Skripsi

KLASIFIKASI CITRA MRI TUMOR OTAK MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORKS

---

Brain tumor classification aims to identify abnormal and normal cell growth in brain tissue. Magnetic Resonance Imaging (MRI) is used because it can produce high-resolution images without ionizing radiation, but the classification process, which is still performed manually by radiologists, is time-consuming and prone to error. To address this issue, this study applied Convolutional Neural Networks (CNN) with three architectures, namely VGG-16, MobileNet-V2, and Xception. The dataset consisted of 7,828 MRI images with four classes: glioma, meningioma, pituitary, and no tumor. Testing was conducted through 12 scenarios combining batch size (32 and 64) and learning rate (0.001 and 0.0001). The results showed that the Xception architecture with a batch size of 32 and a learning rate of 0.0001 produced the highest performance with an accuracy of 98.08%, precision of 98.01%, recall of 98.06%, and an F1-score of 98.03%. The VGG-16 architecture recorded an accuracy of 97.32%, while MobileNet-V2 achieved an accuracy of 96.94% with stable precision, recall, and F1-score values. These results indicate that the use of a CNN architecture with optimal hyperparameter settings can be an effective solution in the process of predicting and classifying brain tumors based on MRI images.

---

Availability

Detail Information

Series Title

-

Call Number

T1867342025

Publisher

Indralaya : Prodi Teknik Informatika, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xvi, 125 hlm.; ilus.; tab.; 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

616.075 707

Content Type

Text

Media Type

unmediated

Carrier Type

other (computer)

Edition

-

Subject(s)

Prodi Teknik Informatika
Diagnosis Medis--Radiologi MRI

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment