Detection of Juxtapleural Nodules in Lung Cancer Cases Using an Optimal Critical Point Selection Algorithm

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Image Processing

Background:

• Lung cancer detection relies heavily on accurate image processing techniques.
• Lung segmentation is a critical pre-processing step for computer-aided diagnosis (CAD) systems.
• Identifying nodules, especially those near the lung boundary, remains a challenge.

Purpose of the Study:

• To develop an automated system for lung cancer detection from CT images.
• To improve the segmentation and detection of lung nodules, including juxtapleural ones.
• To reduce the computational complexity and time required for lung nodule detection.

Main Methods:

• An automated system using computed tomography (CT) images.
• Nodule segmentation via an optimal critical point selection (OCPS) algorithm and bidirectional chain code (BDC).
• Classification of nodules using support vector machine and random forest classifiers based on extracted shape and size features.

Main Results:

• The OCPS algorithm effectively detects shape- and size-based juxtapleural nodules.
• The combined OCPS and random forest classifier approach demonstrated superior performance metrics (precision, recall, accuracy, F-measure) compared to existing methods.
• The proposed method significantly reduces computation time and complexity.

Conclusions:

• The automated system effectively identifies lung cancer nodules from CT scans.
• The OCPS algorithm combined with a random forest classifier offers a more efficient and accurate approach to lung nodule detection.
• This method holds promise for improving early lung cancer diagnosis.