#41
1 Like
#42

Would people here use this new service (currently initiated in Sacramento, CA):

1 Like
#43

GTRilNOaYAobVmL
GTRilNOaYAobVmL692×692 83.2 KB

#44

ChatGPT has been an unbelievable help to me. In the course of 6 months, I’ve been able to help my mom and her doctors improve 3 separate chronic diseases / conditions she has had for years. As a smart and interested non-medical professional, I was able to dig into her symptoms and current treatments and find alternatives that might work better and then had educated conversations with her doctors that went very well. I didn’t show up saying “let’s do this because the computer says so” but I used the information that I found as jumping off point for the conversation. The doctors largely appreciated the discussion and questions and in the end a lot of improvements to her treatment came from it.

If this is how good it is now in 2024, I can’t wait to see the 2030 version.

6 Likes
#45

University of Oxford is trialing ARTICULATE PRO, an AI technology for detecting and grading prostate cancer… but this is just the beginning.

1 Like
#46

and if you want to predict your glucose try January AI

3 Likes
#47

Friends of mine built this: https://vera-health.ai/

For doctors only. I understand that you can sign up and try it, so doctors here might be interested (@DrFraser @Dr.Bart and others?).

4 Likes
#48

Just remember that whatever the glycemic index of a food, that individuals respond differently. So with a CGM (January Pro) this could be good.

https://pubmed.ncbi.nlm.nih.gov/20681028/

“Glycemic Index (GI) values are, therefore, relative and are not necessarily a reliable guide to the person’s actual individual AUC when consuming a food. Without knowledge of the person’s characteristic blood glucose responses, reliance only on the GI may be misleading.”

#49

I was part of their beta with CGM a long time ago. They were way behind Levels and their initial idea of training an AI on people with CGMs is not going to work in the end. There is way too much variability between people and between meals for glucose response prediction even if you enter exactly what you eat.
The order in which you eat/drink, the speed of eating, if/when you did a workout before/after, etc. makes it impossible to make any meaningful glucose prediction.

I switched to Levels. There is no substitute for a real CGM + calibration with fingerpick glucose.

#50

I think it has a lot of potential - CarePods. In accordance with our mantra…test, test, test, anything that makes testing easier and cheaper is a good thing.
" And, making scans affordable for the masses, Forward’s competitive price point could encourage adoption beyond biohackers."
As long as they keep improving it with new tech and new tests.
“Aoun has ambitious plans for CarePods, and the company’s proven ability to iterate and develop new solutions shows it’s likely agile enough to work through early adopter issues. CarePods has a surprising number of features, and Forward Health will keep creating more.”

Here’s how it works: Users walk up to the door and unlock it with their mobile device. They step in alone and find a large touchscreen facing a chair. A glowing ring on the floor indicates the location of the full-body scanner. A hidden drawer on one of the sides delivers different medical tests as needed. A friendly female voice guides the patient through the process.

The CarePod’s screen serves up different apps available to the user: full body scan, heart health, thyroid testing, blood pressure, weight management, diabetes screening, COVID-19 test, HIV screening, kidney, and liver health — there seems to be an app for many ailments. Let’s say you pick heart health. The drawer opens and presents the patient with a sensor. The touchscreen instructs the user to place the sensor against their heart, and after a few tests, the diagnosis is displayed on the screen. And if more treatment is required, one of Forward’s doctors will review the findings in real time and issue a prescription or additional instructions.

Blood tests are performed with a single-use, needleless collection device. Blood pressure is measured using a wireless arm cuff. Skin testing uses a scanner. Various swabs are available to test for COVID-19, strep and other ailments. The company says it also tests for STDs but does not detail which type or how it completes the diagnosis.

Each CarePod is staffed with an attendant who can answer basic questions and services the device between uses.

After they get the kinks worked out, the attendant could probably be replaced on-screen with an AI chatbot and avatar. Should technically be able to do some simplified form of a DexaScan or VO2 Max test.
A CT scanner is too big and expensive to allow CAC tests but I wonder if they could invent some kind of automated, robotic machine to do CIMT tests. That would be great.
They could even have specialized cameras to do FaceAge or retina scanners to do RetinaAge and certainly could do eye tests and hearing tests. Or some functional tests like grip strength, balance or any number of others. Could be an easy stop in the mall to get your BioAge engraved on a little plaque that you could hang in your kitchen (certainly better than that damn plaque in your arteries!).

#51

Well your friend needs to give me an invite code … Would love to sign up - but I’m not a member of the exclusive club with a code. I’d love to get one, and happy to give feedback. The use of AI searches is helping me in my work - it’s not the end all, but is doing things faster and better than the old way.

2 Likes
#52

You can join the wait list here and they’ll give you access: Vera Health

#53

Okay — I’m wait listed … so sad, never been wait listed before. Time to accept my position and identify with being a member of the proletariat.

3 Likes
#54

A new AI tool based on research papers:

From: I'm troubled by something. Longevity Quotient and mTOR per cell, Please poke holes - #14 by RapAdmin

#55

Elicit also looks good.

https://doctoraimd.com/elicit-ai-vs-consensus-ai-literature-reviews/

And here are some more that medical students use:

https://forums.studentdoctor.net/threads/gpt-4-and-medical-research-apps.1485558/

(https://drgupta.ai/)

(https://www.droracle.ai/)

(https://www.openevidence.com/)

And there’s this:

https://www.euronews.com/next/2024/01/20/best-ai-tools-academic-research-chatgpt-consensus-chatpdf-elicit-research-rabbit-scite

Which adds:
(https://scite.ai/)

(https://www.researchrabbit.ai/)

I think @desertshores is happy with Dr Oracle but if anyone has personal experience with the others and comparisons between them, let us know.

1 Like
#56

I just tried it with the following tag.

metanalyses on berberine, last 5 years. Exclude papers form china

Alas, it included some papers from China, I’m wondering if there si a way around this, maybe just skip those results?

#57

“form china” or “from china” - perhaps retry and see what happens, or try other boolean operators (“not china”, “exclude china”)…

#58

The papers are not “from China” per se and I’m not sure the model takes into account the affiliations of the authors. You could try “Only include the highest quality evidence published in the best journals by the best researchers”.

1 Like
#59

I retried but the AI seems to ignore the prompts to exclude china, not china, Chinese authors and so on. Maybe it has been trained to ignore potential racial suggestions!

I also pasted the longest sentence Only include the highest quality evidence published in the best journals by the best researchers. I don’t know how much the AI considered that, the caption for the results only included the 1st part of the enquiry ( Meta-analyses on Berberine from the Last 5 Years)

#60

I assume that the AI has no idea what are good quality journals. The developers should take into account impact factor + number of citations of each paper.