Small, open-source #LLMs offer an accessible solution for the curation of local research databases, said @ucsfci2.bsky.social's Dr. Maggie Chung in @nature.com. "They achieve human-level accuracy while using only desktop-grade hardware & 100 training reports."

A community of collaborators & innovators, the Chen Institute Scholars Program brings together those in the scientific community excelling in clinical study & #ArtificialIntelligence. Dr. Andreas Rauschecker speaks on the power of #AI ⬇️ www.youtube.com/watch?v=H5EW...

#DiagnosticRadiology & #ImagingInformatics combine to guide the development & deployment of #AI applications in healthcare, as described by Laurens Topff of @nkinl.bsky.social at the last @ucsfci2.bsky.social Pillar Meeting.

Can #LLMs extract structured data from notes in clinical research? Visit Dr. Maggie Chung's latest study in @nature.com to find out. ➡️

New @ucsfci2.bsky.social blog ➡️ "Finetuned LLMs Create Research Databases."

Dr. Andreas Rauschecker & collaborators compared their three-network framework to synthesis infant brain MRI to prior approaches that used generative adversarial network, a #GenerativeAI architecture where two neural networks compete.

During the December SRG Pillar Meeting, Dr. Laurens Topff of @nkinl.bsky.social joined @ucsfci2.bsky.social to discuss the building of deployable #BrainMetastasis with #AI. Visit bit.ly/4aekyzI to read the recap!

The Chen Institute empowers the pursuit of scientific discovery & those dedicated to making a profound impact in their chosen field of study. @ucsfci2.bsky.social's Dr. Andreas Rauschecker speaks on the power of #AI. www.youtube.com/watch?v=H5EW...

New research in @nature.com by @ucsfci2.bsky.social's Dr. Maggie Chung hypothesizes that, with minimal computational & annotation resources, open-source LLMs could create high-quality research databases.

@ucsfci2.bsky.social's Dr. Andreas Rauschecker & the University of Washington Radiology's Dr. Gunvant Chaudhari collaborated in @rsnasky.bsky.social to research #AI synthesizing of infant MRIs.

Join us for a talk with Dr. Laurens Topff, #Radiologist & Medical Information Officer for AI at the @nkinl.bsky.social. His work focuses on turning imaging #AI innovations into real clinical impact across European healthcare.

Our time at #RSNA25 concludes with Dr. Bonnie Joe's discussion on the assignment of risk-based on traditional models versus age-based & #AI for #BreastCancer screening. @rsnasky.bsky.social

Happening this morning at @rsnasky.bsky.social ➡️ Dr. Maggie Chung presents on screening mammography workflows & #AI applications in #BreastImaging. #RSNA25

@ucsfci2.bsky.social's Dr. Maggie Chung is at #RSNA25. She joins the 1:00 p.m. Science Session as a moderator. See you there! @rsnasky.bsky.social

Dr. Jae Ho Sohn kicks off #RSNA25 moderating the basics of natural language processing in radiology. Join us in the Radiology Learning Lab! @rsnasky.bsky.social

Each November, we highlight clinical advancements in #LungCancer research that utilize #AI. Visit bit.ly/3XCqLxC to read @ucsfci2.bsky.social's Dr. Jae Ho Sohn's studies published in BMC Medical Imaging, @springernature.com & @rsnasky.bsky.social Radiology. #LungCancerAwarenessMonth

"Investing in AI literacy enables wider implementation by empowering radiologists," write @ucsfci2.bsky.social's Drs. Andreas Rauschecker, John Mongan & team in part I of their series of research.

At the latest @ucsfci2.bsky.social SRG Pillar Meeting, @ucsantacruz.bsky.social's Yuyin Zhou presented "The Thinking Eye: AI That Sees, Reads, & Reasons in Medicine." Visit bit.ly/3XDLVLK to read more!

In @bmchealthsystem.bsky.social Medical Imaging, Dr. Jae Ho Sohn & colleagues developed a nomogram that predicts pneumothorax requiring chest tube placement following CT-guided lung biopsy. #LungCancerAwarenessMonth

"Primer Helps #Radiologists Navigate #AI" is available now on our blog! Check it out ➡️

#LungCancer diagnosis is delayed in interstitial lung disease patients on average by three years. Research by @ucsfci2.bsky.social's Dr. Jae Ho Sohn highlights the importance of early detection. Read it here for #LungCancerAwarenessMonth ➡️

"Accurate knowledge is actually far more important than the reasoning capabilities," said @ucsantacruz.bsky.social's Yuyin Zhou on the development of #AI reasoning models. Read the recap of her SRG Pillar Meeting at @ucsfci2.bsky.social.

New in @rsnasky.bsky.social from @ucsfci2.bsky.social's Dr. Jae Ho Sohn & collaborators ➡️ "Balancing Diagnostic Certainty & Local-Regional Recurrence Risk in Stage I Non–Small Cell Lung Cancer." #LungCancerAwarenessMonth

It's #InternationalDayOfRadiology! Today & everyday, we promote the essential role that #radiology plays in the patient journey. @ucsfci2.bsky.social's Dr. Andreas Rauschecker (@drdremdphd.bsky.social) relates the implementation of #ArtificialIntelligence to the advancement in #MedicalImaging.

