Voio Emerges From Stealth to Build Frontier AI for Healthcare

Voio staff’s preliminary open-source launch, Pillar-0, is the world’s most correct AI mannequin for radiology and builds on the success of the staff’s different AI instruments already utilized in 92+ hospitals and a couple of million mammograms worldwide

Voio, a frontier AI lab devoted to healthcare, right now emerged from stealth with $8.6 million in seed funding from Laude Ventures and The House Fund. The firm’s first focus is constructing a unified studying platform that helps radiologists throughout each scan and modality, spinning out from analysis labs on the University of California, Berkeley and University of California, San Francisco (UCSF). Voio’s founding staff right now launched Pillar-0, an open-source AI mannequin that interprets medical pictures immediately to acknowledge a whole lot of circumstances from CT and MRI scans with unprecedented accuracy. The preliminary iteration of Pillar is the world’s most correct AI mannequin for medical imaging with a demonstrated 10% – 17% accuracy enchancment over main fashions from Google, Microsoft, and Alibaba.

Voio was based by Adam Yala, Assistant Professor of Computational Precision Health at UC Berkeley and UCSF, Dr. Maggie Chung, Assistant Professor in Radiology and Biomedical Imaging at UCSF and a practising radiologist, and Trevor Darrell, Professor of Computer Science at UC Berkeley and founding father of Berkeley AI Research (BAIR). The staff’s earlier AI fashions have been validated in 92+ hospitals throughout 30 nations worldwide and their breast most cancers software has been utilized in over 2 million mammograms world wide.

Radiologists can drive higher affected person outcomes throughout the complete healthcare system: quicker and extra correct diagnoses, proactive scientific administration based mostly on refined imaging cues, and higher affected person entry to the most effective of care. Voio goals to empower radiologists with state-of-the-art know-how to push the scientific frontier, creating a better customary of care and addressing the crucial workforce scarcity in radiology. With 375 million CT scans carried out yearly, this capability hole is main to longer turnaround occasions, elevated burnout, and diagnostic delays that negatively impression affected person care.

Currently, radiology reporting requires fixed context-switching between the picture viewer, reporting software program, EHR, and AI instruments. This fragmentation takes time away from picture interpretation and fuels burnout. Voio restores steadiness with a unified studying atmosphere powered by frontier vision-language fashions. Voio’s fashions interpret full exams and draft high-quality experiences so radiologists can evaluation and finalize quicker, with out sacrificing accuracy.

“Radiologists shouldn’t have to select between velocity and high quality,” stated Dr. Maggie Chung, Co-Founder and Medical Lead of Voio. “Our aim is to make radiology reporting seamless by drafting full experiences and connecting pictures, historical past, and prior exams into one clever platform that feels pure to use. By re-designing the reporting expertise from the bottom up, we will scale back unsatisfying grunt work and let radiologists concentrate on affected person care.”

Pillar-0, the staff’s open supply mannequin, achieved a .87 AUC throughout 350+ findings in chest CT, stomach CT, mind CT, and breast MRI scans. This benchmark outperforms all publicly obtainable AI fashions for radiology on the identical take a look at knowledge, together with Google’s MedGemma (.76 AUC), Microsoft’s MI2 (.75 AUC), and Alibaba’s Lingshu (.70 AUC). Moreover, P0 can simply be prolonged to deal with frontier scientific challenges; by positive tuning P0, the staff improved over the state-of-the-art in predicting future lung most cancers, Sybil-1, by 7% in an exterior validation examine at Massachusetts General Hospital. The staff is now scaling the know-how to help multi-modal agentic workflows in radiology and push preventative care.

“The Voio staff is displaying that AI analysis can change healthcare outcomes. This work represents a basic shift in healthcare from reactive analysis to predictive medication,” stated Andy Konwinski, Co-Founder of Laude Ventures, Databricks, and Perplexity. “The means to determine future well being dangers from present imaging earlier than signs seem might remodel how we strategy preventive care, making radiology central to proactive well being administration moderately than simply documenting what’s already unsuitable.”

The know-how builds on the groups’ confirmed monitor file of translating AI analysis into scientific observe. Dr. Yala beforehand created Mirai, a breast most cancers threat prediction mannequin validated throughout 2 million mammograms, and Sybil, which predicts lung most cancers threat from screening scans. Dr. Maggie Chung has led a potential examine demonstrating how AI can scale back diagnostic workup time for high-risk breast most cancers sufferers. Dr. Trevor Darrell based Berkeley’s BAIR lab and led the staff which constructed the Caffe deep studying framework which revolutionized picture processing.

“Voio is the end result of almost a decade pushing the frontier of AI for well being, constructing main fashions for most cancers imaging and validating them globally,” stated Adam Yala, Co-Founder and CEO of Voio. “With Voio, we’re scaling our impression throughout radiology; we’re constructing a frontier AI lab and a unified studying platform that helps radiologists throughout each examine and activity. Over time, this basis will allow richer AI programs that collaborate seamlessly throughout modalities and specialties.”

“Adam, Maggie, and Trevor symbolize precisely what we glance for at The House Fund – area professional Berkeley founders combining world-class AI analysis with confirmed scientific impression. Voio is constructing the infrastructure layer that can outline the subsequent technology of radiology, and we’re thrilled to help their mission,” stated Jeremy Fiance, Founder of The House Fund in Berkeley.

Voio’s open-source strategy allows tutorial researchers to validate and construct upon its fashions, creating transparency in a market the place efficiency claims are sometimes unverifiable. The firm plans to help the event of impartial benchmarks to set up evidence-based requirements for radiology AI.

The submit Voio Emerges From Stealth to Build Frontier AI for Healthcare first appeared on AI-Tech Park.