Application of Fusion Technique with ImageJ Stacks Feature for Brain Tumor MRI Image Optimization

ABSTRACT


Introduction
Brain tumors are abnormal growths of cells in brain tissue, which need to be examined with an MRI of the brain.Symptoms include severe headaches, behavioral changes, visual or hearing impairments, difficulty speaking, seizures, or movement coordination problems.An MRI Brain examination provides a detailed picture of a brain create projections from stacks of images with different types of projections such as maximum intensity, minimum intensity and average intensity.
In general, app integration Fusion in the MRI modality has become a built-in part in some software.However, software installation Fusion in the MRI modality device requires a significant cost expenditure, reaching around 150 million rupiah.Unfortunately, not all hospitals have enough resources to adopt the software (Nadra et al., 2022).Because of this, there is a need for an alternative to be able to do the Fusion if the MRI modality is not equipped with software Fusion at the same time can specify the type of projection Fusion which is optimal in evaluating brain tumor images.

Research Methods
This study used 17 image samples with a quasi-experimental design of post test only without control group to compare three analysis methods, namely fusion maximum intensity, minimum intensity, and average intensity.Each image sample is analyzed using these three methods, so that each method has 17 analysis results from the same image sample so that the most suitable projection can be determined.The evaluation of image quality was carried out through a histogram which was then analyzed with a crucal-wallis and the Mann Whitney u test, while the analysis of pathological information used a crucalwallis, followed by a post hoc test and continued with Mann Whitney u for further analysis.

Result
The image processing process begins by burning MRI images stored in the local database in dicom format.The research sample was MRI images of Axial T2-FLAIR and Axial T1-GD (T1-weighted post-contrast) brain tumors which had a field of view (FOV) = 220 mm and slice thickness = 5 mm.MRI images of Axial T2-FLAIR and Axial T1-GD sequence brain tumors can be seen in figure 1 as follows:  Stacks Feature The flow of the fusion process with the Stacks feature in the ImageJ software above begins by opening the ImageJ application, then opening the two images, namely the MRI image of the Axial T1-GD (T1-weighted post-contrast) and Axial T2-FLAIR sequence brain tumors, then the downscale image is carried out as a pixel size demonstration with the images to stack method which is in the Stacks feature {(copy(center)}, if the pixel size is different between the two images to be fused.The next step is to do various methods found in the Stacks feature, namely Stack to Images, the implementation can be done by selecting "Image" > "Stacks" > "Stack to Images", used to convert the image stack into separate images.
Once in a single separate image, the images to stack method can be performed by selecting "Image" > "Stacks" > "Images to Stack", used to combine the two images that have been generated previously.Then the last method is carried out by the "Z Project" method by selecting "Image" > "Stacks" > "Z Project", used to make projections on the image that has been stacked into the form of maximum intensity, minimum intensity and average intensity images, After the fusion image display appears, the last step is to save the fusion image results which can be done with various storage formats.The display of fusion image results on MRI brain tumors with the Stacks feature on ImageJ with maximum intensity, minimum intensity and average intensity projections can be seen in figure 3 as follows:

