An improved method for IR image filtering from living beings.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Thermodynamics

Background:

• Thermal imaging is crucial for diagnosing superficial cancers, relying on heat conduction in tissues.
• Existing filtering methods often overlook heat convection, a component related to blood flow.
• Improving thermal image quality is essential for accurate oncological diagnosis.

Purpose of the Study:

• To present a new nonlinear filtering method for enhancing infrared thermal images.
• To incorporate both heat conduction and convection into the filtering process.
• To improve the speed and effectiveness of noise reduction in thermal imaging for medical applications.

Main Methods:

• Developed an improved nonlinear filtering algorithm.
• Integrated heat convection (blood flow) with heat conduction in the filtering model.
• Employed an iterative process to minimize noise in thermal images.
• Compared the algorithm's convergence speed and contour accuracy with anisotropic filtering.

Main Results:

• The proposed algorithm iteratively minimizes noise more rapidly than existing methods.
• Demonstrated improved thermal image quality through iterative filtering.
• Successfully applied the method to patient data, including a case of papillary carcinoma.
• Showcased the algorithm's effectiveness in preserving and enhancing thermal contour shapes.

Conclusions:

• The novel nonlinear filtering method effectively enhances infrared thermal image quality.
• The algorithm's consideration of heat convection improves noise reduction efficiency.
• This technique shows promise for improved diagnosis of superficial cancers by providing clearer thermal contours.