket because different AIs offer different qualities of response. Some are more detailed than others. So I tend to use 2-3 AIs at a time and select the most helpful reply. Some of the AIs I use are ChatGPT, Gemini AI, CoPilot AI, Grok AI and Perplexity AI. They all have websites where you can type in your question/query/prompt.

You said something that needs to be clarified by me too. I never ask an AI for an interpretation of an image. That would be placing the AI in the position of a doctor (radiologist, in this case). Instead, I provide the AI with a copy of the written report, direct it to the part titled IMPRESSIONS (usually at the end), and ask it to simplify the contents by translating it into Layman’s English.

My doctors appreciate it when I use AI as a translator/educator to supplement their communications with me. I don’t think that they would appreciate it as much if I provided a competing diagnosis (image interpretation) during our discussion.

Does that make sense? Thanks, again, for the question.

4

Another question then, if you don’t mind. Do you copy the image report straight to these AIs? As you can tell, I’m quite ignorant about them.

5

At some AI websites, you can cut and paste that portion of the written test results that you want the AI to focus its translating powers at and insert it into the body of your question. While at other AI websites, you can include a screensaved image of the written report, either in it entirety or selected portion(s).

And if I can acquire a .pdf copy of the test result from the radiology company that performed the CtScan, MRI or MRA, I will occasionally use that instead and attached it to my question. Since I am a dinosaur of sorts, however, I find myself cutting and pasting a lot. :slight_smile:

6

Thanks Peter! Yes I’ve been playing with two of them. I’ll have to wait a few years for my next images to see what those say. It was helpful, a good bit. They both gave me the same translation. Although I would have still asked my Neurosurgeon to review the images because it didn’t specify seeing my stent and my NP seemed to be as confused as I was, it would have reduced some of the stress I was going through. Thanks again!

7

Think nothing of it, the Divine Miss M (dating myself here). Great album too.

Whenever I am the beneficiary of something, I like to pay it forward and spread the wealth, as they say. And with regard to AIs, it can be a most helpful tool for a patient to become better informed and self-empowered. Big advocate of it, as you can tell.

Regarding the hidden-stent caper, try your hand at placing this conundrum on the plate of an AI. The quality of an AI’s output is always dependent on the quality of the input. Suggested question/prompt:

“If I had a stent (type?) surgically placed at (indicate location) and the findings of a recent MRI/MRA report did not state that it found one there, what are the possible reasons for that to be the case?”

My last MRI, post my endovascular procedure, IDed what it thought was an intracranial stenosis at the neck of my coiled Annie, but my Interventional NeuroRadiologist correctly IDed it as the stent that he placed there.

Curious what an AI would have to say about your hidden stent, aren’t you?

8

Just chiming in here, I personally have not done any of these myself, so I cannot talk about the accuracy, privacy, reliability, authority, and cost concerns (why, yes, I do have a LOT of questions! :rofl:).

However, beyond those that @Peter has mentioned I am also starting to see AI models using LLMs specifically geared towards reading MRIs, thus potentially more accurate. Which makes sense! I mean, I presume my team is using AI to read them, so why can’t I use AI to better understand?!

Separately, I found this AI chat for radiology:

Note, some of these do require uploading images.

However, the other side of me gets into my bug-a-boo as to who is collecting this data behind the scenes, and what is the intent behind the collection? Will my data be for sale? [A simple “yes, it is confidential!” isn’t sufficient enough to quell my concerns. Why, yes, I did learn in econ that there is no such thing as a free lunch, however did you guess??]

So far, the privacy side has won! However, I know the writing is on the wall, so I will cave at some point. :zany_face:

Happy Thanksgiving to all of you …and AI models too!

Fin Whale Fan :whale:

9

Good morning, and thank you for your insightful “chimes”. I love learning something that I never knew, and you gave me some very interesting stuff to chew and digest. :slight_smile:

I absolutely agree with you that a patient’s independent use of an AI will likely come with some understandable privacy agita. It’s definitely a legitimate concern, to be sure. But I have the habit of conducting an informal risk-benefits-analysis (RBA) for moments like this one, and I’ve discovered that the benefits that I personally receive far outweigh the risks that may come with my utilization of AI. Your RBA may produce a different result, of course.

What would be another benefit that I enjoy receiving, for example? Glad you are mentally asking. :slight_smile:

AIs help me to formulate the best questions that I can ask of my doctor the next time that i see him or her. Case in point, my annual blood tests via LabCorp or Quest. I will submit the written results to a couple of AIs, accompanied by the following question/query/prompt:

“If you were a very well-informed patient who has an appointment with your doctor to go over the blood test results that I am attaching to this query, what would be a good series of questions that you would want to pose to the doctor regarding the results? What would be a good follow-up test or two, if any, that the doc might be interested in considering from the attached results?”

Can you see how an AI can be of valuable assistance, in this instance, in helping the patient to form relative and helpful questions to ask of his or her doc? I’ve known many family members who often admitted to me that they did not know what questions to ask in a follow-up appointment with their doc. That landscape has now changed, thanks to the advent of AI.

Thank you, again, for your thought-provoking “chimes”. And please give Captain Ahab my regards should your paths cross.

P.S. I am only consuming my first cup of Half Caff coffee thus far, so my cognitive skills might need some fine tuning.

10

@Peter I now have Friends in my brain! So I’m guessing close enough to your age as she’s always been one of my favorites, great actress as well.

You make a great point, any computer program is only as good as what the programmer enters. I’ll try your suggestion this week.

My first introduction to computers was in one of my Psych classes and they were the old punch cards if you can remember those horrible things. Drop a box of them and everything went awry. But the computer had the best ac on campus in the tallest building north of San Francisco at the time. Had its own generator for power failures too. Just don’t get stuck on the elevator or your class would end up being heard over the intercom in a hot stuffy box. Ask me how I know :joy:

In the USN my rate was involved with computers, I’ve been told the rate (job) doesn’t exist anymore. By then terminals had become computers in their own right. But we were taught that anything put on a computer never truly goes away. Always be careful of what you enter into it. Also tried to teach one of my nephews to stay off FB if he wanted to make rank. He thought I was a dinosaur and didn’t know anything spouting nonsense about Free Speech. (Not really nonsense unless you’re in a job where you really do ah e to mind your p’s and q’s) Tried to remind him he belonged to the government as soon as he did his second oath. His sister and her husband (military different branch, now teaching computer crypto stuff) also tried to tell him. He didn’t listen, he never made rank like the other nephew, stayed in longer too. Big Brother truly is watching, probably more now than when I was in.

In a way, it’s the same as what is done here, when we ask folks not to use their full name so as to try to protect identities. There are always very bright folks who can figure out things via the internet, sometimes things we may not want known to others.

Which all falls into the category @FinWhaleFan has presented, doesn’t it? How and who is collecting and keeping our data? Who can access it? We certainly aren’t signing a HIPPA form to keep our information semi private.

Thank you both for giving lots of insight not only for me, but everyone reading! Well done and since you’re both in the States, have a very Happy Thanksgiving!

1 Like
11

thanksgiving-29

4 Likes