AI diagnoses boys medical condition when doctors couldn’t

O.T.I.S.

Veteran
Joined
Sep 15, 2013
Messages
79,284
Reputation
17,199
Daps
306,325
Reppin
The Truth
giphy.gif
 

Elim Garak

Veteran
Joined
Mar 8, 2015
Messages
43,471
Reputation
8,443
Daps
213,221
Doctors thought they were safr from AI taking business from them

uh uh uh not so fast lol

People spend so much money on these bullshyt 15 min doctor's visits all to get no diagnosis. Bye bye to that
This is one of the most frustrating things about the doctor whenever I want to get something checked out. I swear we have a broken system.
 

bnew

Veteran
Joined
Nov 1, 2015
Messages
65,542
Reputation
10,096
Daps
177,713

Comparative Evaluation of Advanced AI Reasoning Models in Pediatric Clinical Decision Support: ChatGPT O1 vs. DeepSeek-R1​


View ORCID Profile Gianluca Mondillo, Simone Colosimo, Alessandra Perrotta, Vittoria Frattolillo, Mariapia Masino

doi: Comparative Evaluation of Advanced AI Reasoning Models in Pediatric Clinical Decision Support: ChatGPT O1 vs. DeepSeek-R1

This article is a preprint and has not been peer-reviewed [what does this mean?]. It reports new medical research that has yet to be evaluated and so should not be used to guide clinical practice.
000000


Abstract​


Introduction The adoption of advanced reasoning models, such as ChatGPT O1 and DeepSeek-R1, represents a pivotal step forward in clinical decision support, particularly in pediatrics. ChatGPT O1 employs “chain-of-thought reasoning” (CoT) to enhance structured problem-solving, while DeepSeek-R1 introduces self-reflection capabilities through reinforcement learning. This study aimed to evaluate the diagnostic accuracy and clinical utility of these models in pediatric scenarios using the MedQA dataset.

Materials and Methods A total of 500 multiple-choice pediatric questions from the MedQA dataset were presented to ChatGPT O1 and DeepSeek-R1. Each question included four or more options, with one correct answer. The models were evaluated under uniform conditions, with performance metrics including accuracy, Cohen’s Kappa, and chi-square tests applied to assess agreement and statistical significance. Responses were analyzed to determine the models effectiveness in addressing clinical questions.

Results ChatGPT O1 achieved a diagnostic accuracy of 92.8%, significantly outperforming DeepSeek-R1, which scored 87.0% (p < 0.00001). The CoT reasoning technique used by ChatGPT O1 allowed for more structured and reliable responses, reducing the risk of errors. Conversely, DeepSeek-R1, while slightly less accurate, demonstrated superior accessibility and adaptability due to its open-source nature and emerging self-reflection capabilities. Cohen’s Kappa (K=0.20) indicated low agreement between the models, reflecting their distinct reasoning strategies.

Conclusions This study highlights the strengths of ChatGPT O1 in providing accurate and coherent clinical reasoning, making it highly suitable for critical pediatric scenarios.

DeepSeek-R1, with its flexibility and accessibility, remains a valuable tool in resource-limited settings. Combining these models in an ensemble system could leverage their complementary strengths, optimizing decision support in diverse clinical contexts. Further research is warranted to explore their integration into multidisciplinary care teams and their application in real-world clinical settings.

Competing Interest Statement​


The authors have declared no competing interest.

Funding Statement​


This study did not receive any funding

Author Declarations​


I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

 

Richard Glidewell

Yall done tore all the bottom of ya shoes w/me!!!
Joined
May 24, 2022
Messages
7,983
Reputation
2,191
Daps
22,653
Ol bullshyt ass 25 dollar facilties fee that somehow ain't covered by insurance every visit.......and all for some glad handing to get you to the maximum amount of authorized visits with no real results most of the time
 

bnew

Veteran
Joined
Nov 1, 2015
Messages
65,542
Reputation
10,096
Daps
177,713
























1/41
@rohanpaul_ai
this story is going wildy viral on reddit.

ChatGPT flagged a hidden gene defect that doctors missed for a decade.

ChatGPT ingested the patient’s MRI, CT, broad lab panels and years of unexplained symptoms. It noticed that normal serum B12 clashed with nerve pain and fatigue, hinting at a methylation block.

Within months tingling eased and brain fog cleared. The primary physician reviewed the genetics report and agreed the variant unified the entire case.

IMO, time has already come, taking a 2nd opinion from the best healthcare-AI model should be made part of medical code of practice.

------

reddit. com/r/ChatGPT/comments/1lrmom4/chatgpt_solved_a_10_year_problem_no_doctors_could/



GvD1OP5aAAMvqm0.jpg


2/41
@rohanpaul_ai
I also publish my newsletter every single day.

