JelloX Biotech Inc., a Taiwan-based startup focused on cancer pathology, announced involvement with two new research breakthroughs that demonstrate expanded potential for its specialized technology combining 3D imaging and AI.
3D imaging and digital pathology are currently viewed by the healthcare industry as prohibitively expensive due to the need for costly graphics tools, but JelloX’s solution for this is MetaLite® — a cost-effective software platform that leverages AI to analyze 3D-rendered thick tissue images rapidly, saving pathologists hours while improving diagnostic precision.
Recommended AI: aelf Announces the form of aelf DAO, Enhancing Decentralization of Governance and Ecosystem Growth
The company’s work is timely, as a recent study released in The Harvard Gazette shows that cancer risks are rising with each new generation.
Advancing precision medicine beyond oncology
While focusing mainly on cancer-specific applications, JelloX believes strongly in the possibilities of 3D imaging and AI. Faced with the recent prospect of demonstrating this technology’s potential beyond oncology, the company was eager to provide fast and informative service to its customers interested in this area. This included Dr. Lin-Shien Fu, Director of the Pediatric Nephrology and Immunology Department at Taiwan’s Taichung Veterans General Hospital, whose recent work highlights new potential of 3D pathology.
Published in February 2023, Dr. Fu et al. released a paper that showed that 3D pathology can facilitate pathological analysis of kidney inflammation from lupus, also known as lupus nephritis.
“3D imaging from tissue that is hundreds of microns thick captures much greater information compared with a traditional 4-micron-thin slide, and the study is a key milestone in showing its usefulness to identify and pinpoint where exactly disease is affecting the body, in this case for kidney inflammation from lupus,” notes Dr. Yen-Yin Lin, CEO of JelloX. “If 3D imaging can ultimately help doctors administer medicine with extreme accuracy across a wide variety of diseases, the possibilities are endless for healthcare — from keeping chronically ill patients healthy to saving lives.”
In the study, researchers compared conventional 2D biopsy samples with 3D morphology imaging of samples, which yielded observations that the 3D samples provided more comprehensive spatial information on important components of chronic inflammation. In turn, this means an improved prognosis for lupus nephritis.
The study also paves the way for the new 3D pathology method to address a major unresolved area of research: how chronic kidney damage progresses in different kidney diseases.
Foray into cross-staining solutions for precision medicine
In its forte of cancer pathology, JelloX experts published a recent research paper to demonstrate that the AI deep learning techniques powering JelloX’s technology, and AI more broadly, can effectively analyze a combination of image types and distinguish between them.
Specific to the growing threat of breast cancer, this study combined two different types of staining dyes — bright-field hematoxylin and eosin (H&E) and dark-field fluorescent (FLUO) dyes — for 2D physical biopsy analysis via light microscopy and confocal microscopy, respectively, and applied a cross-staining inference AI model to the samples. Color variation between the two staining technologies causes challenges for cross-analysis, but JelloX overcame this by applying color normalization and nucleus extraction methods.
Recommended AI: GoodFirms Unlocks the Best Task Management Software with Rich Features
The results demonstrated that the cross-staining inference AI model had high precision in identifying tumor features. This achievement is a breakthrough in that it demonstrates the world’s first universal cross-staining solution for precision identification using both H&E and FLUO images, which paves the way for AI to have a greater impact on tumor recognition and pathology in the future.
Recommended AI: Adelante Enhances Capabilities with Zendesk Setup Solution
[To share your insights with us, please write to sghosh@martechseries.com]
Comments are closed.