| Autor: |
S. Hemapriyaa, V. Sindhu, M. Singaravelan, J. Ramadevi, S. Vinitha |
| Rok vydání: |
2020 |
| Předmět: |
|
| Zdroj: |
Springer Tracts in Nature-Inspired Computing ISBN: 9789811550966 |
| DOI: |
10.1007/978-981-15-5097-3_9 |
| Popis: |
Recently, a large number of inspection methods are implemented to examine the brain MRI slices recorded using Flair/T2 modality. The Flair/T2 modality provides more visible brain abnormalities than other modalities. This work implemented a heuristic algorithm based examination procedure. In this, the Bat Algorithm (BA) is used to improve the visibility of the tumor-pixels using the Kapur’s Entropy (KE). The improved tumor section is then collected using the Watershed Segmentation Method (WSM). Later, a comparison linking the tumor portion and available Ground-Truth Image (GTI) is executed and the Quality Measures (QM) is computed individually for Flair and T2. During this investigation, the MRI slices with 240 × 240 × 1 pixel resolution are considered and this technique is implemented on 400 MRI slices (Flair = 200 + T2 = 200). The average result of QM confirmed that the BA-based technique helped to get superior result (Accuracy >95% for Flair/T2 modality slices). In future, this methodology can be implemented to analyze the MRI slices obtained from the hospitals. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
