An Efficient Brain Tumour Extraction in MR Images using Ford-Fulkerson Algorithm
Journal Title: Transactions on Machine Learning and Artificial Intelligence - Year 2016, Vol 4, Issue 2
Abstract
Brain tumor division intends to discrete the diverse tumor tissues, for example, dynamic cells, necrotic core,and edema from typical cerebrum tissues of White Matter (WM), Gray Matter (GM), and Cerebrospinal Fluid (CSF). MRI-basedbrain tumor division studies are drawing in more consideration lately because of non-invasiveimaging and great delicate tissue difference of Magnetic Resonance Imaging (MRI) pictures. With the improvement of very nearly two decades, the inventive methodologies applying PC supported strategies for the sectioning brain tumor are turning out to be more develop and coming closer to routine clinical applications. The reason for this exploration work is to give a far-reaching review to MRI-based brain tumor division techniques. To consider and characterize the tumor pictures, a couple of well-known Edge Detection Techniques have been proposed as of late. This examination work has distinguished KWT (K-Means, Watershed, and Texture) Segmentation Technique and executed and contemplated. From our test comes about, this examination work uncovered that this model neglects to make productive groups order force causes poor tumor characterization precision. This is one of the real issues to anticipate the tumor design and to address this issue, the Ford-Fulkerson Segmentation Technique is proposed and concentrated altogether as far as Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and Classification Accuracy. Results set up that the proposed Ford-Fulkerson Segmentation Technique beats KWT in terms of Mean Squared Error (MSE), Peak Signal to Noise Ratio (PSNR) and Classification Accuracy.
Authors and Affiliations
C. Latha, K. Perumal
Keywords
Related Articles
- EP ID EP277223
- DOI 10.14738/tmlai.42.1875
- Views 87
- Downloads 0
Inference-Based user’s Recommendation in E-Learning Systems
This paper proposes a technique of user’s recommendation for Elearning systems, which makes it possible to identify the best qualified profiles in a given field, the method is based on artificial intelligence in order to...
Publishing Student Graduation Projects Based on the Semantic Web Technologies
Linked Open Data is already widely available in several industries, such as libraries, biomedicine and Linked government data, etc. for data sharing, promoting semantic web development and maintaining a global culture of...
A Dynamic Clustering Approach for Maximizing Scalability in Wireless Sensor Networks
Scalability is an important and crucial issue which in routing protocols for Wireless Sensor Networks (WSNs). In this paper, we present an approach to achieving a balanced energy consumption rate using dynamic clustering...
Architecture of A Semantic Annotation of Handwritten Documents Based on the Ontology "OMOS"
In this work we propose to present an application that supports the representation of manuscript documents according to an ontological approach. The implementation of this application makes it possible to annotate semant...
Bio-Inspired Temporal-Decoding Network Topologies for the Accurate Recognition of Spike Patterns
In this paper will be presented simple and effective temporal-decoding network topologies, based on a neuron model similar to the classic Leaky Integrate-and-Fire, but including the spike latency effect, a neuron propert...