Ankit Sakhuja

Schedule

The October academic activities in nephrology at @uottawa.ca @ottawahospital.on.ca

Feat @nephcrit.bsky.social @dialysisbloke.bsky.social and @asakhuja.bsky.social

#NephGR #NephSky

Off to a great start of #POCUS immersion course at #SWNC with @nephrothaniel.bsky.social ‘s opening talk

NephJC Editorial Internship — NephJC

Announcing the #NephJC internship, @nephjc.bsky.social and @kireports.bsky.social joining forces

www.nephjc.com/intern-app for details and to apply!

#NephSky

Proud to be part of this pivotal study on persistent severe #AKI in critically ill patients, alongside leading experts in the field. Excited to contribute to research that pushes the boundaries of our understanding of this critical condition!

pubmed.ncbi.nlm.nih.gov/39966171/

Love these maps. Puts it into a perspective

Honored and humbled to have been a part of the faculty for the #SCCM #AI PreCourse at #SCCM2025. Grateful for the opportunity to discuss critical role of #AI in resource-constrained settings. Exciting times ahead for innovation in critical care!

⁦‪Mount Sinai #ICCM‬⁩ alumni dinner at #SCCM2025. Privileged to work with such amazing colleagues and so proud of the spectacular achievements of our fellows- both past and present

Still time left to join our 2-day #pocus program @ SW Nephrology conference in Phoenix 3/20-3/21 @NKFArizona!

swnephrologyconference.org
$125 discount for trainees!

Guaranteed surfeit of hemodynamic assessment!

Featuring:
@argaiz.bsky.social
@asakhuja.bsky.social
@michaelturk6.bsky.social

Do they identify reasons for increased morality? Delayed labs, orders, chart review etc? #NephJC

Shows the importance of built in redundancies in IT systems (including paper based downtime procedures) #NephJC

That’s like the old times. We do that occasionally during EHR downtimes. Never fun #NephJC

Ankit Sakhuja, intensivist/nephrologist from NYC #nephjc

We have had patients on ECMO

Can carefully developed electronic alerts help with management of #AKI? Yes, but developing an effective alert is both an art and science. Read about our experience in #ACI www.thieme-connect.com/products/ejo...
@girishnadkarni.bsky.social #informatics

Had so much fun discussing how #AI is shaping medicine and sharing the work from our lab at #apexpathshala @akforg.bsky.social @girishnadkarni.bsky.social

Pleasure to be a part of the phenomenal conference #apexpathshala organized by @akforg.bsky.social

Prompting strategies
Zero shot perform a task without examples
Adding examples increases the accuracy

Encourage the model think aloud or follow a logical sequence

Add the prompt “lets think step by step”

Provide context
Who are you talking to

On to transformer models. Much more complex.

Need inputs not only from right now, but from previous time. Short term memory. Recurrent neural network.
What about long-term memory: long Short-Term Memory Network.

Is the issue that there is less demand for nephrologists (and hence we should decrease fellowship spots) or is there a decreased demand for nephrology?

Extracting International Classification of Diseases Codes from Clinical Documentation using Large Language Models - PubMed Current LLMs have poor performance in extraction of ICD-10-CM codes from inpatient notes when compared against the human coder.

Are #LLMs as good as human coders in extracting ICD codes? Of 6 LLMs we tested, GPT-4 achieved highest agreement with human coder but still only at 15.2%. Read our full study in #ACI at - pubmed.ncbi.nlm.nih.gov/39608761/ @girishnadkarni.bsky.social

#coffeewiththeexperts

Get ready to know more about #AI in nephrology. Are you up-to-date with the new-age tech that keeps changing every quarter moon?

Hear Dr Sakhuja
at #ApexPathshala

🤖Have we been ‘Naturally Stupid’ so far?
⏰7.30-10am
🗓️Dec 8, ‘24
Agree or disagree? @isn-india.bsky.social

Our first #NephSky post - introducing our Fall 2024 Newsletter

Check out what's been happening in our Division 👇

issuu.com/mountsinaido...

#SinaiStrong #Nephrology

A Primer on Reinforcement Learning in Medicine for Clinicians npj Digital Medicine - A Primer on Reinforcement Learning in Medicine for Clinicians

Excited to share our review article, which provides a step-by-step comprehensive overview of reinforcement learning tailored specifically for clinicians, presented in a clear and non-technical manner - www.nature.com/articles/s41... @girishnadkarni.bsky.social #AI #RL

Same

This is great work @chriscarrollmd.bsky.social As the world in general came to a standstill then, curious if you might have access to air pollution data during that time and it’s effect on decrease in asthma exacerbations then?

I agree that rates of ODS are low in general (which is good), but if the pathophysiology ties it to rate of correction of Na, then it should be a #neverevent

That’s the primary issue with retrospective data. But it is likely what happened as clinicians are usually concerned more about patients that are sicker.

This is a great figure Joel. So clinicians were much more concerned for (and hence corrected the serum Na level much slower) for patients that were likely much sicker (and hence had higher mortality)