A Comparative Study of Lung Cancer Detection using Machine Learning Algorithms

Pr, Radhika; Nair, Rakhi. A. S.; Veena, G

doi:10.1109/icecct.2019.8869001

Cited by 73 publications

(25 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The number of persons who smoke many cigarettes in rapid succession is di-rectly proportionate to the overall number of cases of lung cancer. A number of different classification methods, including Naive Bayes, Decision Tree, SVM, and Logistic Regression, were used in the analysis of the lung cancer prediction [1]. Using CT scans derived from the SPIE-AAPM-LungX data set, lung nodules are classified in [2].…”

Section: Related Workmentioning

confidence: 99%

Pattern Knowledge Discovery Based Lung Cancer Classification System

Mangayarkarasi¹,

Chithrakkannan²,

Karthikeyan³

2023

View full text Add to dashboard Cite

An accurate detection of abnormal lung nodule detection is very important for effective treatment and surgical procedure to remove the nodules. This paper introduces an efficient deep learning model to classify lung cancer in both left and right lung. It consists of three important stages; preprocessing, lung region detection and abnormal lung nodule detection. Further, a detailed discussion about the performance of the system is given using two benchmark databases; 30 lung CT images taken from the ELCAP dataset and 130 lung CT images taken from the LIDC dataset. An algorithmic framework is first created for the purpose of segmenting left and right lung region by a morphological algorithm after removing the noise by a wiener filter. A well defined deep learning architecture is designed for effective classification or detection of abnormal lung nodule detection by semantic classification. The proposed system is validated on LIDC and ELCAP database and provides an average accuracy of 97.86%.

show abstract

Section: Related Workmentioning

confidence: 99%

Pattern Knowledge Discovery Based Lung Cancer Classification System

Mangayarkarasi¹,

Chithrakkannan²,

Karthikeyan³

2023

View full text Add to dashboard Cite

show abstract

“…The average intensities in interior and exterior areas of a mask B (x, y) are computed at each point. To optimize [9] the total energy of the contour, the mask is considered in each point separately, and the point is moved to decrease the energy function. This energy function is defined as follows:…”

Section: Segmentation Processmentioning

confidence: 99%

“…Experimental results indicated that there is a small difference between the length and short diameter of a nodule, whereas the difference would be higher for vascular areas [16,17] because vascular areas continue to be connected in successive slices of 2D CT images [9]. Thus, the length and short diameter of remaining suspicious nodule candidates are extracted as features.…”

Section: Figure 9 Feature Parameters Of a Nodulementioning

confidence: 99%

“…Length = n × slice thickness (9) n is the number of slices that transverses a nodule candidate. In this study, a slice thickness of data is mm.…”

Section: Figure 9 Feature Parameters Of a Nodulementioning

confidence: 99%

“…Radhika et al [9] proposed their work on "A Comparative Study of Lung Cancer Detection using Machine Learning Algorithms". In this work, the Authors worked on four classifiers viz SVM model, Naï ve Bayes model, Decision Tree model, Logistic Tree model, and applied these classifiers on BRATS Dataset, OASIS Dataset, NBTR Dataset are taken from UCI Machine learning repository.…”

Section: Literature Surveymentioning

confidence: 99%

See 2 more Smart Citations

An Extensive Review on Lung Cancer Detection Using Machine Learning Techniques: A Systematic Study

Joshua¹,

Chakkravarthy²,

Bhattacharyya³

2020

RIA

View full text Add to dashboard Cite

The Main Objective of this research paper is to investigate the accuracy levels of various machine learning algorithms. To find out the accuracy levels of various classifiers we evaluated the various models employed by researchers were listed out and they have few limitations and their drawbacks were listed out. After a systematic literature study, we found out that some classifiers have low accuracy and some are higher accuracy but not reached nearer of 100%. Therefore, we need to employ a more strategic way for the better classification of Lung cancer nodule. Through a systematic literature survey, the low accuracy levels were due to improper dealing of Dicom images. After an extensive study, we found that ensemble classifier was outperformed when compared with the other machine learning algorithms. Thus, by taking consideration of all the classifiers. The findings we drew was that major machine learning algorithms gave accuracy which was not close to 90%. A better model needs to be employed to increase the accuracy level, and need to be revised should be reliable and meaningful draw insights for the tumour diagnosis, so it reflects our better understanding of the classification of lung cancer. And, lastly, intensive research should be done on the field of Oncology for the better classification of benign and malignant tumours.

show abstract