<img height="1" width="1" style="display:none;" alt="" src="https://px.ads.linkedin.com/collect/?pid=4958233&amp;fmt=gif">
Case study

ZEISS Enhances Cloud-Based Image Processing With AI Denoising

Healthcare and Life Sciences

Challenge

AI presents major opportunities for the life sciences. To improve its cloud-based microscopy solutions platform, ZEISS, the world's leading manufacturer of optical systems, enlisted us to develop a set of modules for automatically improving the quality of images.

Outcome

Our engineers integrated with ZEISS personnel to form a single team. The collaboration produced new training modules and pre-trained models for automatically removing the noise from microscopy images.

ZEISS and APEER: simplifying customised microscopy solutions

ZEISS is the leading manufacturer of optical systems and optoelectronics. Their hardware and software solutions are applied in fields spanning semiconductor manufacturing, medical technology, material research, industrial metrology, and research into micro- and nanostructures using microscopy.

Microscopy users are no longer solely interested in obtaining high-quality images; they are now seeking complete application solutions. The field of microscopy applications is vast and varied, and it is not feasible for any single organisation to provide all necessary software solutions. As a result, ZEISS created APEER, a cloud-based platform that simplifies the development and implementation of customized microscopy solutions.

Developing high-performance microscopy modules

Endava’s advanced technical proficiency in AI, machine learning, and data science, plus our keen understanding of microscopy users’ needs, stood out to the ZEISS team.

 

The Endava developers and ZEISS collaborated closely as a single team to create a deep learning denoising module for APEER, along with other supportive modules. These included a training module for denoising images with the Noise2Void algorithm, as well as pre-trained models for denoising LSM and electron microscope images.

 

Endava also developed various other image processing modules, such as a watershed-based object segmenter and a signal-to-noise estimator. Our teams assisted in organizing these modules into user-friendly workflows – available for review here, here and here.

 

Improving through innovation

ZEISS sets a high standard which Endava is pleased to have met. Endava looks forward to further innovative collaborations with ZEISS in the future.

 

Team APEER and our users are very happy with the performance of the modules and workflows created so far with Endava. We highly appreciate these module improvements which have made our products better.

Dr. Roman Zinner

Senior Product Manager, ZEISS

Interesting? We love when people share.

We can help you reach new heights

Let's connect