Google has opened source a Python library called EZ WSI DICOMWeb to simplify operations and make IT easier for developers to access and retrieve full-slide image WSI information from cloud DICOM (Medical Digital Image Transfer Protocol) storage. So as to promote the development of the application of digital pathology artificial intelligence.

Because of the huge capacity of full glass WSI high-resolution image, it is not a simple thing to retrieve specific WSI information from DICOM storage using DICOMweb. Therefore, the EZ WSI DICOMWeb Python library developed by Google aims to simplify these operations, efficiently and simply access WSI block images, and make WSI easy to share and access.

Retrieving block images locally not only increases network traffic usage costs, but also creates more latency and takes up a lot of storage space compared to the traditional method of downloading the complete WSI from DICOM storage. The EZ WSI DICOMWeb information base can directly retrieve the required WSI block image through the API, so the image data can be used intuitively and concisely. Developers do not need to deeply understand the data structure and API of DICOM, but can focus more on application development to further promote collaboration and knowledge transfer. And make it easier for researchers to use the data in machine learning techniques to drive AI applications in healthcare.

Note: A pathological biopsy is a tissue sample cut into very thin slices that are stained and viewed under a microscope as part of medical diagnosis. WSI is a technology that digitizing traditional pathological slices and storing them locally or in the cloud for remote diagnosis, education and research.