Brain Tumor Detection Using Image Processing Techniques

137

In their data analysis study, Sangeeta and Nagendra [65] compared the

accuracy and processing times of k-means and Fuzzy C-means clustering tech-

niques in brain tumor detection. In terms of accuracy both of them obtained

the same results and achieved 80% accuracy. However, in terms of processing

time the situation is different. Since Fuzzy C-means requires more segmenta-

tion than k-means, it has longer processing time. Thus, the k-means clustering

stands out with its performance in this study.

Nyo et al. [66] employed Otsu’s thresholding technique to develop an au-

tomated system for segmenting tumors in brain MRI scans. The initial step

involves converting the RGB image to grayscale and resizing it. Next, the

noise removal stage utilizes a median filter to eliminate any unwanted noise.

The segmentation stage employs Otsu’s thresholding method to distinguish

between tumor and non-tumor regions. Finally, a morphological operation is

applied to accurately identify the tumor regions. The system was evaluated

using the 2015 BRATS dataset and achieved an accuracy of 68.7%.

4.5

Conclusion

The growth of brain cells that is an abnormal and uncontrolled manner is

referred to as a brain tumor. This condition can occur in different anatomical

regions within the brain. MRI images are highly effective in detecting even

the smallest irregularities in the body, and they are used to visualize the brain

and identify any abnormalities, particularly tumors. The research on tumor

imaging aims to accurately determine the location, size, shape and type of the

tumor.

The identification of brain tumors is a crucial task in the field of medical

image processing. The segmentation of medical images is a complex and chal-

lenging stage in the study of tumor detection, and various approaches have

been proposed in this area. The use of brain tumor segmentation techniques

has already demonstrated promise in the detection and analysis of tumors

in clinical images. In order to enhance the accuracy of segmentation, it can

be considered to employ a combination of segmentation techniques or make

modifications in future studies.

Bibliography

[1] R. K. Garg and A. Kulshreshtha, “A review of automated mri image

processing techniques employing segmentation & classification,” Inter-

national Journal of Computer Science Trends and Technology (IJCST),

vol. 5, no. 2, pp. 117–120, 2017.