Title: LESH - feature extraction and cognitive machine learning techniques for recognition of lung cancer cells

Authors: Ummadi Janardhan Reddy; B. Venkata Ramana Reddy; B. Eswara Reddy

Addresses: Department of Computer Science and Engineering, JNTUA, Ananthapuramu, India ' Department of Computer Science and Engineering, Nalanda Institute of Engineering and Technology (NIET), Guntur, India ' Department of Computer Science and Engineering, JNTUACE, Kalikiri, Chittoor, India

Abstract: The novel local energy-based shape histogram (LESH) feature mining strategy was proposed for different cancer predictions. This paper stretches out unique work to apply the LESH system to distinguish lung cancer using machine learning approaches. As the traditional neural network systems are complex and time consuming, machine learning approaches are considered in this work, which atomises the process of tumour identification. Before LESH feature extraction, we upgraded the radiograph pictures utilising a complexity constrained versatile histogram adjustment approach. Subjective machine learning classifiers are chosen specifically extraordinary learning machine approach; support vector machine (SVM) connected utilising the LESH impassive features for effective analysis of right therapeutic state in the X-ray and MRI pictures. The framework comprises of feature extraction stage, including choice stage and order stage. For including extraction/choice distinctive wavelets capacities have been connected to locate the noteworthy exactness. Grouping K-nearest neighbour calculation has been created/used for arrangement. The informational collection used in the proposed work has 114 knob regions and 73 non-knob districts. Precision levels of more than 96% for characterisation that have been accomplished which exhibit the benefits of the proposed approach.

Keywords: local energy-based shape histogram; LESH; feature extraction; cancer cell; K-nearest neighbour; classification.

DOI: 10.1504/IJCAET.2021.115943

International Journal of Computer Aided Engineering and Technology, 2021 Vol.15 No.1, pp.32 - 45

Received: 13 Jun 2018
Accepted: 18 Oct 2018

Published online: 12 May 2021 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article