Biopsy result - or not… venting

Firstly any form of digital diagnosis is far from ideal.

Don’t get your hopes up or feel disappointed.

I stay well away from digital diagnosis because it’s not accurate for you, it’s American and it can’t see you.

My advice is to be patient.

I realise you need to talk to someone so talk away. The forum is here to help and we are very experienced and supportive. Talk to someone you know and trust, or maybe a counsellor or partner. But stay away from your mouse.

Good luck

Hello Chairboy1d4982 

That's very shabby from your NHS Trust. The appointment was both a waste of your time and the Consultants and I would be taking my complaint to the top.

As to Chat GPT/AI I agree with Mr U - leave it well alone. Whilst it can give a general guide you can never get an accurate diagnosis as it searches the entire internet to find answers and quite frankly there's so much wrong information out there.

Have you and your team considered the fact you may be a none (or low) PSA secretor (yes we do have some on the group)  and thus the MRI/Biopsy is still considered the "gold standard" for diagnosis.

I wish you well and I do hope in the end you find you are cancer free.

Best wishes - Brian.

 Thank you for your replies. I agree to an extent with Chat GPT being unreliable but used it for a synthesis of peer reviewed articles and read several of them myself looking at characteristics of false positive PIRADS 5 MRIs. Me not having a lesion but a diffuse spread is indicative of a non cancer although there’s a chance it’s a non PSA secretor. 

interestingly the doctor said had I been older, with an MRI like that, even though coded as PIRADS 5 they may have advised against biopsy due to what they saw and the low PSA, balancing risks in biopsy against low likelihood of it  being clinically significant and suggested monitoring and repeat imaging. 

I have an interest in AI, and use them frequently. Note "them".

Quite often I will copy and paste my prompt in one or more, and compare the results.

I always ask them to list the sources, and then compare those too.

The 3 I use most at Gemini, Copilot and Perplexity.

Sometimes I take the output from one AI and copy it into one of the others, with the additional request to review and check.

In the end I usually get information that I cane use.

PS: Copilot is now universally referred to as "Colin" in my family, courtesy of staff members from the local police force.

Essentially, I am using them to find sources that I can check and learn from.

So the biopsy came back last Monday. I can see it exists on patient portal but can’t read it. It hasn’t escaped onto Airmid like my MRI did.

So I rang up the Consultant’s secretary today and I haven’t made weekly MDT meeting for today. Reading between the lines that was another error (possibly due to industrial action) to add to the ‘no results’ appointment I attended. The journey has been quite long now, referred on the two week fast track by GP 27/09/25. MRI and Ultrasound done quickly but first appt with urologist 24/10/29 so the 28 day to rule in/out fell there. Biopsy 27/10/25 and I am assured I’ll be discussed at MDT on the 24/11/25.

I’m pretty sure it’s a false positive with the MRI not being typical despite being scored PI-RADs 5, and my PSA of 1.1. The biopsy surgeon said it didn’t ‘feel’ malignant to his Digital exam. So the irritating delay isn’t too bad for me.

To keep myself “amused”, and I don’t recommend this for everyone, I have run my MRI and PSA results through Chat GPT 5, Perpexity Pro and Copilot “Pro” now. Chat GPT reckons <5% clinically significant chance, the other two 50-70% likely. It’s been interesting to then put there reasoning back to each-other, Chat GPT is the most robust in trying to convince me the other two aren’t using enough evidence. Perplexity disputes a piece of evidence Chat GPT sees as very strong. 

Interestingly Chat GPT was “pushable” to get to <5% as it started around 25% but seemed to want to give me what it thinks people want to hear. Perpexity won’t revise downwards.

So my conclusion so far is - my hospital struggle to sort out admin and AIs, whilst they can find good quality peer reviewed papers may struggle to interpret them - or possibly claim they have used them in their analysis but in reality just use easy to access websites.

All being well I’ll have an appointment offered to get results before the end of November but may well be December. So 8-9 weeks + from referral. I wonder if that’s typical? Shall I ask Copilot

Reply to end this. They must have got me into MDT today. And al clear. Prostatitis! 

Great news! 

Although I suppose it's a little odd to be glad to be celebrating the fact that we "only" have an infected organ. 

I hope that your recovery is speedy and complete.

On the subject of AI I tend to ask them to list their sources, and work through them. 

I have a mental list of the types of sources I am happy to accept. I learn from the sources, and then review the AI results. 

AI is brilliant for summarising results you have found, and presenting them in layman's terms, so sometimes I ask them to review sources they have found, and that I'm satisfied with.