IMPLEMENTASI TRANSFORMASI FOURIER FRAKSIONAL TIPE BERBOBOT DAN MODIFIKASI K-NEAREST NEIGHBOR DALAM IDENTIFIKASI TUMOR OTAK BERDASARKAN MAGNETIC RESONANCE IMAGING (MRI)

LUBIS, ANUGRAH ARIEF YAHYA (2022) IMPLEMENTASI TRANSFORMASI FOURIER FRAKSIONAL TIPE BERBOBOT DAN MODIFIKASI K-NEAREST NEIGHBOR DALAM IDENTIFIKASI TUMOR OTAK BERDASARKAN MAGNETIC RESONANCE IMAGING (MRI). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111840000077-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Kemajuan teknologi dapat mengakselerasi pekerjaan manusia dalam berbagai bidang salah satunya dalam bidang medis tak terkecuali permasalahan gangguan pada otak manusia. Pengolahan citra digital dapat diaplikasikan dalam permasalahan gangguan otak pada otak manusia khususnya tumor otak sehingga dapat membantu teknisi rekayasa medis dalam merancang algoritma identifikasi tumor otak yang efektif. Metode yang menggunakan Transformasi Fourier Fraksional Tipe Berbobot (TFFB) merupakan cara yang relatif baru dalam memperoleh karakteristik citra berupa piksel citra. Dalam Tugas Akhir ini dimplementasikan Transformasi Fourier Fraksional Tipe Berbobot dan modifikasi K-Nearest Neighbor (MKNN) dalam identifikasi tumor otak berdasarkan Magnetic Resonance Imaging (MRI). Terdapat 4 jenis citra MRI yang digunakan, yaitu citra tumor otak glioma, meningioma, pituitary, dan citra tanpa tumor otak. Citra terbagi ke dalam 2 jenis proses yaitu proses pelatihan dan pengujian. Pada proses pelatihan, citra latih diekstraksi fitur-fiturnya menggunakan TFFB selanjutnya diseleksi fitur-fiturnya menggunakan Principal Component Analysis (PCA) untuk memperoleh fitur-fitur yang representasi terhadap citra dengan dimensi yang lebih rendah. Fitur-fitur yang telah diseleksi kemudian diklasifikasikan dengan MKNN sehingga diperoleh model klasifikasi MKNN. Setelah diperoleh model klasifikasi MKNN, selanjutnya pada proses pengujian, citra uji dilakukan pengujian dengan model klasifikasi MKNN yang diperoleh pada proses pelatihan. Hasil terbaik pada Tugas Akhir ini mampu mengidentifikasi tumor otak yang memiliki rata-rata untuk keseluruhan performansi sebesar 98,4562% menggunakan ekstraksi fitur sudut sumbu-x bernilai 0,8 dan sumbu-y bernilai 0,6 pada TFFB.
==============================================================================================================================
Technological advances can accelerate human work in various fields, one of which is in the medical field, including the problem of disorders of the human brain. Digital image processing can be applied to the problem of brain disorders in the human brain, especially brain tumors so that it can assist medical engineering technicians in designing effective brain tumor identification algorithms. The method using the Weighted Type Fractional Fourier Transform (TFFB) is a relatively new method of obtaining image characteristics in the form of image pixels. In this final project, a Weighted Type Fractional Fourier Transform and modification of K-Nearest Neighbor (MKNN) is implemented in the identification of brain tumors based on Magnetic Resonance Imaging (MRI). There are 4 types of MRI images used, namely glioma, meningioma, pituitary brain tumor images, and images without brain tumors. The image is divided into 2 types of processes, namely the process of training and testing. In the training process, the features of the training image are extracted using TFFB and then the features are selected using Principal Component Analysis (PCA) to obtain features that are representative of the image with lower dimensions. The features that have been selected are then classified with MKNN so that the MKNN classification model is obtained. After obtaining the MKNN classification model, then in the testing process, the test image is tested with the MKNN classification model obtained in the training process. The best results in this Final Project are able to identify brain tumors that have an average overall performance of 98.4562% using the x-axis angle feature extraction with a value of 0.8 and the y-axis value of 0.6 on TFFB.

Item Type:	Thesis (Other)
Additional Information:	RSMa 006.42 Lub i-1 2022
Uncontrolled Keywords:	Magnetic Resonance Imaging, Modifikasi K-Nearest Neighbor, Transformasi Fourier Fraksional Tipe Berbobot, Tumor Otak. Magnetic Resonance Imaging, Modified K-Nearest Neighbor, Weighted- Type Fractional Fourier Transform, Brain Tumor.
Subjects:	Q Science > QA Mathematics
Depositing User:	Mr. Marsudiyana -
Date Deposited:	08 Jun 2026 06:07
Last Modified:	08 Jun 2026 06:07
URI:	http://repository.its.ac.id/id/eprint/133631

Actions (login required)

View Item