AUTOMATIC FEATURE EXTRACTION FOR BREAST DENSITY SEGMENTATION AND CLASSIFICATION
Journal Title: Asian Journal of Pharamceutical and Clinical Research - Year 2017, Vol 10, Issue 12
Abstract
Objective: Cancer is the uncontrollable multiplication of cells in human body. The expansion of cancerous cells in the breast area of the women is identified as breast cancer. It is mostly identified among women aged above 40. With the current advancement in the medical field, various automatic tests are available for the identification of cancerous tissues. The cancerous cells are spotted by taking the photo imprint in the form of X-ray comprising the breast area of the woman. Such images are called mammograms. Segmentation of mammograms is the primary step toward diagnosis. It involves the pre-processing of the image to identify the region of interest (ROI). Later, features are extracted from the image which involves the learned features that may be statistical and textural features [7]. When these features are used as input to the simple classifier, it helps us to predict the risk of cancer. The support vector machine (SVM) classifier was proved to produce a better accuracy percentage with the features extracted.Methods: The mammograms are subjected to a pre-processing stage, where the images are processed to identify the ROI. Next, the features are extracted from these images to identify the statistical [9] and textural features. Finally, these features are used as input to the simple classifier, it helps us to predict the risk of cancer.Results: The SVM classifier was proved to produce the maximum accuracy of about 88.67% considering 13 features including both statistical and textural features. The features taken for the study are mean, inverse difference moment, energy, entropy, root mean square, correlation, homogeneity, variance, skewness, range, contrast, kurtosis, and smoothness.Conclusion: Computer-aided diagnosis is one of the most common methods of detection of cancer with mammograms, and it involves minor human intervention. The dataset of mammograms was analyzed and found that SVM provided the highest accuracy of 88.67%. A wide range of the study is progressing in the field of cancer as this disease causes a high threat of human life in this era.
Authors and Affiliations
Aswathy K Cherian, Poovammal E, Malathy C
Keywords
Related Articles
- EP ID EP605509
- DOI 10.22159/ajpcr.2017.v10i12.19699
- Views 74
- Downloads 0
EXPRESSION OF CANCER STEM CELL MARKER OCTAMER-BINDING TRANSCRIPTION FACTOR 4 IN HIGH-GRADE TRANSITIONAL CELL CARCINOMA
Objectives: The objectives of this study were to provide an outlook of urothelial carcinoma through the immunohistochemical expression patterns of octamer-binding transcription factor 4 (OCT4) in high-grade transitional...
PROBIOTICS: THE UNDERVALUED CONQUERORS
ABSTRACTProbiotics or “beneficial microorganisms†have recently acquired the status of blockbusters in the pharmaceutical industry. Despite being commerciallyavailable, little is known about the magnitude to which th...
IN VITRO EVALUATION OF CYTOTOXIC AND GENOTOXIC EFFECTS OF PLANT EXTRACTS FROM NOTHAPODYTES FOETIDA (WIGHT) SLEUMER (FAMILY: ICACINACEAE)
Objective: The objective of this study is to investigate cytotoxic and genotoxic properties of aqueous and methanolic extracts of the aerial parts of Nothapodytes foetida (Wight) Sleumer plant.Methods: The cytotoxic effe...
MOLECULAR TARGET AND THERAPEUTIC ASPECTS OF RHEUMATOID ARTHRITIS: A REVIEW.
 Rheumatoid arthritis (RA) is an autoimmune and chronic inflammatory disease. Cytokine plays a significant role in modulating the immune systemand specifically the inflammation. The role of cytokines in disease has led...
EFFECT OF NERVE CONDUCTION STUDIES IN HYPOTHYROIDISM - AN OBSERVATIONAL STUDY
Objective: This observational study was conducted during the year 2016–2017 to assess the electrodiagnostic evidence of peripheral nerve dysfunction among newly diagnosed hypothyroid patients attending a tertiary care...