Related Experiment Video
Updated: Oct 8, 2025
High-Throughput, Multi-Image Cryohistology of Mineralized Tissues
Published on: September 14, 2016
A generative adversarial approach to facilitate archival-quality histopathologic diagnoses from frozen tissue
Kianoush Falahkheirkhah1,2, Tao Guo3, Michael Hwang4
1Department of Chemical and Biomolecular Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, 61801, USA.
Abstract:
In clinical diagnostics and research involving histopathology, formalin-fixed paraffin-embedded (FFPE) tissue is almost universally favored for its superb image quality. However, tissue processing time (>24 h) can slow decision-making. In contrast, fresh frozen (FF) processing (<1 h) can yield rapid information but diagnostic accuracy is suboptimal due to lack of clearing, morphologic deformation and more frequent artifacts. Here, we bridge this gap using artificial intelligence. We synthesize FFPE-like images ("virtual FFPE") from FF images using a generative adversarial network (GAN) from 98 paired kidney samples derived from 40 patients. Five board-certified pathologists evaluated the results in a blinded test. Image quality of the virtual FFPE data was assessed to be high and showed a close resemblance to real FFPE images. Clinical assessments of disease on the virtual FFPE images showed a higher inter-observer agreement compared to FF images. The nearly instantaneously generated virtual FFPE images can not only reduce time to information but can facilitate more precise diagnosis from routine FF images without extraneous costs and effort.
Insights
Artificial intelligence creates "virtual FFPE" images from fresh frozen (FF) tissue, mimicking high-quality formalin-fixed paraffin-embedded (FFPE) slides. This AI approach accelerates diagnostics and improves diagnostic accuracy without extra costs.
Area of Science:
- Computational pathology
- Artificial intelligence in histopathology
- Medical imaging
Background:
- Formalin-fixed paraffin-embedded (FFPE) tissue offers superior image quality for histopathology but requires lengthy processing (>24 hours).
- Fresh frozen (FF) tissue allows rapid processing (<1 hour) but results in suboptimal diagnostic accuracy due to artifacts and poor morphology.
- A significant time gap exists between rapid FF processing and the diagnostic quality of FFPE, impacting clinical decision-making.
Purpose of the Study:
- To develop an artificial intelligence method for generating high-quality FFPE-like images from FF tissue samples.
- To bridge the gap between rapid FF processing and the diagnostic utility of FFPE, enhancing histopathological analysis.
- To improve diagnostic accuracy and inter-observer agreement in histopathology through AI-generated virtual FFPE images.
Main Methods:
- A generative adversarial network (GAN) was employed to synthesize FFPE-like images from FF images.
- The study utilized 98 paired kidney tissue samples from 40 patients for training and validation.
- Five board-certified pathologists conducted a blinded evaluation of the generated virtual FFPE images.
Main Results:
- The generated virtual FFPE images exhibited high quality and closely resembled real FFPE images.
- Clinical assessments using virtual FFPE images demonstrated improved inter-observer agreement compared to standard FF images.
- The AI method rapidly generated virtual FFPE images, significantly reducing the time to obtain diagnostic information.
Conclusions:
- AI-driven synthesis of virtual FFPE images effectively bridges the diagnostic gap between FF and FFPE tissue processing.
- This approach facilitates more precise diagnoses from routine FF images, offering a faster alternative to traditional FFPE.
- The virtual FFPE method provides a cost-effective and efficient solution for accelerating clinical diagnostics and research in histopathology.
