Related Experiment Video
Updated: Oct 4, 2025
A Rapid Method for Multispectral Fluorescence Imaging of Frozen Tissue Sections
Published on: March 30, 2020
On the Acceptance of "Fake" Histopathology: A Study on Frozen Sections Optimized with Deep Learning
Mario Siller1, Lea Maria Stangassinger2, Christina Kreutzer3
1Department of Information Technology and System Management, Salzburg University of Applied Sciences, Salzburg, Austria.
Background:
The fast acquisition process of frozen sections allows surgeons to wait for histological findings during the interventions to base intrasurgical decisions on the outcome of the histology. Compared with paraffin sections, however, the quality of frozen sections is often strongly reduced, leading to a lower diagnostic accuracy. Deep neural networks are capable of modifying specific characteristics of digital histological images. Particularly, generative adversarial networks proved to be effective tools to learn about translation between two modalities, based on two unconnected data sets only. The positive effects of such deep learning-based image optimization on computer-aided diagnosis have already been shown. However, since fully automated diagnosis is controversial, the application of enhanced images for visual clinical assessment is currently probably of even higher relevance.
Methods:
Three different deep learning-based generative adversarial networks were investigated. The methods were used to translate frozen sections into virtual paraffin sections. Overall, 40 frozen sections were processed. For training, 40 further paraffin sections were available. We investigated how pathologists assess the quality of the different image translation approaches and whether experts are able to distinguish between virtual and real digital pathology.
Results:
Pathologists' detection accuracy of virtual paraffin sections (from pairs consisting of a frozen and a paraffin section) was between 0.62 and 0.97. Overall, in 59% of images, the virtual section was assessed as more appropriate for a diagnosis. In 53% of images, the deep learning approach was preferred to conventional stain normalization (SN).
Conclusion:
Overall, expert assessment indicated slightly improved visual properties of converted images and a high similarity to real paraffin sections. The observed high variability showed clear differences in personal preferences.
Insights
Deep learning methods can convert frozen sections into virtual paraffin sections, improving diagnostic accuracy for surgeons. Pathologists found these AI-generated images highly similar to real ones, aiding clinical assessment.
Area of Science:
- Digital Pathology
- Artificial Intelligence in Histology
- Medical Imaging
Background:
- Frozen sections enable intraoperative histological assessment but often have reduced quality and diagnostic accuracy compared to paraffin sections.
- Deep neural networks, particularly generative adversarial networks, can enhance digital histological images.
- AI-driven image optimization shows promise for computer-aided diagnosis and visual clinical assessment.
Purpose of the Study:
- To investigate the effectiveness of deep learning-based generative adversarial networks in translating frozen sections into virtual paraffin sections.
- To assess pathologists' evaluation of the quality of these virtual paraffin sections.
- To determine if experts can differentiate between AI-generated virtual sections and real digital pathology images.
Main Methods:
- Three deep learning-based generative adversarial networks were employed for image translation.
- Forty frozen sections were processed and translated into virtual paraffin sections.
- Pathologists evaluated the quality and diagnostic utility of the generated virtual sections.
Main Results:
- Pathologists' detection accuracy for virtual paraffin sections ranged from 0.62 to 0.97.
- In 59% of cases, virtual sections were deemed more appropriate for diagnosis.
- The deep learning approach was preferred over conventional stain normalization in 53% of images.
Conclusions:
- Expert assessment revealed slightly improved visual properties in converted images, with high similarity to real paraffin sections.
- The study highlights the potential of AI for enhancing frozen section quality for clinical use.
- Significant variability in expert preferences underscores the need for further research into personalized AI applications.
