Klasifikasi Tumor Kulit Berdasarkan Komponen Melanosit pada Citra Dermoskopi dengan Menggunakan Metode Radial Basis Function SVM

Rahajeng, Andhryn Celica Dewi (2020) Klasifikasi Tumor Kulit Berdasarkan Komponen Melanosit pada Citra Dermoskopi dengan Menggunakan Metode Radial Basis Function SVM. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
07311640000035_Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (3MB) | Request a copy

Abstract

Kanker kulit akhir-akhir ini menjadi salah satu jenis kanker yang sering muncul dan mematikan. Di Indonesia, kanker kulit menempati urutan ketiga setelah kanker leher rahim (serviks) dan kanker payudara. Pendeteksian dan penanganan secara dini sangatlah penting. Angka kematian dari penderita kanker kulit dapat dikurangi apabila pendeteksian dan penanganannya secara dini dan tepat. Segmentasi lesi kulit biasanya dilakukan pada citra yang tergolong melanosit (mls), padahal lesi kulit yang tergolong non melanosit (nomls) juga sama pentingnya. Contoh lesi kulit yang tergolong melanosit adalah melanoma malignant, sedangkan yang tergolong non melanosit adalah basal cell carcinoma dan squamous cell carcinoma. Langkah pemrosesan pada penelitian ini yaitu pre-processing, segmentation, feature extraction, dan classification. Proses classification dilakukan dengan menggunakan metode Support Vector Machine (SVM) untuk membedakan lesi kulit pada citra dermoskopi apakah lesi tergolong melanosit (mls) atau non melanosit (nomls). Keluaran dari penelitian ini yaitu menganalisa tingkat akurasi, sensitivitas, spesifikasi, dan presisi dari tiap kernel SVM yang digunakan dan penyesuaian parameter C dan gamma (γ) yang digunakan untuk mendapatkan nilai yang paling baik. Dari hasil pengujian klasifikasi dari semua kernel dan nilai parameter yang digunakan, nilai yang paling baik didapatkan ketika menggunakan kernel radial basis function dengan menghasilkan akurasi 85%, sensitivitas 86%, spesifikasi 84%, dan presisi 88%. Metode k-Fold Cross Validation digunakan untuk mengetahui nilai k optimal dari model yang digunakan. Hasil yang didapat adalah nilai k=7 dan k=8 merupakan nilai optimal dengan menghasilkan akurasi 83%. ================================================================================================================== Skin cancer has recently become one of the types of cancer that often appears and is deadly. In Indonesia, skin cancer placed third after cervical cancer and breast cancer. Early detection and handling is very important. Mortality from skin cancer patient could be reduced if the detection and treatment is early and appropriate. Segmentation of skin lesions is usually carried out on images that have classified melanocytic (mls), whereas skin lesions that are classified as nonmelanocytic (nomls) are equally important. Examples of skin lesions which classified as melanocytic are malignant melanomas, whereas those classified as nonmelanocytic are basal cell carcinoma and squamous cell carcinoma. Required processing in this study are pre-processing, segmentation, feature extraction, and classification. The classification process is carried out using the support vector machine (SVM) to differentiate skin lesions in dermoscopic images whether they are classified as melanocytic (mls) or nonmelanocytic (nomls). The output of this research is analyzing the level of accuracy, sensitivity, specification, and precision of each SVM kernel used and the adjustment of C and gamma (γ) parameters used to get the best value. From the results of the classification testing of all kernels and the parameter values used, achieving classification of the melanocytic and nonmelanocytic image with accuracy of 85%, sensitivity of 86%, specification of 84%, and precision of 88% using radial basis function kernel. The k-Fold Cross Validation method is used to determine the optimal k value of the model. The results obtained are the value of k=7 and k=8 is the optimal value with an accuracy of 83%.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: dermoskopi, kanker kulit, lesi, pengolahan citra, klasifikasi, support vector machine, dermoscopy, skin cancer, lesions, image processing, classification
Subjects: T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Biomedical Engineering > 11410-(S1) Undergraduate Thesis
Depositing User: Andhryn Celica Dewi Rahajeng
Date Deposited: 24 Aug 2020 03:17
Last Modified: 24 Aug 2020 03:17
URI: http://repository.its.ac.id/id/eprint/79957

Actions (login required)

View Item View Item