Automatic Quantification of High S:N Images Using VCS

DESCRIPTION (provided by applicant): Theoretically unbiased stereology has been applied to the morphometric analysis of biological tissue sections in the Stereologer, a commercially available, integrated hardware-software-microscopy system. A general limitation of this and other such systems lies in the time- and labor-intensive requirement for identification of targeted objects by trained non-expert users. This Phase 1 SBIR will test the accuracy, efficiency, and feasibility of a novel modification -- Verified Computerized Stereoanalysis (VCS) - for automatic quantification of stereological parameters for objects with a high signal-to-noise ratio (S:N). This automatic combination of image analysis and unbiased stereology will be implemented on the Stereologer system and evaluated through a two-step process: systematic-random sampling under automatic stage control; and, automatic quantification of mean object size and variation. To avoid the collection of spurious data, a verification process after step one will ensure that minimal S:N of equal to or > 90% exists prior to automatic data collection. Accuracy will be evaluated through parallel automatic and semi-automatic analyses on routine tissue sections and by external beta test sites. The VCS approach has the potential to drastically improve throughput of stereological studies, without a loss of accuracy, and thereby stimulate progress in a wide range of biomedical research projects.