Get started
Work

Breast Cancer Detection and Classification by Features Non-Linear Mapping with Random Forest Classifier

Year: 2023
Type: article
Abstract: The current imaging technique of choice for breast cancer screening is mammography. Mammography’s primary encouraging results are masses and calcification. If breast cancer cases are solely relied upon for diagnosis, a sizable portion will be overlooked or incorrectly identified due to the varying appearance of lumps and calcification. Mammography has demonstrated encouraging results with the application of deep learning technology in the quantitative assessment of parenchymal density, categorization, detection, diagnosis, and prognosis of breast cancer risk, allowing more accurate patient management. The idea of deep learning has also improved the workflow efficiency of interpretation by lowering interpretation time and the workload. To definitively demonstrate the efficacy of deep learning, more thorough research is needed. The classification of mammography using a deep learning process is covered in this article. And how it can be used for mammography interpretation, as well as the difficulties it is now facing in actual use. In proposed approach use Long Short-Term Memory (LSTM) based sequence learning and Convolution Neural Network (CNN) based Non-linear feature mapping it improve Mammographic Image Analysis Society (MIAS) dataset accuracy 5%, Precision 6% and recall 4.6%. INBREAST and Digital Database for Screening Mammography (DDSM) dataset using different performance metrics by these experiments validate our approach. In comparison with existing approaches proposed approach improves accuracy by 2-3%, precision by 2% and recall 3-4% in DDSM dataset. It improves accuracy by 3-4%, precision by 2-3%, and recall by 4% in the INBREAST dataset.
Language: English
Open Access status: diamond
OpenAlex ID: W7216773896
FWCI:
Cites: 1
Cited by:
Related to: 10