AI in Healthcare and Scienceโ€ขโ€ข5 min readโ€ข818 words

๐Ÿš€ Medical Imaging - CERVINO Study Results

โšกDirect Technical Summary

The CERVINO study, presented at the #IMS26 plenary session, reported efficacy and safety results from phase 3 in relapsed or refractory multiple myeloma (RRMM). The study compared

๐Ÿš€ Medical Imaging - CERVINO Study Results

The CERVINO study, presented at the #IMS26 plenary session, reported efficacy and safety results from phase 3 in relapsed or refractory multiple myeloma (RRMM). The study compared etentamig to investigator-choice SAT in triple-class exposed, BCMA-naive RRMM patients. The results showed an ORR of 74% vs 46% (P<0.0001) and a PFS HR of 0.40, with median NR vs 6.2mo.

Key Points:

  • CERVINO Study Design: The CERVINO study was a phase 3 trial that compared etentamig to investigator-choice SAT in triple-class exposed, BCMA-naive RRMM patients.

  • Efficacy Results: The study reported an ORR of 74% vs 46% (P<0.0001) in favor of etentamig.

  • Safety Results: The study reported a PFS HR of 0.40, with median NR vs 6.2mo, and a 12-mo OS of 87.9% vs 72% (early signal).

  • Adverse Events: The study reported a CRS of 39.5%, mostly low-grade.

๐Ÿ”— Resources:


๐Ÿค– Clinical AI - CDI Query Engine

At Becker's Executive Summit, Abridge previewed its CDI Query Engine, built to weigh when a documentation gap is worth surfacing to a clinician, and when to stay silent. The engine is designed to provide restraint in clinical AI, earning trust through careful decision-making.

Key Points:

  • CDI Query Engine: The CDI Query Engine is a tool designed to weigh documentation gaps and decide when to surface them to clinicians.

  • Restraint in Clinical AI: The engine is designed to provide restraint in clinical AI, earning trust through careful decision-making.

  • Clinical AI Trust: Restraint is a key factor in building trust in clinical AI systems.

๐Ÿ”— Resources:


๐Ÿš€ Behavior Recognition Models

Metofico has developed several behavior recognition models that are generalizing well to unseen footage. The models will be available to try at SFN, and attendees are encouraged to bring footage to see how they work.

Key Points:

  • Behavior Recognition Models: Metofico has developed several behavior recognition models that are generalizing well to unseen footage.

  • Model Availability: The models will be available to try at SFN.

  • Footage Testing: Attendees are encouraged to bring footage to see how the models work.

๐Ÿ”— Resources:


๐Ÿš€ Regulatory Technology & Innovation

DNAnexus will discuss the data integration foundation that makes AI in global eSubmissions work at the Global Pharmaceutical Regulatory Affairs Summit in Brussels. The discussion will cover semantic alignment strategies, structured, standards-driven exchange, and building the data foundation.

Key Points:

  • Data Integration Foundation: DNAnexus will discuss the data integration foundation that makes AI in global eSubmissions work.

  • Semantic Alignment Strategies: The discussion will cover semantic alignment strategies that let AI understand submission data.

  • Structured, Standards-Driven Exchange: The discussion will cover the importance of structured, standards-driven exchange in AI-assisted review.

  • Building the Data Foundation: The discussion will cover how to build the data foundation for AI in global eSubmissions.

๐Ÿ”— Resources:


๐Ÿš€ Cloud Layer for Modern Regulatory Collaboration

DNAnexus is enabling faster global submission through its cloud layer for modern regulatory collaboration, Trusted Regulatory Spaces.

Key Points:

  • Cloud Layer for Regulatory Collaboration: DNAnexus is enabling faster global submission through its cloud layer for modern regulatory collaboration.

  • Trusted Regulatory Spaces: The cloud layer is designed to provide Trusted Regulatory Spaces for modern regulatory collaboration.

๐Ÿ”— Resources:


๐Ÿš€ Genome-Wide Map

Recursion Pharmaceuticals has created a genome-wide map that required over 1 trillion neurons, roughly the equivalent of 12 human brains. The map is designed to study diseases by working with cells that actually resemble them.

Key Points:

  • Genome-Wide Map: Recursion Pharmaceuticals has created a genome-wide map that required over 1 trillion neurons.

  • Disease Study: The map is designed to study diseases by working with cells that actually resemble them.

๐Ÿ”— Resources:


๐Ÿš€ Multiple Myeloma Study Results

The #IMS26 plenary session presented the first efficacy and safety results from phase 3 CERVINO study in relapsed or refractory multiple myeloma (RRMM). The study compared etentamig to investigator-choice SAT in triple-class exposed, BCMA-naive RRMM patients.

Key Points:

  • CERVINO Study Results: The study reported efficacy and safety results from phase 3 in RRMM.

  • Efficacy Results: The study reported an ORR of 74% vs 46% (P<0.0001) in favor of etentamig.

  • Safety Results: The study reported a PFS HR of 0.40, with median NR vs 6.2mo, and a 12-mo OS of 87.9% vs 72% (early signal).

๐Ÿ”— Resources:

๐Ÿ“‚Source / Implementation:AI in Healthcare and Science / resources-262.md
GitHub Repositoryโ†—

Related AI in Healthcare and Science Breakdowns

Drishtant Ghosh (Drix10)
Drishtant Ghosh (Drix10)โ€ขAuthor & Engineer

Technical founder and engineer working across AI systems, developer infrastructure, and cybersecurity.

PortfolioยทGitHubยทLinkedInยทXยทEmail