A video imaging and image-data-processing system for use in dermatological examinations acquires and processes images of patients' skins. The system processes a current image of a selected area of a patient's skin along with an image of the same area acquired previously, generating a comparison image that highlights any nevi (moles) that have appeared, disappeared, grown, and/or shrunk in the interim.
Most nevi are harmless, but changes into malignant melanoma can occur. Early detection is essential for effective treatment of malignant melanoma. Some patients have hundreds of nevi, so that it is difficult to track changes in the nevi that could signify the development of melanoma. The present image system assists the physician in identifying changes in nevi; the physician can then concentrate on examining any changed nevi.
The system includes a video camera, which is used to acquire the images. Typically, an image covers the patient's back from neck to waist. The first image of this area is digitized and stored on a computer disk. In a subsequent examination, a second image of the same area is digitized, stored, and processed along with the first image to generate the comparison image.
Because of differences between the relative alignments of the camera and patient in acquiring the two images, these images are not exactly registered with each other. Therefore, processing begins with the manual selection of a number (typically between 2 and 20) of nevi for use as landmarks in computing a nonlinear geometric transformation between corresponding points in the two images; the transformation is exact at the landmarks and a close approximation between them.
The average of the intensities of the central pixels of the landmarks in each image is computed, and is designated intensity A. The average of the intensities of all other pixels in each image is computed and designated intensity B. A linear transformation between the A and B intensities of both images is computed and used to correct the contrast between the images after they have been geometrically transformed into registration. A local shading correction is also applied; from the intensity of each pixel in each image, the median pixel intensity of all pixels in a small rectangular area around the pixel is subtracted.
From the shading-corrected, contrast-corrected intensity of each geometrically transformed pixel in the second image, the shading-corrected intensity of the corresponding pixel in the first image is subtracted. To bring negative values into the range of visibility, a middle gray-level intensity is then added to all pixels. The result is the comparison image. Wherever a nevus has grown or appeared between the time of acquisition of the first and second images, a dark spot is generated in the comparison image. Similarly, wherever a nevus has shrunk or disappeared, a bright spot can be seen in the comparison image.
Optionally, the comparison image can be rendered in false color to facilitate examination. Instead of adding a gray level, the negative changes in intensity (darkening changes, which could signify melanoma) are displayed in red, while positive changes in intensity (brightening, signifying disappearance or shrinkage of nevi) are displayed in green. The image can be enhanced even further by restricting red and green coloring to those pixels for which the differences between the first and second images exceed a specified threshold value.
This work was done by Robert H. Selzer of Caltech for NASA's Jet Propulsion Laboratory. For further information, access the Technical Support Package (TSP) free on-line at www.techbriefs.com under the Life Sciences category, or circle no. 139 on the TSP Order Card in this issue to receive a copy by mail ($5 charge).
In accordance with Public Law 96-517, the contractor has elected to retain title to this invention. Inquiries concerning rights for its commercial use should be addressed to
Technology Reporting Office
Mail Stop 122-116
4800 Oak Grove Drive
Pasadena, CA 91109
Refer to NPO-19729, volume and number of this NASA Tech Briefs issue, and the page number.
This Brief includes a Technical Support Package (TSP).
Image-matching system highlights changes in skin lesions
(reference NPO19729) is currently available for download from the TSP library.
Don't have an account? Sign up here.