We use cookies to understand how you use our site and to improve your experience. This includes personalizing content and advertising. To learn more, click here. By continuing to use our site, you accept our use of cookies. Cookie Policy.

Features Partner Sites Information LinkXpress hp
Sign In
Advertise with Us
PURITAN MEDICAL

Download Mobile App




AI Performs Virtual Tissue Staining at Super-Resolution

By LabMedica International staff writers
Posted on 03 Jul 2025

Conventional histopathology, essential for diagnosing various diseases, typically involves chemically staining tissue samples to reveal cellular structures under a microscope. More...

This process, known as “histochemical staining,” is laborious, time-intensive, and requires expensive chemical reagents. It also damages the tissue, making it unusable for further analysis. To address these drawbacks, a technique called “virtual staining” has gained prominence. This method uses computational tools to convert images of unstained tissue into digital versions that resemble chemically stained samples, eliminating the need for physical dyes or chemical treatments. Now, researchers have introduced an AI-based technique that virtually stains unlabeled tissue samples at a resolution much higher than the original input image, completely bypassing the use of chemical dyes or staining methods.

This pixel super-resolution virtual staining technique, developed by researchers at the University of California, Los Angeles (UCLA, Los Angeles, CA, USA), transforms low-resolution autofluorescence images of unstained tissue into high-quality, higher-resolution brightfield images that accurately mimic histochemically stained tissue, including the widely used hematoxylin and eosin (H&E) stain. This approach results in a 4- to 5-fold increase in spatial resolution, significantly improving the clarity and diagnostic value of the resulting images.A key innovation lies in the model’s control over the randomness typically found in diffusion models. Through the use of novel sampling methods—such as mean sampling and averaging—the researchers were able to minimize variation between images, providing consistent and reliable results suitable for clinical diagnostics. In blind tests using human lung tissue, the diffusion-based pixel super-resolution virtual staining model outperformed existing techniques in terms of resolution, structural similarity, and perceptual accuracy. The findings, published in Nature Communications, showed that a board-certified pathologist found complete agreement between the AI-generated images and those obtained through conventional staining across a range of tissue structures.

The model’s versatility was also validated through successful transfer learning on human heart tissue, demonstrating consistently high accuracy and resolution across different tissue types. This AI-driven method removes the need for chemical staining, offering advantages in terms of time, cost, and preservation of tissue integrity. The innovation holds significant promise for streamlining digital pathology workflows, particularly in settings where resources are limited or rapid diagnostics are critical. By integrating pixel super-resolution with virtual staining, this approach enables high-definition digital pathology and moves closer to enabling precision medicine, without relying on a lab stocked with chemical reagents. The research highlights the revolutionary potential of generative AI in computational pathology and establishes a new benchmark for high-resolution, reliable virtual staining of unstained tissue samples.

“Diffusion models are powerful, but their randomness is a double-edged sword,” said senior author Professor Aydogan Ozcan. “We introduced a way to tame that randomness, giving us control and consistency during inference-which is essential for clinical applications.”

Related Links:
UCLA


New
Gold Member
Hematology Analyzer
Medonic M32B
Serological Pipet Controller
PIPETBOY GENIUS
New
Urine Chemistry Control
Dropper Urine Chemistry Control
New
Rapid Test Reader
DIA5000
Read the full article by registering today, it's FREE! It's Free!
Register now for FREE to LabMedica.com and get access to news and events that shape the world of Clinical Laboratory Medicine.
  • Free digital version edition of LabMedica International sent by email on regular basis
  • Free print version of LabMedica International magazine (available only outside USA and Canada).
  • Free and unlimited access to back issues of LabMedica International in digital format
  • Free LabMedica International Newsletter sent every week containing the latest news
  • Free breaking news sent via email
  • Free access to Events Calendar
  • Free access to LinkXpress new product services
  • REGISTRATION IS FREE AND EASY!
Click here to Register








DIASOURCE (A Biovendor Company)

Channels

Hematology

view channel
Image: The microfluidic device for passive separation of platelet-rich plasma from whole blood (Photo courtesy of University of the Basque Country)

Portable and Disposable Device Obtains Platelet-Rich Plasma Without Complex Equipment

Platelet-rich plasma (PRP) plays a crucial role in regenerative medicine due to its ability to accelerate healing and repair tissue. However, obtaining PRP traditionally requires expensive centrifugation... Read more

Immunology

view channel
Image: The 3D paper-based analytical device has shown high clinical accuracy for adult-onset immunodeficiency (Photo courtesy of National Taiwan University)

Paper-Based Device Accurately Detects Immune Defects in 10 Minutes

Patients with hidden immune defects are especially vulnerable to severe and persistent infections, often due to autoantibodies that block interferon-gamma (IFN-γ), a key molecule in immune defense.... Read more
Copyright © 2000-2025 Globetech Media. All rights reserved.