panorad.io

Automated, personalized, radiology-pathology correlation

[panorad.io @panorad_io]

@AlexThomasNDUC @jonherochung You should check out PATHFINDER - we simplify and automate rad-path correlation. We agree that it is critical for radiologist education and continual improvement!

[Nikhil Krishnan @nikillinit]

Hospital that's “building a startup internally”:

[panorad.io @panorad_io]

Heading to the @RadiologyACR Quality and Safety + Informatics Conference in San Diego this week? Stop by and say hi!

We’re excited to show how our customers use PATHFINDER to drive QI, diagnostic accuracy, and research data projects.

Learn more at https://t.co/MDs57xv4GZ

[panorad.io @panorad_io]

@olgabrook @martafloryMDrad @BIDMCRad @CsirEtc We agree! We think that all radiologists should be doing routine review of rad-path cases and have made it easy with #PATHFINDER. Learn more at https://t.co/MDs57xv4GZ

[panorad.io @panorad_io]

@martafloryMDrad @FraRigi @olgabrook @Dr_3absi2 @hadiskavandi @sethb_md @twloehfelm Thanks for the shoutout! We’re super happy to be live with #PATHFINDER @StanfordRad!!

[panorad.io @panorad_io]

@heacockmd @RogueRad We have some ideas:

See here for starters: https://t.co/UlXRJRLreK

…and https://t.co/MDs57xv4GZ for more!

[panorad.io @panorad_io]

@mehran__jalali Our PATHFINDER application allows fine tuning radiology and pathology report classifiers from any Huggingface base model!

[panorad.io @panorad_io]

@PBalthazarMD @hanssenli @johntmoon @judywawira @EAKrup @ZachBercuMD @angiowoman @ImonBanerjee6 @EmoryRadiology @CI_Journal Very cool! We want in on the next one! #DYK that #PATHFINDER calculates readability metrics for every report impression automatically? Our platform-based approach to report processing streamlines work like this and enables it to be deployed at scale and in real time!

[panorad.io @panorad_io]

@jonherochung @francisdeng @MatthewDLi @kalpathy1 @MinLangMD @MarcSucciMD @ajh25_MghGrub @BJR_Radiology We should talk for real - our base use case is path feedback to radiologists. We think that alone is important. We use those annotations & processing pipelines to do automated dx accuracy assessment & peer learning. Transformer-based classifiers are another layer on all top that.

[panorad.io @panorad_io]

👇💯 Contact us for a demo!

[Jonathan Chung, MD @jonherochung]

@francisdeng @MatthewDLi @kalpathy1 @MinLangMD @MarcSucciMD @ajh25_MghGrub @BJR_Radiology Important paper. I would love for someone to create a NLP generator which you could train by yourself by loading in like a thousand clinical reports with labels. Has a lot of potential for quality where 100% accuracy is not essential, general trends are more important.

[panorad.io @panorad_io]

@jonherochung @francisdeng @MatthewDLi @kalpathy1 @MinLangMD @MarcSucciMD @ajh25_MghGrub @BJR_Radiology Our zero-code workflow = define, label, train, & deploy transformer-based report classifiers! Our customers classify PE reports as +/- to monitor CDS, RUQ US for acute chole to prep for research, and path report as malig or benign, diagnostic or inadequate for downstream QA work.

[The Radiology Review Podcast @radrevpodcast]

Unbiased Review of Panorad’s Pathfinder — The Radiology Review Insider.

Guest Post by Mike Larson, MD PhD from UC Davis

⁦@UCDRadiology⁩, ⁦@UCDRadRes⁩ https://t.co/vrMj1gwOV7

[panorad.io @panorad_io]

@radrevpodcast @UCDRadiology @UCDRadRes Thanks Drs Larson and Covington! We’re proud of the ways PATHFINDER facilitates radiologist-directed learning, QA, and research, and of the exciting work happening on our platform @UCDavisHealth.

We now have 2 transformer-based text classifiers deployed to production🤗

[panorad.io @panorad_io]

@heacockmd @Azure @OpenAI We can help you find old cases - search your entire archive of radiology and pathology reports instantly. We'll even let you know when a case you read two years ago finally received path proof - no searching necessary.

[panorad.io @panorad_io]

@searsvillemayor @bodyradgal We can help with that!

[panorad.io @panorad_io]

PATHFINDER - follow-up is good even when the diagnosis is never in doubt.

Strengthen the good synapses and prune the bad ones!

Gleason 4+5 prostate cancer on CECT, PSMA PET, and MRI

@SAR_ProstateDFP #RadPath

[panorad.io @panorad_io]

@nellytanmd We could help you monitor whether you’re actually finding ccRCC at the expected rates when you assign a ccLS!

[panorad.io @panorad_io]

@scottlee_md @AmyMehollinRay Gladly! Tell @hanssenli to hit me up!!

[panorad.io @panorad_io]

@AmyMehollinRay @scottlee_md We couldn’t agree more - our PATHFINDER software automates this process for you so you can easily see path outcomes for all your radiology reads

[panorad.io @panorad_io]

PATHFINDER - because you should learn your blindspots

Here's an RCC met to the L4 lamina on arterial-phase
CECT that was missed on the first opportunity to identify it.