MRI,
agreement of volunteers and patients,
major parameters and evaluation
MRI was performed on a 1.5T scanner with eight-channel torso phased array coil.
All volunteers approved MRCP scan and agreed to drink 100cc of Manganese Chloride Tetrahydrate (MCT) to reduce signal intensity of GI tract overlying biliary and pancreatic images after our explanation of this purpose.
Under approved our local institutional board,
all patients also agreed with 3D-Cube MRCP scan in addition to the routine MRCP scan including usage MCT and presented written informed consent for this protocol.
Major parameters were FOV=360mm,
Matrix=320x192,
Slice thickness=1.6mm,
slab thickness≧50,
NEX=1,
Bandwidth=62.5kHz,
Fat saturation=chemical shift selective(CHESS).
After obtained source image of MRCP was presented with maximum intensity projection images (MIP)and rotating views,
one radiographer and one radiologist evaluated image quality with agreement.
Phantom
To evaluate appropriate parameters of MRCP on 3D-Cube,
phantom was applied.
For phantom,
isotonic sodium chloride solution as a model of the bile and pancreatic juice and MCT diluted solution(0.1mM) as a model of the abdominal parenchyma organs were used.
After obtained images on 3D-Cube,
signal intensity was measured by placing circular regions of interest (ROIs) on a and b,
contrast ratio was evaluated with the formula:
contrast ratio=(SIa-SIb) / (SIa+SIb)
SIa indicated signal intensity of a,
and also SIb indicated signal intensity of b(Fig.3).
Each parameter on 3D-Cube and 3D-FRFSE was as follows:
RFA(100。,110。120。),
TE (300ms,400ms,450ms,500ms) and ETL (60,80,100,120) on 3D-Cube and TE (300ms,
400ms,
500ms) on 3D-FRFSE was performed.
Other parameters were listed in Table 1(Table 1).
Evaluating appropriate parameters of 3D-Cube,
phantom data revealed that RFA seldom affected contrast ratio (Fig.4).
3D-Cube revealed ETL seldom affected contrast ratio on phantom data
(Fig.5).
Various TEs of phantom data revealed that the longer TE increased the contrast ratio.
Basted on this phantom data,
3D-Cube improved contrast ratio compared with 3D-FRFSE.
(Fig.6)
Healthy volunteers
3D-Cube and 3D-FRFSE was performed under respiratory trigger among two healthy volunteers.
To evaluate parameter,
3D-Cube was scanned with RFA(100。,
110。,
120。) and various TE (300ms,
400ms,
450ms,
500ms).
3D-FRFSE was scanned with TE (450ms,
500ms) in order to compare with 3D-Cube images.
As a result of phantom evaluation that ETL did not almost decrease contrast ratio,
ETL was selected the maximum range of 60 and 80 depended on the respiratory rate changes (Table 2).
After these various TE and RFA was scanned for each volunteer,
contrast ratio between signal intensity of the GB,
CBD and pancreatic duct and signal intensity of background on 3D-Cube and 3D-FRFSE was evaluated quantitively with this formula:
Contrast ratio =(SIeach organ – SIbackgroudn) / (SIeach organ + SIbackground)
Each signal intensity was measured by placing ROIs on the center of the GB and CBD,
and on the enlarged pancreatic duct to eliminate the signal intensity of background.
From each ROI,
the mean value was calculated for evaluation (Fig.7).
As the result of the phantom data,
3D-Cube parameters applied to volunteers revealed that RFA seldom affected contrast ratio (Fig.8).
As the result of the phantom data,
various TEs among volunteers revealed that the longer TE increased contrast ratio of various organs such as the GB,
CBD and pancreatic duct.
3D-Cube images improved contrast ratio compared with 3D-FRFSE in each organ.
Especially the contrast ratio of the pancreatic duct was improved.
With using minimum TE 450ms,
contrast ratio of the pancreatic duct was 0.532 on 3D-Cube and 0.320 on 3D-FRFSE (Fig.9).
Compared with 3D-Cube images using various TE (300ms,
400ms,
500ms),
it was revealed that longer TE improved the appearance of the pancreatic duct due to suppressed the signal intensity of the background tissue (Fig.10).
Patients
MRCP was performed for 34 patients with an age ranged of 29-89 years (median 70.5 years) from September 2017 to December 2018 who approved the MRI examinations.
They were made diagnosis with gall stone in six,
GB polyp in two,
adenomyomataosis of GB in four,
CBD dilatation in three,
pancreatic duct dilatation in one,
IPMN in seven,
nonspecific pancreatic cyst in eight and normal appearance in five as listed (Table3).
MRI parameters
We determined the appropriate parameters for 3D-Cube image based on phantom and volunteers data that indicated longer TE acquired higher contrast ratio.
Therefore,
we set “MSK-PD” for Cube enhance to achieve longer TE.
To obtain the shortest acquisition time,
TE 400ms was selected for all patients,
because TE400ms was confirmed among volunteers.
Setting “MSK-PD” for Cube enhance was able to use variable RFA,
however,
FA was selected 100。 to obtain the shortest acquisition time because RFA did not almost reflect contrast ratio.
ETL was selected the maximum depending on the respiratory rate changes
as same as volunteers scan.
Other parameters were listed in Table 4 (Table4).
Case 1 of 79-year-old female with pancreas IPMN.
Compared with 3D-FRFSE,
3D-Cube was able to diminish the signal intensity of background tissue,
so that contrast ratio was improved between the IHBD and pancreatic duct (Fig.11).
3D-Cube was able to reduce blurring effects,
to obtain T2 long region clearly and to demonstrate the precise images of the IHBD and CBD.
Continuity was also revealed between the pancreatic duct and tiny pancreatic cyst on 3D-Cube.
(Fig.12)
Case 2 of 63-year-old male was diagnosed as adenomyomatosis of GB on screening MRCP.
3D-Cube revealed multiple hepatic cysts and the pancreatic duct due to diminish the signal intensity of surrounding tissue of the hepatic and splenic vein (Fig.13).
As our result of quantitative data analysis,
contrast ratio of various organs including the GB,
IHBD,
CBD and pancreatic duct was shown in Table 5 to compare contrast ratio between 3D-FRFSE and 3D-Cube for all 34 cases of MRCP performed.
Contrast ratio of 3D-Cube was higher than 3D-FRFSE in each organ:
Contrast ratio of the GB on 3D-Cube 0.815±0.059 was higher than 0.787±0.074 on 3D-FRFSE (P =0.001),
contrast ratio of the CBD on 3D-Cube 0.808±0.045 was higher than 0.768±0.050 on 3D-FRFSE (P <0.001).
Contrast ratio of the IHBD on 3D-Cube 0.775±0.094 was higher than 0.695±0.090 on 3D-FRFSE (P <0.001).
Contrast ratio of the pancreatic duct 0.636±0.150 on 3D-Cube was much higher than 0.534±0.161 on 3D-FRFSE (P <0.001).
P value of <0.05 was considered significant statistically between 3D-FRFSE and 3D-Cube using paired t test (Table 5).
Contrast ratio of various organs with comparison between 3D-Cube and 3D-FRFSE was shown graphically.
The increasing rate was 3.5% of the GB,
5.3% of the CBD,
11.5% of the IHBD and 19.1% of the pancreatic duct between 3D-Cube and 3D-FRFSE respectively (Fig14).
For the reason of increased contrast ratio of the IHBD and pancreatic duct,
the center of k-space is sampled with centric order on 3D-Cube “MSK-PD”,
so that employing high refocusing FA until the center of k-space is acquired.
FAs prior to sampling the center of k-space affect the effective TE,
therefore,
this approach is possible to select long TE and to emphasize T2 contrast [18].
We tried to apply for other type of Heavily T2WI such as MR urography,
MRU.
On 3D-Cube,
the left ureter was visible clearly due to suppressed signal intensity around it and also the left ureterovesical junction was able to be confirmed.
Bilateral renal pelvis was not seen on 3D-FRFSE,
however it became visible clearly on 3D-Cube (Fig.15).
On 3D-Cube,
bilateral renal pelvis was demonstrated because overlying signal intensity of duodenum was diminished (Fig.16).