Azmi, Taufik (2019) Segmentasi Tumor Otak Pada Citra Mri Menggunakan Fully Convolutional Network. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06211540000107-Undergraduate_Theses.pdf Download (3MB) | Preview |
Abstract
Tumor otak adalah pertumbuhan jaringan abnormal pada sel-sel otak yang terus tumbuh dan berlipat ganda tanpa ter¬kendali. Deteksi tumor otak dapat dilakukan dengan pemeriksaan laboratorium dan pemeriksaan radiologis. Magnetic resonance imaging (MRI) adalah salah satu pemeriksaan radiologis yang dapat memindai area otak untuk mendeteksi lokasi tumor otak. Masalah medisnya adalah untuk menentukan tumor otak secara lebih tepat, yang dilakukan dengan memisahkan segmen pada otak (sebagai Region of Interest atau ROI) pada segmen lain dalam MRI. Dalam penelitian ini, klasifikasi ROI dan Non-ROI dilakukan dengan menggunakan Fully Convolutional Network (FCN). Metode ini dapat memproses data dengan pola grid dalam konstruksi matematika. Klasifikasi dengan FCN akan divalidasi dengan menghitung Correct Classification Ratio (CCR) yang dilakukan dengan membandingkan hasilnya dengan ground truth. Hasil klasifikasi dan segmentasi menggunakan FCN dapat mengenali tumor otak pada gambar MRI dengan nilai rata-rata CCR hingga 94,65%.
=================================================================================================================================
Brain tumors are an abnormal tissue growth in brain cells that continue to grow and multiply uncontrollably. The detection of brain tumor can be done under the laboratory examination and radiological examination. Magnetic Resonance Imaging (MRI) is one of the radiological examinations that could scan the brain area to detect the location of the brain tumor. The medical problem is how to specify the brain tumor more precisely, that is accomplished by separating the brain segment (as the Region of Interest or ROI) to the other segment in the MRI. In this study, the classification of ROI and Non-ROI is done by employing the Fully Convolutional Network (FCN). This method can process data with a grid pattern in a mathematical construction. Classification with FCN would be validated by calculating the Correct Classification Ratio (CCR) which is done by comparing the segmentation results with the ground truth. The results show that the classification and segmentation using FCN could recognize the brain tumor in the MRI image with average CCR value of 94,65%.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Fully Convolutional Network, Magnetic Resonance Imaging, Segmentasi, Tumor Otak |
| Subjects: | H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects ) Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Taufik Azmi |
| Date Deposited: | 27 Sep 2024 04:21 |
| Last Modified: | 27 Sep 2024 04:21 |
| URI: | http://repository.its.ac.id/id/eprint/64475 |
Actions (login required)
 |
View Item |