Page 276 - 2024-Vol20-Issue2

272 |
Fig. 3. Proposed model architecture.
Fig. 4. Average metrics results comparison of the proposed
model.
comprehensive understanding of the model’s performance
across various evaluation metrics, including precision, recall, (Hybrid-LCSCDM) that comprises detection modules and
and F1 score. The goal of this study was to use CT scans to
identify and categorize lung cancer, which is proven to be one
of the main causes of death worldwide. The outcomes showed
that the suggested model could accurately predict CT images
with the necessary sensitivity.
Deep learning is now often employed in a variety of machine
vision applications, including segmentation, object recogni-
tion, and image categorization [32]. Using deep neural net-
works in databases, it is possible to classify images with high The proposed Hybrid-LCSCDM module utilizes neural net-
accuracy. Machine vision and deep learning have been used
to automatically diagnose cancer in a number of works. In
this domain, the outcomes of these study efforts are commend-
able. For the purpose of detecting and classifying lung cancer
Qadir, Abdalla & Abd
in this investigation, we created a hybrid model. When the
proposed model’s performance is compared to that of earlier
research for the same purpose, it outperforms all of the work
that has been done. The suggested system was trained us-
ing two techniques: XGBoost classifier for classification and
VGG16 for feature extraction. The findings indicated that
the hybrid approach-based proposed solution had the highest
overall accuracy value at 98.54 percent.
V. CONCLUSION
In conclusion, this research concentrated on developing a
Hybrid Lung Cancer Stage Classifier and Diagnosis Model
stage classifiers with the aim of achieving a reduction in the
rate of death caused by lung cancer. Lung cancer, as a fatal
disease, is caused by the unlimited distribution of cells in the
lung tissue, which is the main cause of this type of cancer.
Early detection of lung cancer is vital since it may improve
patient survival. This paper presents a hybrid transfer learning
and machine learning model that works on computed tomog-
raphy images in order to detect lung cancer.
works using a hybrid learning approach as a predictor and
stage classifier. The process of classification includes feature
extraction using a pre-trained model, namely VGG16, pre-
ceded by the CT image classification through the use of the
XGBoost algorithm. The suggested model more accurately
diagnoses lung cancer via computed tomography images com-
pared to the previous studies. The findings of the performance
evaluation show the accuracy of the model as high as 98.54%;