TLDR: Modella AI and illumiSonics Inc. have announced successful research demonstrating the compatibility of illumiSonics’ Multi-Laser Imaging (MLI™) system with Modella AI’s PathChat™ DX platform. This breakthrough supports a fully digital, AI-assisted pathology workflow, enabling diagnostic interpretation without the need for traditional chemical staining or tissue damage. The PathChat DX platform, which utilizes generative and agentic AI, has received Breakthrough Device Designation from the U.S. FDA.
BOSTON & WATERLOO, Ontario – Modella AI, a prominent entity in generative and agentic artificial intelligence for biomedicine, and illumiSonics Inc., a leader in label-free optical imaging and computational pathology, have jointly announced the successful completion of research demonstrating the seamless compatibility between illumiSonics’ Multi-Laser Imaging (MLI™) system and Modella AI’s PathChat™ DX platform. This significant advancement paves the way for an entirely digital pathology workflow, from initial tissue imaging to AI-enabled diagnostic interpretation, eliminating the need for chemical staining or potential tissue damage.
The PathChat™ DX platform, developed by Modella AI, is a generative AI-powered pathology diagnostic system that has been granted a Breakthrough Device Designation by the U.S. Food and Drug Administration (FDA). The illumiSonics’ MLI™ system, currently for research use only, specializes in non-destructive, label-free imaging for high-content tissue analysis.
In the collaborative study, the illumiSonics’ MLI system generated virtual H&E (Hematoxylin and Eosin-like) images from unstained skin tissue resections, which exhibited various malignant features. These de-identified images were then analyzed by the PathChat DX platform, notably without any prior exposure or training on virtual histology. The diagnostic interpretations provided by PathChat DX showed extremely high concordance when compared to the consensus reads of eight board-certified pathologists specializing in dermatologic pathology, who reviewed both the virtual images and corresponding chemical H&E slides.
Jill Stefanelli, CEO of Modella AI, emphasized the transformative nature of this research, stating, “Enabling diagnostic-grade AI inference from label-free slides is a transformative step forward and advances the vision of digital-first pathology workflows that preserve tissue integrity and can streamline diagnostic throughput.” John Mackey, CEO of illumiSonics, added, “Our MLI system produces rich, label-free images suitable for computational analysis. The compatibility study with PathChat DX demonstrates the potential of virtual pathology in automating diagnostic workflows, opening new possibilities for precision medicine and translational research.”
Also Read:
- Mayo Clinic Pioneers Medical AI with NVIDIA Blackwell Supercomputing Deployment
- AI-Powered ‘Virtual Scientists’ Accelerate Biological Discovery
This novel research underscores the advantages of label-free imaging over traditional chemical staining, which is irreversible, labor-intensive, and can interfere with subsequent molecular testing. The MLI platform’s ability to produce high-resolution virtual H&E images without dyes or tissue destruction highlights its potential value in clinical settings where tissue preservation for additional molecular testing is crucial. Both PathChat™ DX and MLI™ platforms are currently for research use only and have not yet received FDA approval for commercial use.
