Klasifikasi Trombosit Pada Citra Hapusan Darah Tepi Berdasarkan Gray Level Co-Occurrence Matrix Menggunakan Backpropagation

Fitri, Zilvanhisna Emka (2017) Klasifikasi Trombosit Pada Citra Hapusan Darah Tepi Berdasarkan Gray Level Co-Occurrence Matrix Menggunakan Backpropagation. Masters thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 2215205005-Master_Thesis.pdf]
Preview
Text
2215205005-Master_Thesis.pdf - Published Version

Download (6MB) | Preview

Abstract

Penelitian ini bertujuan untuk melakukan klasifikasi sel trombosit pada citra hapusan darah berdasarkan tekstur fitur yaitu menggunakan Gray Level Co-occurrence Matrix (GLCM). Fitur yang digunakan adalah Angular Second Moment (ASM), Invers Different Moment (IDM), dan Entropi. Fitur-fitur tersebut menjadi masukan pada proses klasifikasi. Backpropagation digunakan untuk mengklasifikasi antara sel leukosit, sel trombosit normal dan sel trombosit raksasa. Hasil pengujian, backpropagation mampu mengklasifikasikan jenis sel dengan akurasi pada leukosit 90.31%, trombosit normal, 93.88% dan trombosit raksasa 86.22%. Berdasarkan jenis citra AB 85.71%, citra AL, 87.75%, citra BG 84.69% dan citra RG 83.67%. Jadi sistem klasifikasi ini mampu digunakan sebagai alat bantu bagi dokter atau analis medis untuk mempercepat proses diagnosis trombosit pada bidang kesehatan.

=====================================================================

This study aims to classify platelet cells in blood smear image based on feature texture using Gray Level Co-occurrence Matrix (GLCM). The features used are Angular Second Moment (ASM), Inverse Different Moment (IDM), and Entropy. These features become input to the classification process. Backpropagation is used to classify between leucocyte cells, normal platelet cells and giant platelet cells. Test results, backpropagation able to classify cell types with accuracy on leukocyte 90.31%, normal platelet 93.88% and giant platelet 86.22%. Based on AB image type 85.71%, AL image, 87.75%, BG image 84.69% and image RG 83.67%. So this classification system can be used as a tool for doctors or medical analysts to accelerate the process of platelet diagnosis in the field of health.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Backpropagation, Gray Level Co-occurrence Matrix (GLCM), Giant Platelet, Trombosit raksasa
Subjects: R Medicine > RB Pathology
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions: Faculty of Industrial Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User: Zilvanhisn Emka Fitri
Date Deposited: 18 Sep 2017 03:29
Last Modified: 05 Mar 2019 04:29
URI: http://repository.its.ac.id/id/eprint/42887

Actions (login required)

View Item View Item