Movement of abdominal organs due to breathing on MRI examination on the abdomen causes motion artifacts that may reduce image quality. Abdominal MRI protocol to reduce respiratory motion artifacts is by using respiratory gating, howeverit results in longer time scan. PROPELLER is a technique for filling data matrices radially with a blade coverage value that is able to reduce artifacts.

To study the application of PROPELLER on abdominal MRI to reduce motion artifacts, get optimal anatomical and image quality information with faster time scan.

This was an experimental study with one group post test design. The samples in the study were 8 respondents, each of which received 8 treatments. One treatment used respiratory gating as a standard protocol and 7 treatmentsapplied PROPELLER using coverage blade variations of 50%, 75%, 100%, 125%, 150%, 175% and 200%. Assessment included time scan, anatomical information, artifacts, SNR, and CNR. Statistical data analysis usedKruskal Wallis test, Wilcoxon test and Spearman test.

There were differences in anatomical information, CNR, SNR, artifacts and time scan between the use of abdominal MRI image and respiratory gating compared to PROPELLER on 100% blade coverage.

The application of PROPELLER with a 100% blade coverage value was able to reduce motion artifact so as to produce optimal anatomical information and image quality with faster time scan.

Keywords: Abdominal MRI, Motion Artifact, PROPELLER, Respiratory Gating.

The selection of MRI examination parameters can be related to the emergence of artifacts that affect image quality and diagnostic information (Ruan, 2001) [28]. In addition, an MRI operator must pay attention to three basic principles in producing an MRI image, namely good image quality, pathological information that is capable to be displayed and comprehensive anatomical information on the area being evaluated (Brown and Samelka, 2003) [7]. Optimal image quality helps accuracy in diagnosing, so it can avoid errors in diagnosis (Bourne, 2010). On abdominal MRI examination the organs always move due to the patient's breathing and it becomes a problem since it causes motion artifacts that can significantly reduce image quality (Low et al., 1997).

There are two techniques commonly used in the abdominal MRI examination protocol to eliminate artifacts due to breath movements, namely breath-hold and respiratory gating techniques (Moeller et al., 2003). The breath-hold technique is performed by the patient holds his breath during the scanning process. Good breath holding performed by patients will result in good image quality (Grand et al., 2012) [13]. Patients who are not cooperative so that they are unable to hold their breath will produce signals that become motion artifacts in the MRI images (Brown and Samelka, 2003) [7]. Meanwhilem the respiratory gating technique is a procedure for taking an image of an organ at a certain time during the patient's breath cycle. The position and duration of image capture at each breath cycle is determined by monitoring the patient's breath movements (AAPM, 2006). Respiratory gating causes the image capture could not to be carried out continuously so that the use of this technique takes longer, but results in better image quality than breath-hold technique (Kandpal et al., 2009 and AAPM 2006) [16].

In 1990 a new technique was developed in the process of filling the data matrix on MRI imaging. The data filling technique is known as PROPELLER (Periodically Rotated Overlapping Parallel Lines with Enhanced Reconstruction) (Elster, 2015) [11]. PROPELLER is a technique of filling the data matrix (k-space) radially by a number of parallel lines (blade coverage) of data obtained from the scanning process. PROPELLER produces excess data at the center of k-space. The same data on each coverage blade is used to correct other data so that data as a result of movement can be known. The differencesof data from each blade coverage are identified as inconsistencies within data which is interpreted as motion artifacts, so that the data can be reduced. The scaning time needed by using PROPELLER is influenced by blade coverage and Echo Train Length (ETL) values ​​(McRobbie et al., 2006). According tothose problems a study on the optimization of abdominal MRI images will be carried out with the application of PROPELLER.

The objective of this study was to examine the use of PROPELLER on abdominal MRI to reduce motion artifact, obtain optimal anatomical information and image quality with faster time scan. This study was conducted by applying PROPELLER on abdominal MRI examination with variations in blade coverage values ​​of 50%, 75%, 100%, 125%, 150%, 175% and 200%. The images obtained were compared to the results of the examination by using the respiratory gating technique which is the standard protocol of abdominal MRI examination. Analysis was carried out on several parameters including anatomical information, Signal to Noise Ratio (SNR), Contrast to Noise Ratio (CNR), artifact and time scan. This was an experimental study with one group post test design. The sampling procedure was conducted by purposive sampling technique which means that the sampling is done based on the decision of the researcher, which in his opinion appears to represent the population (Budijanto, 2010). This study used eight types of treatments so that the researcherdetermined the number of samples to be used as many as 8 samples.

This study obtained information that changes in blade coverage values ​​provided a significant difference in time scan with p< 0,001, anatomical information with p< 0,001, motion artifacts with p< 0,00, SNR with p = 0,007 and CNR with p = 0,004. Furthermore, the assessment of the use of respiratory gating compared to the application of PROPELLER showed the following results:

Tabl e 1 Comparison of Assessment o n the Use of Respiratory Gating with PROPELLER application

To prove that interrelated variables are then spearman test with the following results:

Table 2 Variable Correlation Test Results

It can be seen in the table that in the anatomical information assessment of 50% blade coverage had a mean value of 40,12 lower than the mean value on the use of respiratory gating of 47,88. Thus, the 50% blade coverage should not be used on Abdominal MRI examinations since it produced lower anatomical information than the standard protocol. Meanwhile in the SNR and CNR assessments, all variations of blade coverage yielded a greater and significantly different value than the use of respiratory gating. Therefore the SNR and CNR assessments did not provide limits on the use of blade coverage. The assessment of the use of 100% blade coverage values artifacts showed the artifact values of 3,25 and 200% blade coverage values artifacts showed the artifact values of 3,00. The use of 100% - 200% blade coverage values is recommended because it produces lower artifacts compared to the use of respiratory gating. The selection of blade coverage that is able to reduce motion artifacts, produce optimal anatomical information and image quality should selected from blade coverage that is able to produce better images than the use of respiratory gating with the shortest time scan.

In the table it can be seen that the 200% blade coverage produced the highest mean of anatomical information, artifacts and anatomical information but it also required significantly higher time scan than the use of respiratory gating with p = 0.011. Meanwhile, the implementation of 125% -175% blade coverage values showed that there was no difference in time scan compared to the use of respiratory gating with a p values of 1.00 (125% blade coverage), 0.910 (150% blade coverage) and 0.778 (blade coverage 175%), respectively. The 100% blade coverage resulted in better anatomical information, artifact and image quality compared to the use of respiratory gating. Besides, the time scan required on a blade coverage of 100% of 3.25 minutes was significantly faster (p=0.035) than the use of respiratory gating of 4.33 minutes. Therefore, the application of PROPELLER with 100% blade coverage was able to reduce motion artifacts, produce the most optimal anatomical information and image quality in abdominal MRI examination with the shortest time scan. The Spearman test results also showed a correlation between variables of time scan, anatomical information, artifacts, SNR and CNR so that the use of 100% blade coverage with scanning time of 3.25 is related to the resulting image with p <0.05 in all tests between variables.

.

a. b.

Figure 1. a. Motion artifacts still present ed in the use of respiratory gating b. Abdom inal MRI image by usin g PROPELLER with 100% blade coverage