Klasifikasi Tingkat Keganasan Kanker Payudara Melalui Breast Histopathology Image Menggunakan Convolutional Neural Network

Noviandini, Farah (2021) Klasifikasi Tingkat Keganasan Kanker Payudara Melalui Breast Histopathology Image Menggunakan Convolutional Neural Network. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 01111740000031-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (1MB) | Request a copy

Abstract

Kanker payudara menjadi penyebab utama kematian bagi wanita di Indonesia. Prediksi tenaga medis terhadap kanker payudara memerlukan keahlian untuk mengklasifikasikan jenisnya melalui citra histopatologi payudara (BreakHis) dengan akurasi tinggi secara cepat. Penelitian ini bertujuan untuk mengetahui klasifikasi tingkat keganasan BreakHis termasuk dalam kelas jinak (benign) atau kelas ganas (malignant) dengan menggunakan alogaritma CNN (Convolutional Neural Network) dan mengetahui hasil optimasi nilai akurasi kelas jinak dan kelas ganas BreakHis dengan arsitektur MobileNetV2 dan ResNet50V2. Pada penelitian tugas akhir ini digunakan 7891 dataset BreakHis dengan faktor pembesaran 40, 100, 200, dan 400 yang dapat diakses melalui wesite Kaggle. Keseluruhan gambar diresize menjadi ukuran 224224. Perangkat lunak yang digunakan adalah Jupyter dengan bahasa pemrograman Python. Hasil penelitian menunjukkan bahwa hasil akurasi tertinggi diperoleh pada model ResNet50V2 dengan nilai akurasi pelatihan 100%, 95,8% uji dan 97% validasi.
===================================================================================================
Breast cancer is one of the main causes of women’s death in Indonesia. Medical personnel's prediction of breast cancer needs the expertise to classify the type through the breast histopathology image (BreakHis) with high accuracy quickly. This study aims to determine the classification of BreakHis’ malignancy, including in the benign or malignant class using CNN (Convolutional Neural Network) algorithm and determine the optimization’s results of the accuracy benign class and malignant class using architectures of MobileNetV2 and ResNet50V2. In this final research, 7891 BreakHis datasets are used with 40x, 100x, 200x, and 400x factors, which can be accessed through the Kaggle website. The whole image is resized to 224x224 pixels. The software used is Jupiter with the Python programming language. The results showed that the highest accuracy is in the ResNet50V2 model, with accuracies in training, testing, and validation were 100%, 95,8%, and 97%, respectively.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Convolution Neural Network, Histopathology , Breast Cancer, Convolution Neural Network, Histopathology , Kanker Payudara
Subjects:	Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science) Q Science > QC Physics Q Science > QR Microbiology > QR 201.T84 Tumors. Cancer
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Physics > 45201-(S1) Undergraduate Thesis
Depositing User:	Farah Noviandini
Date Deposited:	11 Aug 2021 05:41
Last Modified:	11 Aug 2021 05:41
URI:	http://repository.its.ac.id/id/eprint/85419

Actions (login required)

View Item