Image Quality
The quality of the image in this study was measured from the gray level value in each slice of the entire MRI fusion image of brain tumors produced from the Stacks feature on ImageJ fusion results with projection type maximum intensity, fusion results with projection type minimum intensity and fusion results with projection type average intensity.The gray level value is obtained from the display of the value displayed on the histogram using ImageJ software.
The histogram results of MRI fusion images of brain tumors produced from the Stacks feature on ImageJ are fusion results with projection type maximum intensity, fusion results with projection type minimum intensity and fusion results with projection type average intensity can be seen in figure 4 as follows: The information obtained from the display of the MRI image of the brain tumor above is the mean value of the density of the brain tumor image in fusion with the maximum intensity being at the highest value of 163.37 which means that the intensity range provides the maximum contrast between the tumor and the surrounding normal tissue.The mean value of fusion with a minimum intensity is 114.52 which means that the intensity range highlights the subtumor details well.And the mean value of fusion with an average intensity is 139.19 which means that it reflects a balanced combination of contrast between the tumor and normal tissue.The histogram also displays the standard deviation value, the lowest density value and the highest density.The size of the image matrix and the range of density (gray level) from 0-2301 that explain the configuration of the dicom image bit from the MRI image of the brain tumor.
The results of the overall image quality assessment of MRI fusion of brain tumors are shown from the average (mean) gray level value which can be seen in table 1  There was an average gray level of MRI fusion images of brain tumors with a projection type maximum intensity of 185.12, for the average gray level of MRI brain tumors with a projection type minimum intensity of 114.35 and then the average gray level for MRI of brain tumors with a projection type average intensity of 149.98.
The quality of MRI fusion images of brain tumors with projection type maximum intensity, minimum intensity and average intensity was analyzed using the crucial-wallis test from the mean data of the gray level value.The test results are seen in table 2 as follows: The test results showed that there was a significant difference in the quality of MRI fusion images of brain tumors with projection type maximum intensity, minimum intensity and average intensity because it had a p-value of < 0.001.
Follow-up tests to determine the difference in the quality of MRI fusion images of brain tumors with projection types maximum intensity, minimum intensity and average intensity were carried out by analyzing the Mann Whitney U test.The results of the Mann Whitney test can be seen in table 3 as follows: These results provide information that the difference in the quality of MRI fusion images of brain tumors with projection type maximum intensity and minimum intensity is obtained p-value = < 0.001, which means that there is a significant difference between the quality of MRI fusion images of brain tumors with projection type maximum intensity and minimum intensity.The difference in the quality of MRI fusion images of brain tumors with projection type maximum intensity and average intensity was obtained p-value = < 0.001, which means that there was a significant difference between the quality of MRI fusion images of brain tumors with projection type maximum intensity and average intensity.The difference in the quality of MRI fusion images of brain tumors with projection type manimum intensity and average intensity was obtained p-value = < 0.001, which means that there was a significant difference between the quality of MRI fusion images of brain tumors with projection type minimum intensity and average intensity.

Pathology Information
Pathology information is assessed by a radiologist on the overall slice of the brain tumor imagery by looking at the clarity of the structure and boundaries of the edges of the brain tumor image, whether it is bright and firm so that it clearly visualizes the brain tumor and the lesion tissue around the tumor.The assessment is divided into 3 categories, namely unclear, quite clear and clear.
The results of the assessment of pathological information from two observers obtained pathological information from the results of the fusion of MRI images of brain tumors with projection types maximum intensity, minimum intensity and average intensity seen in the appearance of the structure and boundaries of the edges of the brain tumor image, whether it is bright and firm so that it clearly visualizes the brain tumor and the lesion tissue around the tumor.Before data analysis, the assessment of pathological information was carried out by two observers and then a kappa test was carried out to obtain a close relationship or harmony from the two observers.The results of the alignment test between observers can be seen in table 4 as follows: The results of the alignment test between observers with the kappa method obtained excellent agreement between the two observers in providing an assessment of pathological information because it has a kappa value of 0.76 with a p-value = <0.001.
The results of the assessment of pathological information as a whole, MRI fusion images of brain tumors for each variable can be seen in table 5 as follows: Pathology information on MRI fusion image of brain tumors with projection type maximum intensity had a total score of 95 with a percentage (93.14%).The assessment of pathological information of MRI fusion image of brain tumors with a minimum intensity projection type had a total score of 40 with a percentage (39.22%).The assessment of pathological information of MRI fusion image of brain tumors with projection type average intensity had a total score of 69 with a percentage (67.65%).
Information on MRI fusion pathology of brain tumors with projection type maximum intensity, minimum intensity and average intensity was analyzed using the crucial-wallis test and the results obtained can be seen in table 6 as follows:

Table 6. Analysis of the Results of the Difference Test of Pathological Information of Brain Tumor MRI fusion with projection types Maximum Intensity, Minimum Intensity and Average Intensity
The results of the analysis in the Krukal-Wallis test obtained the lowest pathological information in the fusion MRI of brain tumors with a projection type minimum intensity with a mean rank value of 23.50 and the highest pathological information in the fusion The results of the pathological information analysis found that the difference in the pathological information of MRI fusion of brain tumors with the projection type maximum intensity and minimum intensity was obtained p-value = < 0.001 which means that there is a significant difference between the pathological information of MRI fusion of brain tumors and the projection type maximum intensity and minimum intensity.The difference in brain tumor MRI fusion pathology information with projection type maximum intensity and average intensity was obtained p-value = < 0.001 which means that there was a significant difference between brain tumor MRI fusion pathology information with projection type maximum intensity and average intensity.The difference in brain tumor MRI fusion pathology information with projection type manimum intensity and average intensity was obtained p-value = < 0.001, which means that there is a significant difference between brain tumor MRI fusion pathology information with projection type minimum intensity and average intensity.

Discussion Image Quality
Assessment of MRI image quality of brain tumors based on differences in grayness (Gray Level).Average Gray Level for imagery Fusion MRI with a maximum projection of intensity was 185.12, higher than the minimum projection of intensity (114.35) and average intensity (149.98).Optimal image quality on 16-bit imagery is determined from the value of the Gray Level distributed in the range of 0 to 65,535 shades of gray (Utama  , 2020).Fusion of MRI imagery is important for combining information from different types of imaging, such as conventional MRI, MRI with contrast, and PET/CT, to produce a more comprehensive picture of the tumor and its structure.
Valuation Gray Level allowing in-depth analysis and better understanding of the characteristics and structure of MRI brain tumors (Widhiarso et al., 2018).The selection of the right projection according to the evaluation objectives is helpful in the diagnosis, treatment planning, and monitoring of tumor progression.Difference in mean value Gray Level of the three types of projection reflect different approaches to highlighting pixel intensity.Maximum Intensity Projection (MIP) displays the highest value of each voxel, emphasizing high-intensity areas that are important for diagnosis.Minimum Intensity Projection (MinIP) displays the lowest value, highlighting areas with low intensity, while Average Intensity Projection provides a balanced picture of intensity.The maximum intensity projection is optimal because it focuses on the areas with the highest intensity, allowing for clear identification of areas with high tumor activity, which is important for accurate diagnosis and treatment planning.The results of the Kruskal-Wallis test showed a significant difference between the three projections with a p-value < 0.001.Further analysis using the Mann-Whitney U test showed a significant difference between all projection pairs with a p-value < 0.001.
The use of fusion imagery can combine the advantages of each imaging technique, improve the ability to determine the location, size, biological properties, and response to brain tumor treatment, as well as assist in precise treatment planning and monitoring of tumor progression.The results of the analysis showed that the maximum intensity projection was more optimal than other projections because it provided maximum contrast between the tumor and normal tissue, with a high and sharp histogram peak, clearly separating the tumor from the normal tissue.
Statistical tests show significant differences between these three types of projection, with maximum projection intensity producing the best image quality, essential for accurate diagnosis and proper treatment planning.The fusion method using ImageJ software, specifically the Stacks feature, allows merging multiple images into a single stack that makes image analysis easier.Fusion of radiological imagery provides a more comprehensive picture, making it easier to identify and locate tumors, improving the efficiency and accuracy of diagnosis and treatment planning.Therefore, the maximum projection intensity is assessed as the optimal fusion method for MRI brain tumors.This is in accordance with the results of the study which explains that Maximum Intensity Projection (MIP) will be able to improve image quality such as contrast and image sharpness.Thus, these projections will provide a clearer and more detailed picture of areas with high tumor activity so that it can ultimately improve the accuracy of diagnosis, therapy planning and effective patient monitoring (Naeem et al., 2022).

1) Pathology Information
Assessments from two observer radiologists show that the image Fusion MRI of brain tumors with a projection of maximum intensity is more optimal than the other two projections.This projection obtained a score of 95 with a percentage of 93.14%, while the minimum intensity projection and average intensity obtained scores of 40 (39.22%) and 69 (67.65%), respectively.Observer stated that the maximum intensity projection displayed the structure and periphery of brain tumors clearly and firmly, visualizing the tumor and surrounding lesions well.MRI is used to visualize brain tumors accurately, but without techniques Fusion, radiologists have to rely on individual MRI sequences that take longer to interpret (Nadra et al., 2022).Fusion MRI imagery provides a Nur Wahyu Tajuddin, Bambang Satoto, Rini Indrati, Donny Kristanto Mulyantoro, Darmini, Gatot Murti Wibowo comprehensive overview of tumor structure and characteristics, aiding in diagnosis, treatment planning, and patient monitoring (Özbay & Özbay, 2023).
Maximum Intensity Projection (MIP) highlights the areas with the highest intensity, it is important to identify active areas such as vascular focuses.MinIP highlights the lowest intensity that may indicate necrosis, while the Average Intensity Projection provides a balanced picture of the intensity distribution.The cruskal-wallis analysis showed significant differences in pathological information between these three types of projections.The maximum intensity projection has the highest mean rank (78.50), while the lowest minimum intensity (23.50), and the average intensity between the two (52.50).The mann-whitney u test showed a significant difference between these three projections with a p-value < 0.001.
This difference occurs because each projection approach highlights a different aspect of pixel intensity.Maximum Intensity Projection (MIP) focuses on the area with the highest intensity, which is useful for highlighting highly active parts of the tumor.Minimum Intensity Projection (MinIP) highlights the areas with the lowest intensity, helping to identify areas that are dormant or degenerated.The Average Intensity Projection provides a balanced picture of the structure of the tumor.Based on the results of the study, MIP is a more optimal type of fusion projection because it is able to highlight areas with high tumor activity, providing a clear picture and details that are very important for diagnosis and treatment planning.
An advantage of maximum intensity projection that the other two fusion types do not have is its ability to highlight areas of the highest intensity, providing clear details about tumor activity which is important for diagnosis and treatment planning.This projection helps radiologists identify the location, size and characteristics of tumors more accurately.Thus, the diagnosis and treatment planning of patients become better.These projections also provide a highly accurate and detailed picture of the pathology, which in turn increases the effectiveness of diagnosis and therapy.
The results of this study are supported by the opinion of other researchers who show that the use of Maximum Intensity Projection (MIP) technique in medical images can improve the contrast and sharpness of images as well as pathological analysis.This confirms that the use of MIP in Fusion MRI is the optimal choice to improve precision in the diagnosis and treatment planning of brain tumor patients (Naeem et al., 2022).

Conclusion
The role of law in the formation of social identity in urban communities is complex and multidimensional.Law serves as a framework that regulates social interaction and maintains public order.Effective law enforcement plays an important role in creating a sense of justice and equality among city dwellers, helping to strengthen the social identity of individuals and groups by providing them with a sense of security and protection guaranteed by the law.In addition, the law also has an impact in regulating the way individuals and groups interact with each other, shaping social norms that are integral to the social identity of urban communities.

Figure 1 .
Figure 1.MRI images of brain tumor sequences (a) Axial T1-GD (T1-weighted post-contrast) and (b) Axial T2-FLAIR sequences MRI images of brain tumors of the Axial T1-GD sequence (T1-weighted postcontrast) and Axial T2-FLAIR sequence were carried out by the Stacks feature in the ImageJ software to be applied to fusion techniques with projection type maximum intensity, minimum intensity and average intensity.The process of processing fusion images with the Stacks feature in ImageJ software can be seen in figure 2 as follows:

