



RESEARCH PLAN 

Principal Investigator/Program Director Williams, Robert W. 

Gross morphologic measures as dependent measures As part of ongoing research in our laboratory, we have consistently used various gross morphological measures as dependent measures in our studies. For example, we have been interested for a number of years in the biological substrates of anatomical asymmetry. We were initially struck by the simple notion that with respect to asymmetric brain regions, symmetry can result developmentally from either an increase in the normally smaller side, a decrease in the usually larger side, or a combination of the two processes. In the first case, the measure of total brain area of symmetric brain regions would be larger than their asymmetric counterparts whereas the opposite would be true in the second case. Brain areas would be similar if the third scenario were true. To test these alternatives, we examined photographs of the planum temporale to measure total planum area in the left and right sides . Similar to previous findings, we found a leftward asymmetry in the majority of cases. When we plotted the total planum area (right + left) against a measure of directionless asymmetry (magnitude), we found a significant negative correlation, indicating that as asymmetry increased, the total planum area decreased. These results demonstrated that symmetric brains were larger than their asymmetric counterparts and, further, support the hypothesis that asymmetry was the result of the production of a small side rather than the production of a large side. We have since replicated these findings in both rats and mice . We have also been aware for some time of the complications involved in estimating volume from twodimensional structures. We were initially troubled by the potentially different methods of computation involved in estimating volume from crosssectional areas. We compared the efficacy of each of the common methods for estimating volume by using mathematical simulations as well as actual morphometric measures from brain regions. As it turned out, there were no meaningful differences between the various estimates when many sections are used; with fewer sections, however, Cavalieris estimator was most accurate. Yet while the Cavalieri approach provides a better approximation of volume under some circumstances, it requires equally spaced sectionsa criterion sometimes difficult to meet in the real world (e.g., missing sections affect accuracy). We therefore devised methods for the estimation of brain volume from unequally spaced sections; these methods are quite accurate when large numbers of sections are used . These studies demonstrate that in our experience, useful information can be obtained from studies involving gross morphometric measures. Moreover, we have been concerned about the computational problems involved in stereological measures and have devised methods to supplement these efficient and accurate methodologies.


NEXT TOPIC 