Dr. Jae Ho Sohn of @ucsfci2.bsky.social concludes half of the #LungCancer cases in interstitial lung disease patients were diagnosed with significant delay, due primarily to parenchymal masking & subtle imaging findings. #LungCancerAwarenessMonth

@ucsantacruz.bsky.social's Dr. Yuyin Zhou joins @ucsfci2.bsky.social to discuss the transformation of #AI in medicine. Details ➡️

Dr. Jae Ho Sohn & collaborators conclude an open-source, fine-tuned #LLM can generate impressions to a satisfactory level of clinical & grammatical accuracy & stylistic quality. Read the full study in @bmchealthsystem.bsky.social ➡️

"RadSearch successfully retrieved reports containing the specified findings for 83.0% of reports & matching location for 89.8% of reports," writes Dr. Jae Ho Sohn & collaborators in a new study in @rsnasky.bsky.social.

According to Mass General Brigham's Dr. Katherine Andriole, clinical deployment of #AI requires "the data scientists' expertise, but also clinical domain expertise & implementation of informatics."

Drs. Leo Sugrue & Andreas Rauschecker participated in a @jamanetworkopen.com study on the link between social media use & cognitive performance in adolescents. Check it out:

Happening at 8:00 a.m. at #SIIM25 ➡️ a scientific abstract presentation from @ucsfci2.bsky.social researchers on the rise of LLMs & #AgenticAI in #radiology. See you there!

Join us this afternoon at #SIIM25 for a @ucsfci2.bsky.social scientific abstract presentation on #AI for clinical prediction & workflow enhancement.

@ucsfci2.bsky.social's Pierre Nedelec & collaborators review real-world performance, bias & #AI governance in a scientific abstract presentation this morning at #SIIM25. Join us!

Dr. Katherine Andriole of Mass General Brigham joined @ucsfci2.bsky.social to discuss examples of clinical deployment of #AI.

It's #NationalMammographyDay! @ucsfci3.bsky.social & @ucsfimaging.bsky.social's Dr. Maggie Chung studied the intersection of #AI & mammography screenings. Read more:

New in @bmchealthsystem.bsky.social from @ucsfci2.bsky.social researchers — "An open-source fine-tuned large language model for radiological impression generation: a multi-reader performance study."

In a retrospective study, @ucsfci2.bsky.social's Dr. Jae Ho Sohn developed a scalable method for training semantic search models for radiology reports. ⬇️

The next five to 10 years will see more opportunities for personalized screening & intervention for #BreastCancer. A recent trial by @ucsfci2.bsky.social & @ucsfimaging.bsky.social's Dr. Maggie Chung is contributing to expediated patient care.

Dr. Katherine Andriole of Brigham & Women's Hospital joins @ucsfci2.bsky.social for the SRG Pillar Meeting next week! Her talk will focus on #ArtificialIntelligence in #radiology.

"Now is such an exciting time with #AI, because with #BreastImaging, a lot of the data is very well curated," said @ucsfci2.bsky.social's Dr. Maggie Chung. Visit bit.ly/3KASj3c to read about her research & predictions for cancer care & intervention. #BreastCancerAwarenessMonth

Interesting SRG Pillar Meeting, "#AI in Radiology: Experiences from Model Development to Clinical Implementation." 👩 Dr. Katherine Andriole 🗓 Monday, October 13 🕒 3:00 - 4:00 p.m. Join us on Zoom or in person! ucsf.zoom.us/j/9127716522...

On October 14, @ucsfci2.bsky.social's Dr. Sharmila Majumdar will join several panelists from @yale.edu to discuss the intersection of healthcare, research & #AI. Join us in person or on Zoom!

The Early Career Global Health Scientist Award has been awarded to @ucsfohns.bsky.social's Dr. Lia Jacobson! Visit bit.ly/3W7kQ2W to learn about her nomination & achievements.

Be sure to check out some of our @ucsfohns.bsky.social poster presentations at The Fall Voice Conference this weekend!

What is multi-omic data? It's a specific type of multimodal data focused on biological 'omics' levels, such as genomics & radiomics. "Integrating these data streams unlock deeper insights," Dr. John Mayfield said. Visit bit.ly/3IAW1t0 for more of his research findings.

Tomorrow afternoon, Dr. Peder Larson (@pezlarson.bsky.social) will host several @ucsfci2.bsky.social members for a 20-30 minute discussion on #AI tools & methods. See you then!

@ucsfci2.bsky.social postdoc Atlas Haddadi Avval presented her research, "Automated MRI-based Classification of Tumor Progression in Diffuse Midline Glioma Using Neural Networks," this past weekend at #ASFNR2025! Congratulations on your hard work & successful presentation! 👏

Dr. John Mayfield of Massachusetts General Hospital joined @ucsfci2.bsky.social for our September SRG Pillar meeting. Our latest blog highlights his research & the clinical applications of classical & quantum #AI methods.

@ucsfci2.bsky.social #postdoc Atlas Haddadi uses #AI & imaging to study diffuse midline glioma. She hopes this work will help doctors detect progression earlier & improve survival for young patients. Happy #PostdocAppreciationWeek!

Happy #PostdocAppreciationWeek! Rabab Rangwala uses #AI to study how speech changes with healthy aging & neurological disease. Her research blends clinical care & neuroimaging to improve understanding & patient support.