Advanced Digital Pathology with CytoViva Hyperspectral Microscopy

Digital Pathology is the fastest growing area of research within pathology and focuses on data analysis of digitized specimen slides. By digitizing the optical image of a pathology slide, machine learning software can be utilized to improve diagnosis, prognosis and prediction of a wide range of diseases. Today, the digitization of pathology slides primarily focuses on capturing simple optical microscopy images of traditional H&E stained slides. This allows the software to stitch these images together so they can be analyzed based on RGB (red, green, blue) color channels or shape of the tissue. While this process can serve to advance the accuracy of disease diagnosis vs traditional qualitative observations of tissue, it is a limited dataset relative to other imaging modalities.?

Hyperspectral Microscopy of specimen slides represents a potential quantum leap in the advancement of digital pathology. Hyperspectral images are digital images that look very similar to a traditional digitized optical image with a very significant difference. Each pixel of a hyperspectral image contains the full visible near-infrared (VNIR) spectrum from 400nm-1,000nm. This spectral data can be produced at a very high spectral resolution of ~2nm and pixel sizes can be as small as 100nm depending on the camera utilized for the image capture. This incredibly data-rich hyperspectral image provides an opportunity for significantly enhanced digital characterization of a wide range of disease tissue states versus simple digital optical images.?

CytoViva, Inc is the world-wide leader in the field of hyperspectral microscopy and a leading developer of hyperspectral imaging-related digital pathology applications. CytoViva provides turnkey hyperspectral microscope systems that include custom microscopy configurations, including its patented enhanced darkfield and traditional brightfield imaging. CytoViva utilizes high-resolution, line scan-based hyperspectral imaging integrated onto the microscope, which provides 2nm of spectral resolution across the entire VNIR 4000nm-1,000nm spectral range. These systems also come equipped with a full suite of custom hyperspectral image analysis functionality, which provides significant utility in pathology-related applications.

The examples in this application note serve to illustrate a simple utilization of hyperspectral microscopy for pathology-based initiatives. The first example in Figure 1 shows a hyperspectral image of H&E stained squamous cell carcinoma tissue under traditional brightfield microscopy illumination.

Figure 2 is a negative control or non-cancerous hyperspectral image of the same tissue.

Figure 3 is a hyperspectral image of the cancerous tissue, with mapping in green of all pixels containing optical spectrum unique to the cancerous elements of the tissue.?The spectrum used for spectral mapping was identified by comparing thousands of pixels in the positive control sample versus all pixels in the negative control sample using a proprietary supervised machine learning algorithm known as Filter Spectral Library. This algorithm identified spectrum unique to the cancerous tissue, which was created as a spectral library. Then, a mapping algorithm known as Spectral Angle Mapper was utilized to map the cancerous elements based on the created spectral library.

In Figure 4, an example mean spectrum is shown to illustrate nuanced spectral differences between the control tissue and areas mapping for cancerous tissue.?

CytoViva, Inc. is the world-wide leader in the field of hyperspectral microscopy and a leading developer of hyperspectral imaging-related digital pathology applications. CytoViva provides turnkey hyperspectral microscope systems that include custom microscopy configurations, including its patented enhanced darkfield and traditional brightfield imaging. CytoViva utilizes high-resolution, line scan-based hyperspectral imaging integrated onto the microscope, which provides 2nm of spectral resolution across the entire VNIR 4000nm-1,000nm spectral range. These systems also come equipped with a full suite of custom hyperspectral image analysis functionality, which provides significant utility in pathology-related applications.

These two examples illustrate how hyperspectral microscopy can be easily applied to quantitatively identify cancerous tissue in both traditional stained and unstained environments. This system and these analysis principles can be applied across many different tissue samples to identify a wide range of disease states. To learn more about hyperspectral microscopy for pathology-related applications, please contact CytoViva at [email protected].???

Visit the Schaefer Technologie GmbH Homepage, Learn More about Different Products from CytoViva and Book a Demo Today: CytoViva | Nanoparticles | Schaefer Group (schaefer-tec.com)