Patient population
The study protocol was approved by the local Ethics Committee (11/11/2010),
and informed consent was obtained from all relatives of the deceased and the living control subjects.
Twenty-one patients,
all of whom died of natural causes (e.g.
cardiovascular arrest),
were examined (11 male,
5 female; mean age 70.2 ± 9.6 years; range: 61 – 85 years,
mean body weight 69.2 ± 16.6 kg).
Exclusion criteria were non-natural dead,
a history of chronic or systemic disease affecting the liver parenchyma (e.g.
hepatitis,
liver tumors or metastasis,
liver cirrhosis),
cardiac pacemaker,
death interval > 6 hours until MRI scan.
The corpses were stored at room temperature until scanning.
Initial MRI scans were conducted within a median time interval of 5 hours postmortem (range 2 to 13 hours) an proceeded in intervals of two hours within the first 16 hours after death.
Core body temperature was rectally measured prior to each MRI examination.
A control group of 5 healthy subjects (3 male,
2 female; mean age 27.3 ± 6.3 years) was scanned for in vivo ADC values.
MR Imaging Protocol
All examinations were performed on a 1.5 Tesla MRI scanner (Magnetom Symphony; Siemens Medical Solutions,
Erlangen,
Germany) using a surface body-coil.
For the morphological orientation a localizer and a half Fourier-acquired single shot turbo spin echo (HASTE) -sequence (TR 1840 ms,
TE 125 ms,
FoV 350 mm,
slice thickness 0.7 cm) in transverse,
coronal and sagittal orientation was performed.
Diffusion-Weighted Imaging
A diffusion-weighted spin echo echo-planar sequence (TR 3000 ms; TE 74 ms; FOV 262 x 350 mm; matrix 115 x 192 mm; slice thickness 6 mm; gap 1.5 mm; voxel size 0.9x0.9x5 mm3; 20 slices,
and NEX 10) was generated in transverse orientation.
A total of 2 averages were acquired and scan time for DWI was about 4 to 6 minutes.
The examination time totaled 30 minutes including time for patient positioning and localization of the liver structures.
The ADC is given by the following equation as shown:
S (b1) = Si (b0)exp(-b1 × ADC)
where S(b1) is the signal intensity of the image measured with a gradient pulse b1 = 50,
400,
and 800 s/mm2 while S(b0) estimates the signal intensity for a b-value of 0 s/mm2. The DWI was performed with a trace weighted sequence type (3 orthogonal directions).
According to this equation,
pixelwise ADC-maps were generated as grey values using Siemens based software (Syngo MR A30A,
Siemens Medical Solutions,
Erlangen,
Germany).
Temperature correction
As water diffusion is dependent on temperature,
the individual diffusion metrics obtained in postmortem cases should be normalized to guarantee comparability within the study group and to in vivo controls,
which has been already applied in previous studies.
The temperature correction was performed according to the algorithm of Kozak et al [8] describing the relationship between temperature and diffusion in unrestricted water.
(i)
T= 2256.74Kln4392.21x10-3mm2/sDmm2/s-273.15K (Figure 1)
Where T is temperature in units of Celsius,
D ist he diffusion in mm2/s.
(ii)
D= 4392.21x10-3exp2256.74T+273.15 (Figure2)
The temperature correction was individually applied for the measured core temperature prior to each scan in the postmortem group.
The body core temperature of in vivo cases was assumed to be 37°,
giving a correction factor (CF) of:
(iii)
CF= 0.0030postmortem core temperature of in vivo cases was assumed to be 37°,
giving a correction factor (CF) as stated in figure 3