It is modeled after the Cornell Image Analysis Task Archive (CIAT)43 and carries the same guarantee of interoperability, inspired by the Cornell Machine Learning Repository (Cornell MLR)44, to provide centralized access to raw microarray image data. Data can be published to CIAT for use by the community. Data modeling is flexible and allows for use of existing data models. CIAT provides a range of quality metrics to assess data quality for use by data providers.
Morphological Profiler is a contour-based and edge-based template-fitting algorithm that takes a grayscale cell-edge image as input, and outputs a bitmap image with semiautomatically contoured edges. For this task, the input image is assumed to be a grayscale cell-edge image, such as the binary image of transmission electron (TEM) micrographs. Only the cell-edge image needs to be aligned to a standard cell format (for example, TEM). For a two-dimensional image set, Morphological Profiler returns contours as a bitmap image of 41 Which was also used in ref. 42 [A simple, inexpensive, nearly-universal approach to biomedical image classification based on machine learning] in ref. Cutting edge micrograph-analysis algorithms can yield comparable results as those presented in this paper; for example, ref. 43 [TEM image analysis using adaptive thresholding and contour processing] in ref. 44 [The Cornell Machine Learning Repository - an extensible, scalable, and well-documented package of machine-learning algorithms] and ref. 45 [A simple, scalable, and well-documented collection of machine-learning algorithms for medical image analysis] in ref. A TEM micrograph of a cell edge. Two contours are drawn, one beneath the cell membrane and one above the cell membrane, labeled 'b' and 'a', respectively. The third contour, labeled 'center' is at the center of the cell and is easy to judge. The contours are drawn automatically by the algorithm. It uses an adaptive thresholding technique (ref. ), and a template-fitting algorithm (ref. ), and does not require any manual intervention. d2c66b5586