Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Table 1.

The latest computational models in lung cancer.

More »

Table 1 Expand

Table 2.

Comparison of Sample Sizes in Various Studies.

More »

Table 2 Expand

Fig 1.

The modelling process framework.

This is a structured flowchart outlining a machine learning pipeline for classifying lung conditions from CT scans, including preprocessing, segmentation, feature extraction, modeling, and classification into six classes.

More »

Fig 1 Expand

Fig 2.

Image and Mask.

This figure shows a CT scan (left) and its corresponding segmentation mask (right), representing the two input modalities to the proposed DbMCA model. The mask isolates the lung region, enabling focused feature extraction for downstream analysis.

More »

Fig 2 Expand

Fig 3.

The Proposed Model - Dual-Branch Model Classification Approach (DbMCA).

This diagram illustrates the DbMCA architecture comprising two parallel branches: a CNN pipeline (left) with four stages and 19 steps for processing image inputs, and a DNN pipeline (right) with seven steps for processing mask inputs. Outputs from both branches are concatenated and passed through a SoftMax layer for final classification into 6 classes of cancer level.

More »

Fig 3 Expand

Table 3.

Metrics values (Accuracy, F1 Score, Recall, and Precession) – Experiment 1.

More »

Table 3 Expand

Table 4.

Metrics values (Accuracy, F1 Score, Recall, and Precession) – Experiment 2.

More »

Table 4 Expand

Fig 4.

Comparison of Metrics Results for all Models for Experiment 1 (20k).

This bar chart compares the proposed Dual-Branch Model Classification Approach (DbMCA) against CNN, DNN, and SVM models using image and mask inputs. Metrics include Precision, Recall, F1 Score, and Accuracy. DbMCA achieved the highest overall performance, with 91.21% accuracy and balanced precision and recall across modalities.

More »

Fig 4 Expand

Fig 5.

Comparison of Metrics Results for all Models for Experiment 2 (5k).

This bar chart compares the proposed Dual-Branch Model Classification Approach (DbMCA) against CNN, DNN, and SVM models using image and mask inputs. Metrics include Precision, Recall, F1 Score, and Accuracy. DbMCA achieved the highest overall performance, with 98.04% accuracy and balanced precision and recall across modalities.

More »

Fig 5 Expand

Table 5.

Paired T-Test Results Summary (Experiment 1).

More »

Table 5 Expand

Table 6.

Table 5- Paired T-Test Results Summary (Experiment 2).

More »

Table 6 Expand