When pathologists examine tissue, they typically study thin, stained sections under a microscope. These provide detailed information, but only show a two-dimensional view of a small part of the tissue. Researchers from TU Delft and Erasmus MC are developing a technique that allows them to examine larger samples of prostate tissue in 3D, without staining. The first 3D images of human prostate tissue mark an important technical milestone.
.

From zebrafish to prostate tissue

“It actually started in 2018, with a PhD researcher studying optical tomography in zebrafish,” explains Jeroen Kalkman from the Department of Imaging Physics at TU Delft. Zebrafish are useful models for studying many diseases, making them particularly interesting for imaging research. They are small enough to image in their entirety, while their organs are sufficiently complex to develop and refine imaging techniques.

Around the same time, Martin van Royen, working with pathologist Arno van Leenders at the Department of Pathology and Clinical Bioinformatics at Erasmus MC, was studying the three-dimensional architecture of healthy prostate tissue and prostate cancer. Following a presentation by Kalkman, the researchers began exploring whether the optical tomography expertise developed in Delft could also be used to image human prostate tissue in 3D.

Through the Convergence Health & Technology Flagship Convergence Imaging Facility and Innovation Center (CIFIC) , researchers from TU Delft and Erasmus MC had the opportunity to develop the technique together.
.

Why examine tissue in 3D?

To determine whether tissue contains abnormal or malignant cells, pathologists typically process it, cut it into very thin sections and apply stains. These stained sections are then examined under a microscope. This approach has formed the basis of modern pathology for more than 150 years and provides highly detailed information.

“However, a thin section only provides a two-dimensional view, offering limited insight into the tissue’s overall architecture,” explains Martin van Royen. “We want to understand how the cells that normally form glands in the prostate are distributed in cancerous tissue, what shape these glands take, how they branch and how they connect.”

This spatial architecture is particularly important in prostate cancer, as pathologists classify tumour malignancy partly by examining glandular growth patterns. Depending on where and how tissue is sectioned, the same three-dimensional structure can look very different in 2D. Three-dimensional imaging could provide additional insight into how tumours grow.

.

A new approach: Optical Diffraction Tomography

The conventional way to create a 3D image of tissue involves imaging numerous individual sections. However, cutting, staining, imaging and digitally reconstructing all these sections is labour-intensive. Small distortions can also occur during sectioning and reconstruction, and the method typically captures only part of the total tissue volume. A more recent alternative uses confocal microscopy and fluorescent staining of optically cleared tissue.

Optical Diffraction Tomography (ODT) takes a different approach. It uses differences in the tissue’s refractive index to create contrast between structures, eliminating the need for stains or fluorescent labels. ODT reconstructs a complete 3D image from projections of transmitted light, without first cutting the sample into numerous sections.

This combination makes the technique particularly interesting: it allows researchers to examine larger, continuous volumes of tissue in 3D, including deeper structures, without staining.

.

Using light to reveal tissue structures

ODT relies on differences in the refractive index of tissue, which determines how light slows down and bends as it passes through a material. Water, fats, proteins and different cell components all have slightly different optical properties. As light travels through tissue, these differences cause subtle changes in the direction and phase of the light wave.

.

To measure subtle differences in refractive index, the imaging system uses digital holography. A laser beam is split into two: one passes through the tissue, while the other serves as a reference. When the beams are recombined, they produce an interference pattern known as a digital hologram. This allows researchers to calculate how the tissue has altered the light wave.

A single measurement is not enough to generate a complete 3D image. The tissue is therefore measured from many different directions, and algorithms combine thousands of measurements into a three-dimensional reconstruction. The principle is similar to a CT scan, which combines images taken from different angles to create a 3D reconstruction. However, while CT scanners use X-rays, ODT uses laser light to measure microscopic differences in refractive index.

What could this technique achieve?

The researchers have successfully produced 3D images of prostate tissue, demonstrating that the entire process works: from tissue clearing and optical measurements to digital reconstruction.

The technique allows researchers to image relatively large structures in human prostate tissue in three dimensions, without staining or labelling. Rather than examining individual sections, they can investigate how structures extend through the tissue and connect with one another. They can also measure characteristics such as shape, size, branching and connectivity. In the longer term, this could contribute to a more quantitative assessment of clinical tissue samples. However, the research has not yet reached that stage. The current images demonstrate that the technique works, but they do not establish whether ODT can independently identify prostate cancer or classify it more accurately than existing pathology methods.

“I see ODT as a complement to conventional pathology rather than a replacement,” Kalkman emphasises. “Extensive research and validation will be needed before it can be used clinically.” For now, the researchers primarily consider ODT a research tool, although their ambition is to develop an approach that could eventually offer clinical value.

.

The next step

The next step is to systematically compare ODT images from more patient samples with stained tissue sections and pathologists’ assessments. This should help establish which growth patterns can be identified in 3D, whether healthy and malignant tissue can be reliably distinguished, and what additional information ODT provides.

If further research demonstrates that the method is sufficiently reliable and practical, ODT could eventually complement conventional pathology by providing a three-dimensional view of tissue.

.

Researchers involved

The project is led by Jeroen Kalkman from the Department of Imaging Physics at TU Delft and Martin van Royen from the Department of Pathology and Clinical Bioinformatics at Erasmus MC. Vahid Khandan worked as a postdoctoral researcher on the development of the ODT platform. Other researchers involved include Alim Yolalmaz, Ofir Heymans, Johan Slotman, Jos van Rooij, Gijs Hooghiemstra and uropathologist Arno van Leenders.