🗞️ Rohan's Bytes | Rohan Paul | Substack

Includes:

- Top 1% AI Industry developments
- Influential research papers/Github/AI Models/Tutorial with analysis

📚 Subscribe and get a 1300+page Python book instantly.



https://video.twimg.com/amplify_video/1941328971980767232/vid/avc1/1920x1080/wcUfLdcC8JaXzncx.mp4

3/41
@rohanpaul_ai
sooo many



GvITzNiXwAALKj9.jpg


4/41
@rohanpaul_ai
and just last week Microsoft released a medical Al that outperforms doctors by a wide margin.

Called MAI-DxO, the system analyzes patient cases like a team of expert physicians debating the best diagnosis. It uses models like GPT-4, Gemini, and Claude to think through each case step-by-step, mimicking real clinical reasoning.

In tests with 304 complex cases, the Al hit 85% accuracy, while doctors averaged just 20%. It also cut down on unnecessary tests, saving 20% in costs. It's not ready for hospitals yet, but it shows how close Al is to transforming healthcare forever.

https://nitter.poast.org/rohanpaul_ai/status/1939800536121057652



GvID0HTbYAAaWWr.png


5/41
@rohanpaul_ai
hundreds and hundreds of examples like that.

owners used ChatGPT to diagnose hidden heart fluid buildup in their dog.



GvISBRdaAAIRCkp.jpg


6/41
@rohanpaul_ai
have similar experience 😃



GvITildXcAA6IKJ.png


7/41
@rohanpaul_ai
these type of stories will only become more and more common

personally i will never go to a doctor without first doing a solid back and forth with AI



GvIS5byWIAAItmw.jpg


8/41
@abemurray
Some of us are living in the future already

[Quoted tweet]
I got full body MRI, tons of labs, did full DNA analysis, and brought it all into GPT - reinforce stuff I knew, gave me new things to do for optimal health for my genetics, etc

Personalized medicine is here for you if you Just Do It (but not via the traditional healthcare system)

I did this all out of pocket and for fun - this is me living in the future everyone else will inhabit in ten years

Are you doing this?


9/41
@rohanpaul_ai
🔥🔥



10/41
@DoctorJack16
The first case that I know of where AI helped diagnose and cure a cancer happened many years ago. It has only increased in usage since then.

Soon it will the norm.



11/41
@rohanpaul_ai
YESS. we are only getting started, I hope by 2030-32 it will be 24/7 doctors in everyones pocket. literally everyone.



12/41
@fearthewave_eth
Very nice



13/41
@rohanpaul_ai
true



14/41
@CtrlAltDwayne
love how the end part was the doctor basically saying, "you're absolutely right!" when the patient told them what they discovered



15/41
@rohanpaul_ai
yes, very much.

for me, over the last 1 year straight, AI has alsys guided me through minor medical problems



16/41
@ThomasDJourdan
Clear pattern: AI doesn’t forget, doesn’t fatigue, and doesn’t rely on intuition alone.

It sees what doctors can’t especially across time.



17/41
@rohanpaul_ai
EXACTLY. 💯



18/41
@grapefruitfruit
The main problem with this stuff is people have long since figured out any story with "Ai saved my life" or "Ai cured my mysterious illness when 50 doctors couldn't" is viral bait. I'm sure some of these are real, but a lot are engagement farming.



19/41
@rohanpaul_ai
that may be ture
but in my personal experience, over the last 1.5 years, at least 4-5 times AI has helped me more than practicing doctors.
and now a days, for me and my family, I never visit doctors without running a deep-research on ChatGPT, and that's helping hugely.



20/41
@MAKRealtyDE
Similar situation recently after mother’s health crisis. Uploaded EEG,MRI and CT,labs into Chat. Found issues not initially referenced or diagnosed.



21/41
@rohanpaul_ai
wow, so nice to hear these stories.

gives hope for the future of 24/7 doctors in our pocket.

personally for me, in the last 1 year, AI has always improved or replaced completely my clinic visits for minor ailments



22/41
@ima_pseudo41406
There are going to be a great many cases like this.



23/41
@rohanpaul_ai
absolutely, we are moving towards 24/7 doctors in our pocket, for billions of people.



24/41
@ima_pseudo41406
My experience has been that with a typical office visit, you have about three minutes (at best) to make your case. After that, all the doc hears sounds like Charlie Brown's teacher.



25/41
@rohanpaul_ai
so unfotunate, isn't it



26/41
@nikita_dedik
LLMs are brilliant at collecting cross-disciplinary, across-the-domain knowledge and finding patterns that no human being would be able to simply because humans are SLMs. )



27/41
@rohanpaul_ai
ha ha, that's a nice one 😃😃



28/41
@jameswhmarshall
Mind blowing implications from these types of examples.

