ADSPAO: Enhanced artemisinin optimization for multi-threshold segmentation of chronic obstructive pulmonary disease

Area of Science:

• Medical Imaging
• Computational Intelligence
• Pulmonology

Background:

• Accurate segmentation of medical computed tomography (CT) images is crucial for diagnosing Chronic Obstructive Pulmonary Disease (COPD).
• Existing segmentation techniques face challenges with optimization efficiency, accuracy, and high computational demands.
• There is a need for advanced algorithms to improve COPD diagnosis through precise medical image analysis.

Purpose of the Study:

• To propose an improved optimization algorithm, the adaptive delivery spiral propagation artemisinin optimization algorithm (ADSPAO), for enhanced medical image segmentation.
• To apply ADSPAO for multi-threshold segmentation of COPD CT images.
• To evaluate the performance of ADSPAO against existing methods in terms of segmentation accuracy and efficiency.

Main Methods:

• Development of the ADSPAO algorithm, integrating spiral propagation (SP) and adaptive delivery (AD) for improved local and global search capabilities.
• Utilizing Renyi entropy as the fitness function for multi-threshold segmentation of COPD CT images.
• Comparative analysis of ADSPAO against other algorithms using the IEEE CEC2017 benchmark and medical imaging datasets.

Main Results:

• ADSPAO demonstrated superior robustness and stability on the IEEE CEC2017 benchmark functions.
• The algorithm achieved significantly better segmentation performance on COPD CT images.
• Quantitative improvements were observed in peak signal to noise ratio (PSNR), structural similarity (SSIM), and feature similarity (FSIM).

Conclusions:

• The proposed ADSPAO algorithm offers an efficient and reliable approach for segmenting COPD CT images.
• ADSPAO overcomes the limitations of current methods, providing enhanced accuracy and reduced computational cost.
• This advancement holds promise for improving the diagnostic capabilities in the field of pulmonology.