ITRGCNN: Integrated Threshold & Region Segmented CNN for Tumour Cells Detection in Brain

  • B. Hanumantha Rao, R. Sivarama Prasad


Continuous monitoring of brain tumour growth and response to treatment is crucial for disease management. Accurate detection facilitates longitudinal analysis, helping medical professionals assess treatment efficacy and make necessary adjustments. Brain tumour detection using image processing involves analyzing and processing brain MRI or CT scans to identify and delineate tumour regions. Segmentation algorithms may produce regions that are too fragmented (over-segmentation) or fail to capture the complete extent of the target object (under-segmentation) [9]. These issues can arise due to the choice of segmentation parameters, the complexity of the object's boundaries, or the presence of noise in the image. The proposed model integrates segmentation techniques based on threshold and region. Integrated segmentation offers a powerful and flexible approach to address the challenges and limitations of individual segmentation techniques. The integration of thresholding and region growing algorithms can be a powerful approach for image segmentation tasks. These two techniques complement each other, with thresholding providing a simple yet effective initial segmentation, and region growing refining the segmentation by incorporating local information and spatial coherence. This integrated segmentation will provide the model with initial segmented regions based on intensity or other image features. Once model has the segmented regions, the model needs to represent them in a format suitable for CNN classification.

How to Cite
R. Sivarama Prasad, B. H. R. (2022). ITRGCNN: Integrated Threshold & Region Segmented CNN for Tumour Cells Detection in Brain . Design Engineering, (1), 4786 - 4800. Retrieved from