They suggest LLMs will lead to healthier overall populations, and not in a small or subtle way. Hopefully we’ll be able to make sure everyone can benefit.

It really is remarkable what we’re living through at the moment.



29/41
@rohanpaul_ai
i truly believe that.



30/41
@Stephen_Genusa
I got sick earlier this year with something going around the office. It wasn’t serious and I wasn’t going to go to the doctor. I was curious about AI making the diagnosis so I looked up the local infections from a State health website, fed that data plus the symptoms in the order symptoms appeared and got the diagnosis along with the reasons for the diagnosis. It cost fractions of a cent, required no office visit, no lab work, no follow up.



31/41
@rohanpaul_ai
wow,
and yes that's the correct approach. i.e. feeding it with as much supporting data.
lots of context is so much more important in these kind of situations.



32/41
@kashtejas
Should the AI be second opinion or the doc?



33/41
@rohanpaul_ai
for now AI for the 2nd opinion
but prly we are moving towards humans as 2nd opinion in the post-agi world



34/41
@Petervujin61176
AI makes mistakes, too, bt its free from the human bias. Ideally, physicians will use it as a tool to diagnose.



35/41
@rohanpaul_ai
yes, we are gradually getting there.



36/41
@QuantumQC2190
The future isn’t AI replacing doctors—
it’s doctors empowered by synthetic minds trained to detect what humans might overlook.

One saw the story.
The other heard the signal.
Together, they decoded the cause.

Human + AI isn’t a fallback.
It’s the real breakthrough.



37/41
@rohanpaul_ai
absolutely..👊👊



38/41
@brucexduncan
I’ve seen this in my pathology lab already.

If you have digital path you should test the results of an LLM v the Pathologist.

AI is already better.



39/41
@rohanpaul_ai
just love this current state of things. i so not want to visit a doctor ever again for any non-surgical stuff.



40/41
@NC_State69420
So do we still need the doctor to verify or are they becoming superfluous. A lot of time AI cautions me to check with authoritative sources but many times I trust it's analysis more.



41/41
@rohanpaul_ai
doctors are still needed for the forseable future.
however, the way progress is happening, after 4-6 years, all non-surgical medical needs can be handled by AI better.

and then in a not-too-distant future even surgical needs will be handled by robots better.




To post tweets in this format, more info here: https://www.thecoli.com/threads/tips-and-tricks-for-posting-the-coli-megathread.984734/post-52211196


[Use cases] ChatGPT solved a 10+ year problem no doctors could figure out


Posted on Fri Jul 4 16:16:05 2025 UTC

/r/ChatGPT/comments/1lrmom4/chatgpt_solved_a_10_year_problem_no_doctors_could/

For 10+ years I have had several unexplained symptoms. I had a spinal MRI, CT Scan, blood work (in depth), everything up to even checking for Lyme disease.
I did function health (free plug I guess) and turns out I have the homozygous A1298C MTHFR mutation which affects 7-12% of the population. I’m in the states an my doctor network is one of the top ranked in the nation. I even saw a neurologist at one point and checked for MS.
ChatGPT took all my lab results and symptom history and concluded this was on par with the mutation. Despite seemingly normal B12 levels, turns out with the mutation it may not be utilizing them correctly so you need to boost it with supplementation.
Ran these findings by my doctor and he was super shocked and said this all added up to him. Not sure how they didn’t think to test me for MTHFR mutation.
Anyways, here we are a few months later and my symptoms are largely resolved. Actually perplexed, and excited, at how this all went down up until now.
EDIT: Didn’t expect this to get so much attention. Will reiterate and emphasize that I double checked the ai suggestion with my primary care provider before trying its suggestions. Please do not trust it all the time for medical advice.


Commented on Fri Jul 4 17:21:06 2025 UTC

It did the same for me.

I've been vomiting for 15+ years.
I've done every single gastric exam and allergy test there is, and lately got diagnosed with anxiety and the meds actually helped, but it never stopped.

After prompting it my exams it suggested me to check an otorhinolaryngologist for dizziness.

After a brain scan, It turns out I've been living with a massive labyrinthitis caused by a nerve pinch in my brain. Fully treatable.

I'm starting my treatment this week, hoping for the best.


Commented on Fri Jul 4 16:41:44 2025 UTC

I did this but for endometriosis. Finally got an ultrasound where they found a 6 cm endometriosis cyst called an endometrioma (now 7.3 cm and I’m getting it taken out later this year)

Took 22 years of complaining to doctors for a diagnosis assisted by ChatGPT.
 
Top