Colorectal cancer is one of the most common types of cancer experienced by the world's population and causes approximately 700,000 deaths each year. Early detection is one of the keys to detecting polyps before they transform into malignant colorectal cancer. Polyp detection is a complicated matter because polyp detection involves many factors. This research attempts to make new improvisations in the polyp detection process in terms of the pre-processing stage and classification stage by utilizing statistical selection techniques for a collection of bag of features and the feedforward neural network classifier function. Based on the experiments that have been carried out, it was found that the proposed method was able to provide the highest accuracy in recognizing polyps of 97.85%.Keywords: Cancer; Colorectal; Polyp; Neural Network AbstrakKanker kolorektal adalah salah satu jenis penyakit kanker yang sering dialami oleh penduduk dunia dan menjadi penyebab kurang lebih 700,000 kematian setiap tahunnya. Tindakan deteksi dini merupakan salah satu kunci untuk mendeteksi adanya polip, sebelum bertransformasi menjadi kanker kolorektal ganas. Deteksi polip merupakan suatu hal yang rumit karena deteksi polip melibatkan banyak faktor. Penelitian ini melakukan improvisasi baru dalam proses deteksi polip dari sisi pre-processing stage serta classification stage dengan mengutilisasi teknik seleksi statistik terhadap kumpulan bag of features serta fungsi classifier feedforward neural network. Berdasarkan percobaan yang telah dilaksanakan didapatkan bahwa metode yang diusulkan mampu memberikan nilai akurasi tertinggi dalam mengenali polip sebesar 97.85%.Kata kunci: Kanker; Kolorektal; Polyp; Neural Network
Copyrights © 2023