Medical Hyperspectral Imaging HyperMed, Inc Hyperspectral Imaging Cube Hyperspectral imaging involves observing light waves at a multitude of wavelengths over a certain spatial region, creating a "hyperspectral cube" of information (below, left). This cube consists of hundreds of slices, or images, each acquired at a different wavelength. By extracting the images from the three wavelengths seen by the human visual system (i.e. red, green, and blue), a visual image can be synthesized (below, right). However, instead of choosing the visual wavelengths, any combination of wavelengths can be selected to best visualize the tissue properties of interest for a given application. Hyperspectral cube of forefinger burn. Red, green, and blue slices extracted from the cube next to a color image composited from the slices. Medical Imaging Spectroscopy Another means of representing the hyperspectral cube involves turning it on its side. Every position (or pixel) in the image represents a real point on an object in space (below, left). By looking at the same pixel at each wavelength, one obtains the full spectral response of the object imaged at the corresponding point (below, right). This is similar to traditional spectroscopy, except that the spectra are obtained at every point throughout the entire image, instead of at just one position. In other words, the imaging component of spectroscopy adds two spatial dimensions to traditional optical spectroscopy. Color image composited from the slices with two points selected. Spectra captured from the two points showing different levels of oxygenation.