Medical Images Breast Cancer Segmentation Based on K-Means Clustering Algorithm: A Review

Abstract

Early diagnosis is considered important for medical images of breast cancer, the rate of recovery and safety of affected women can be improved. It is also assisting doctors in their daily work by creating algorithms and software to analyze the medical images that can identify early signs of breast cancer. This review presents a comparison has been done in term of accuracy among many techniques used for detecting breast cancer in medical images. Furthermore, this work describes the imaging process, and analyze the advantages and disadvantages of the used techniques for mammography and ultrasound medical images. K-means clustering algorithm has been specifically used to analyze the medical image along with other techniques. The results of the K-means clustering algorithm are discussed and evaluated to show the capacity of this technique in the diagnosis of breast cancer and its reliability to identify a malignant from a benign tumor.

Klasifikasi Citra Digital Mammografi Berdasarkan Luas Diameter Kanker Payudara dengan Metode K-Means Clustering

“…Teknik MRI, mamografi, dan ultrasonografi digunakan dalam pencitraan lesi payudara [12]. Mammogram tidak menyembuhkan kanker payudara, tetapi dapat menyelamatkan nyawa pasien dengan mendeteksi kanker payudara sedini mungkin [10].…”

Introducing AI to the molecular tumor board: one direction toward the establishment of precision medicine using large-scale cancer clinical and biological information

“…This is because deep learning technology is particularly strong in image analysis, and when medical AI is practically applied to clinical practice, it is important to use data obtained from physicians’ diagnoses (especially specialists) as training data. Unsupervised learning is subsequently used for clustering and dimensionality compression, and its applications in medical research include patient stratification, medical image categorization, and disease subdivision [ 64 – 68 ]. Semi-supervised learning combines supervised and unsupervised learning.…”

Comparison of G-Means Algorithms and Kohonen Network in Solving Clustering Problems

“…Одной из популярных и часто используемых методик кластеризации является разделительная кластеризация, где на n элементов задается k кластеров, при этом k < n. К простым и эффективным алгоритмам относится K-Means. Он активно используется в задачах медицины [5][6], анализа текста [7][8], информационной безопасности [9] и других.…”