Skip to main content

Efficient storage, access and transmission of whole-slide pathology images

Final Report Summary - PIMCO (Efficient storage, access and transmission of whole-slide pathology images)

Digital pathology images are rapidly becoming an important tool that aid in the diagnosis process of various diseases, such as as cancer, as well as in the understanding of the disease progression.
Despite their many advantages, the large size of the files associated with these images prevent their widespread in clinical and research centers. PIMCO is a project that aims at developing methods to manipulate these large images by compression. During the first two years of this project, important advances to the state-of-the-art have been achieved. These include the development of coding techniques for these images within international compression standards like the High Efficiency Video Coding Standard (HEVC) and JPEG2000. Advances in the area of segmentation have also been proposed, which in turn serve as an important pre-processing tool to help delineate Regions of Interest (RoIs) on the images with minimum manual intervention. These segmentation methods are also important to compress the images differently accordingly to the regions that are most useful to the diagnosis process. RoIs can then be compress at a higher quality than that of other less important regions.
PIMCO has also helped Dr Sanchez to integrate his research into the University of Warwick, UK, and the European Union. This project has been instrumental to strengthen collaborations with universities in Europe and North America, as well as to disseminate research results in international conferences and through the publications of scientific papers in peer-reviewed journals. The second half of PIMCO will focus on continuing developing compression methods for these images with a specific focus on achieving scalable properties. In other words, the main interest will be to encode these images so they can be transmitted and accessed in a progressive manner starting from a very low quality version of the images, e.g. by accessing a limited number of bits, to a full quality version, e.g. by accessing the full bit-stream. Scalability is important when sharing these images through channels of limited bandwidth capacity. Pathologist with a small bandwidth connection can then access these images at a low quality and increase this quality as the bandwidth increases. The second half of this project has supported the integration of Dr Sanchez within the European Union. As part of this integration, current collaborations have been strengthened through research visits and the publications of research results, as well as new collaborations have been created through the disseminations and discussion of results in international conferences.