Brute force absorption contrast microtomography

Abstract

In laboratory X-ray microtomography (XMT) systems, the signal-to-noise ratio (SNR) is typically determined by the X-ray exposure due to the low flux associated with microfocus X-ray tubes. As the exposure time is increased, the SNR improves up to a point where other sources of variability dominate, such as differences in the sensitivities of adjacent X-ray detector elements. Linear time-delay integration (TDI) readout averages out detector sensitivities on the critical horizontal direction and equiangular TDI also averages out the X-ray field. This allows the SNR to be increased further with increasing exposure. This has been used in dentistry to great effect, allowing subtle variations in dentine mineralisation to be visualised in 3 dimensions. It has also been used to detect ink in ancient parchments that are too damaged to physically unroll. If sufficient contrast between the ink and parchment exists, it is possible to virtually unroll the tomographic image of the scroll in order that the text can be read. Following on from this work, a feasibility test was carried out to determine if it might be possible to recover images from decaying film reels. A successful attempt was made to re-create a short film sequence from a rolled length of 16mm film using XMT. However, the *brute force* method of scaling this up to allow an entire film reel to be imaged presents a significant challenge.

Publication
Proceedings of SPIE - The International Society for Optical Engineering ยท September 2014
Date
Links
PDF