Figure 2 .
Figure 2. Process Flow of the ImageJ Fusion technique on MRI Brain Tumor withStacks Feature The flow of the fusion process with the Stacks feature in the ImageJ software above begins by opening the ImageJ application, then opening the two images, namely the MRI image of the Axial T1-GD (T1-weighted post-contrast) and Axial T2-FLAIR sequence brain tumors, then the downscale image is carried out as a pixel size demonstration with the images to stack method which is in the Stacks feature {(copy(center)}, if the pixel size is different between the two images to be fused.The next step is to do various methods found in the Stacks feature, namely Stack to Images, the implementation can be done by selecting "Image" > "Stacks" > "Stack to Images", used to convert the image stack into separate images.Once in a single separate image, the images to stack method can be performed by selecting "Image" > "Stacks" > "Images to Stack", used to combine the two images that have been generated previously.Then the last method is carried out by the "Z Project" method by selecting "Image" > "Stacks" > "Z Project", used to make projections on the image that has been stacked into the form of maximum intensity, minimum intensity and average intensity images, After the fusion image display appears, the last step is to save the fusion image results which can be done with various storage formats.The display of fusion image results on MRI brain tumors with the Stacks feature on ImageJ with maximum intensity, minimum intensity and average intensity projections can be seen in figure3as follows:

Figure 3 .
Figure 3. Fusion MRI Brain Tumor Projections Maximum Intensity with Features of Stacks (a) Manimum Intensity (b) and Average Intensity (c) Fusion images that have been obtained previously, then assessments are carried out from the aspects of image quality and pathological information.Image QualityThe quality of the image in this study was measured from the gray level value in each slice of the entire MRI fusion image of brain tumors produced from the Stacks feature on ImageJ fusion results with projection type maximum intensity, fusion results with projection type minimum intensity and fusion results with projection type average intensity.The gray level value is obtained from the display of the value displayed on the histogram using ImageJ software.The histogram results of MRI fusion images of brain tumors produced from the Stacks feature on ImageJ are fusion results with projection type maximum intensity, fusion results with projection type minimum intensity and fusion results with projection type average intensity can be seen in figure4as follows:

Figure 4 .
Figure 4. Histogram of Brain Tumor MRI Fusion Results Projecting Maximum Intensity with Stacks Features (a) Minimum Intensity (b) and Average Intensity (c)

Table 1 . Average Gray Level Value of MRI fusion image of Brain Tumor with projection type Maximum Intensity, Minimum Intensity and Average Intensity
as follows:

Table 2 . Analysis of the Results of the Quality Difference Test of MRI Fusion Images of Brain Tumors with projection types Maximum Intensity, Minimum Intensity and Average Intensity Variable
p-value Fusion MRI Brain Tumor with Maximum Intensity < 0.001 Fusion MRI Brain Tumor with Manimum Intensity Fusion MRI Brain Tumor with Average Intensity

Table 5 . Assessment of Pathological Information of MRI Fusion Images of Brain Tumors with projection types Maximum Intensity, Minimum Intensity and Average Intensity
N Amount/T.Score Percentage (%)

Table 7 . Analysis of the Results of the Pathological Information Difference Test between MRI Fusion Variables of Brain Tumor with Maximum Intensity, Minimum Intensity and Average Intensity projection types
projection type maximum intensity with a mean rank value of 78.50.Meanwhile, fusion MRI brain tumors with projection type average intensity have pathological information between fusion MRI brain tumors with projection type maximum intensity and fusion MRI brain tumors with projection type minimum intensity, namely a mean rank value of 52.50.The results of this crucial test-wallis test also obtained a p-value = < 0.001.This means that there is a significant difference between MRI fusion of brain tumors with projection types maximum intensity, minimum intensity and average intensity.Follow-up tests to determine the difference between the variables of MRI fusion pathology information of brain tumors with projection type maximum intensity, minimum intensity and average intensity were carried out by analyzing the Mann Whitney U test.The results of the mann whitney u test can be seen in table 7 as follows: