SIFT-GVF-based lung edge correction method for correcting the lung region in CT images

Area of Science:

• Medical Imaging
• Computer Vision
• Image Segmentation

Background:

• Traditional Hounsfield unit thresholding methods exclude juxtapleural nodules from lung segmentation.
• Accurate lung segmentation is crucial for diagnosing and monitoring pulmonary conditions.

Purpose of the Study:

• To develop and validate a new approach for re-including juxtapleural nodules in segmented lung regions.
• To enhance the accuracy of lung segmentation in computed tomography (CT) images.

Main Methods:

• Utilized scale-invariant feature transform (SIFT) to detect key points and supportive boundary lines.
• Employed Fourier descriptors and spectrum energy to identify boundaries requiring correction.
• Applied the gradient vector flow-snake method for precise and smooth correction of recognized boundaries.

Main Results:

• The proposed method successfully detected and corrected juxtapleural regions.
• Experiments on authentic CT images demonstrated precise correction of lung boundaries.
• The approach proved robust and achieved perfect results in re-including excluded regions.

Conclusions:

• The novel SIFT and gradient vector flow-based method accurately corrects juxtapleural regions.
• This technique significantly improves lung segmentation accuracy in CT imaging.
• The findings offer a more comprehensive approach to analyzing lung structures in medical imaging.