Reducing Chatter in Knife-Edge Scanning Microscopy

Date

2014-12-18

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The Knife-Edge Scanning Microscope (KESM) employs a novel form of physical sectioning microscopy: Imaging of tissue while sectioning. KESM was developed in the Brain Networks Lab (BNL) at Texas A&M University. The KESM has been used to section animal tissue embedded in a plastic block using a diamond knife. During each cut, the plastic block containing the tissue contacts the knife and that impact induces vibrations, known as knife chatter. These vibrations introduce noise in the image captured from the cut slice. This research is aimed at determining a metric to quantify knife chatter in the images acquired using the KESM, and to test if the use of a vibrating knife reduces knife chatter.

Knife chatter appears as repeated parallel streaks in the images. A quantitative characterization of knife chatter is difficult since there is no regular pattern with which it appears. Having no regular pattern makes it very challenging to detect the chatter using automated programs. Observing the Fast Fourier Transforms of the images tells us that a narrow vertical band around the central vertical axis contains information exclusively about the chatter, while most of the information about the object in the image is outside the vertical band that represents the knife chatter. Using this information, we can quantitatively characterize knife chatter as a ratio of (1) the width of the region in the Fast Fourier Transform that corresponds to the knife chatter, and (2) the width of the region that corresponds to the object.

Determining if the introduction of vibrations in the KESM diamond knife affects the amount of knife chatter present in the images was achieved by sectioning specimens of nissl-stained zebra-fish embryos embedded into araldite blocks, at different knife vibration frequencies. External sinusoidal wave vibrations were introduced in the KESM knife from a signal generator throughout the sectioning process. These electrical signals were converted to mechanical waves at the tip of the KESM knife blade. Performing such experiments at different oscillation frequencies enabled us to compare data using the metric described above. The results indicate that sectioning tissues with external vibrations does affect the amount of total data bandwidth taken up just by the chatter, and in some cases, reduces the relative width of the bandwidth taken up by the chatter.

Description

Citation