An Interactive Chromosone Scanning System

To the relatively few groups actively engaged in automated chromosome studies, the title of this paper is rather self explanatory because it proposes to address a problem which has become apparent to all of us who have sought to automate cell analysis. The need for human interaction implies that our automated procedures are not error free and require some degree of human verification and correction within the procedure. Indeed this is so. The fact that a problem exists may be attributed to an over-optimistic concept of the capability of machines to duplicate apparently simple human activity. It follows that certainly one purpose of this paper would be to discuss the need for human interaction in automated chromosome analysis and to suggest an approach to the question of where and how interaction should be implemented. However, so that others who may be unfamiliar with automated chromosome analysis can achieve a better understanding, the format of this paper includes a brief overview of human cytogenetics along with an attempt to treat the significance of chromosome abnormalities in terms of the need for population studies. Unfortunately, the justification for automated procedures in cytogenetics has been based upon the need to reduce the labor and tedium involved in large studies without the need for the large studies themselves being clearly established on a cytogenetic basis.