An adaptive pulmonary nodule detection algorithm.

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Radiology

Background:

• Lung cancer diagnosis relies heavily on identifying pulmonary nodules in CT scans.
• Increasing CT resolution generates more suspected nodule data, challenging accurate localization.
• Existing detection methods, like DBSCAN, require manual threshold adjustments, impacting efficiency.

Purpose of the Study:

• To develop an automated and accurate pulmonary nodule detection method.
• To overcome the limitations of manual thresholding in traditional clustering algorithms.
• To improve the early detection of lung cancer through enhanced nodule identification.

Main Methods:

• Utilized Contrast Limited Adaptive Histogram Equalization (CLAHE) for Region of Interest (ROI) extraction.
• Applied a region growing algorithm to process suspected nodular areas and identify key points.
• Implemented an adaptive DBSCAN clustering algorithm using mean region range as threshold and identified center points as starting clusters.

Main Results:

• The proposed method effectively extracts and clusters suspected pulmonary nodules from chest CT images.
• The adaptive DBSCAN algorithm eliminates the need for manual threshold setting, improving workflow.
• Experimental results demonstrate a significant improvement in the accuracy of pulmonary nodule detection.

Conclusions:

• The integrated region growing and adaptive DBSCAN approach offers a more accurate and automated solution for pulmonary nodule detection.
• This method aids in the early and reliable diagnosis of lung cancer.
• The enhanced detection accuracy supports improved patient outcomes through timely treatment.