A fast 3-dimensional neuronal tree reconstruction system that uses cubic polynomials to estimate dendritic curvature.

The main goal of this work was to develop and test the accuracy of our 3DARBOR neuronal tree reconstruction system by comparing it with a very precise but time-consuming method of reconstruction (NEUTRACE). Comparison was performed by reconstructing 18 dendritic trees of frog spinal motoneurons from serial sections with both methods and comparing several morphological summaries of the two reconstructions. In 3DARBOR the planar projection of the dendritic trees was drawn and fed into an IBM-compatible PC through a graphic tablet. Dendritic coordinates along the perpendicular (focus) axis on the plane of drawing were estimated by an interpolation method. The interpolation was based on the lengths of projected dendrites and the coordinates of points where dendrites entered the next section. Focus axis coordinates of these points could automatically be calculated from the serial numbers and thicknesses of sections. 3DARBOR was tested by comparison of the distributions of characteristic points of dendritic trees, segment lengths and branching angles. Product moment analysis on dendritic trees was also performed. It was concluded that 3DARBOR is a fast enough reconstruction system without any systematical error of interpolation that can correctly supply the most morphological parameters and visualize the natural arborizations.

3D reconstruction based on hard tissue microtome cross-section pictures in dentistry.

The aim of this study was to examine the accuracy of the computerized 3D surface analyzing and volume measuring method in dentistry. Two different types of test objects were used in the first part of the measurements. Each sample of the two groups was cross-sectioned using a hard tissue microtome. The sections were photographed on both sides and were projected on a graphical tablet and analyzed using a computer program. The measured and calculated parameters were compared. In the second part, 200 microm thick horizontal sections were prepared from 11 human incisor roots using the hard tissue microtome. This way, five sections were prepared from the apical 2 mm of each root. The effects of section thickness and number were modeled by decreasing the inclusion rate of the obtained number of sections from 10 to 2 and its influence on the calculated results was determined. This method was suitable for the approximation and analysis of 3D parameters. The results indicated that using 200-300 microm section thickness, the measured values were approximately 8-21% lower than the real parameters.