Previous   |    Next

doi:10.2204/iodp.proc.314315316.122.2009

X-ray computed tomography

X-ray CT imaging provided real-time constraints for Expedition 315 core logging and sampling strategies. The following explanation of our methods was derived from the “cookbook” prepared by CDEX/​JAMSTEC for onboard use (X-Ray CT Scanner, version 3.00, 31 July 2007) and practical experience. The cookbook is based on GE Healthcare (2006), Mees et al. (2003), and Nakano et al. (2000).

The X-ray CT instrument on the Chikyu is a GE Yokogawa Medical Systems LightSpeed Ultra 16 capable of generating sixteen 0.625 mm thick slice images every 0.6 s, the time for one revolution of the X-ray source around the sample (see Table T1 for operating parameters). As a result of this high scanning rate, 1.5 m sections of core were typically imaged in <3 min. Data generated for each core consist of core-axis-normal planes of X-ray attenuation values with dimensions of 512 × 512 pixels. Data were stored as Digital Imaging and Communication in Medicine (DICOM) formatted files.

Background

The theory behind X-ray CT has been well established through medical research and is only briefly outlined here. X-ray intensity (I) varies as a function of X-ray path length and the linear attenuation coefficient (LAC) of the target material as

I = I₀ × e^–µL,

where

• I = transmitted X-ray intensity,
• I₀ = initial X-ray intensity,
• µ = LAC of the target material, and
• L = X-ray path length through the material.

LAC is a function of the chemical composition and density of the target material. The basic measure of attenuation, or radiodensity, is the CT number given in Hounsfield units (HU) and is defined as

CT number = [(µ_t – µ_w)/µ_w] × 1000,

where

• µ_t = LAC for the target material and
• µ_w = LAC for water.

The distribution of attenuation values mapped to an individual slice comprises the raw data that are used for subsequent image processing. Successive two-dimensional slices yield a representation of attenuation values in three-dimensional (3-D) pixels referred to as voxels. Visualizations are generally accomplished by varying the manner in which the distribution of CT number is assigned to 256 shades of gray or a color pallet. Onboard analysis of CT images was done primarily using the OsiriX 32-bit DICOM viewer (version 2.7.5) running on a G5 Macintosh (OS 10.4).

Analytical standards used during Expedition 315 were air (CT number = –1000), water (CT number = 0), and aluminum (2467 < CT number < 2487) in an acrylic core mock-up. The schedule for running these was fully implemented on 3 December 2007. All three standards were run once daily after air calibration, and once each week the three standards were run both before and after air calibration. For each standard analysis the CT number was determined for a 24.85 mm² area at fixed coordinates near the center of the cylinder (Table T2).

X-ray CT scan data usage

X-ray CT scans were used routinely during Expedition 315 to

• Provide an assessment of core and core-liner integrity,
• Determine the appropriateness of whole-round samples,
• Identify potentially subtle features and/or features that might warrant special handling during visual core description and sampling, and
• Measure orientations and identify crosscutting relationships.

CT scans provided first-hand information regarding the integrity of sample cores and core liners. Cores that experienced gas expulsion during depressurization could be readily identified (Fig. F1), as could cores in which extensive drilling-induced brecciation had occurred (Fig. F2). In several instances, core liners buckled into the sample along the core axis, displacing the sample, and in one case a liner partially delaminated and became entangled in a serpentine manner within the sample (Movie M1). CT scans also provided a fast and practical means of assessing the coherence of core intervals that contained abundant natural structures and were therefore targeted for additional paleomagnetic analyses.

All whole-round core sections were screened to avoid destructive testing on core samples that might contain critical structural features. This also ensured minimal drilling disturbance of whole-round samples (essential for microbiology as well as physical and mechanical properties) and an assessment of whole-round homogeneity. X-ray CT scanning was done immediately after core cutting so that time-sensitive whole-round samples (e.g., those for interstitial water and microbiological analyses) could be included in this screening process.

The value of X-ray CT scans to identify sedimentary and tectonic features prior to visual core description was quickly recognized (Movie M2). Three-dimensional structure orientation could easily be determined from CT sections, whereas performing the same measurement on the cores generally required cutting orthogonal faces. Furthermore, structures could be classified as a function of their CT contrast, which is presumably related to porosity changes or chemical alteration within shear zones. Observations were noted in the structural geology CT log sheets (Fig. F3) and are commonly referred to in the visual core description sheets.

X-ray CT images have multiple uses, from early assessment of cores to description and synthesis. For this reason, several hundred gigabytes of CT data were durably stored on a local database at the OsiriX interpretation station. All X-ray CT data will be available after the moratorium at the SI07 Web site (sio7.jamstec.go.jp).

Top of page   |    Previous   |    Next