Inverse method for quantitative characterisation of breast tumours from surface temperature data

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Computational Modeling

Background:

• Accurate characterization of breast tumors is crucial for diagnosis and treatment planning.
• Non-invasive methods for estimating tumor properties like size, location, and blood perfusion are highly desirable.

Purpose of the Study:

• To develop and validate a computational method for simultaneously estimating breast tumor size, location, and blood perfusion using surface temperature data.
• To investigate the efficacy of combining steady-state and transient temperature data for improved tumor characteristic estimation.

Main Methods:

• A 2D computational phantom of a breast with a spherical tumor was created, incorporating six tissue layers.
• An inverse algorithm based on the Levenberg-Marquardt method was employed to estimate tumor characteristics from surface temperature distributions.
• The inverse algorithm was extended to incorporate transient temperature data from dynamic infrared imaging, in addition to steady-state data.

Main Results:

• Steady-state temperature data alone were insufficient for simultaneously estimating tumor blood perfusion, size, and location due to similar temperature profiles.
• The integration of transient data with steady-state data enabled the simultaneous estimation of all three tumor characteristics.
• Estimation errors were below 1% for tumors with depths less than 20 mm, but exceeded 10% for deeper tumors (25 mm).

Conclusions:

• The combination of transient and steady-state surface temperature data allows for accurate simultaneous estimation of tumor size, location, and blood perfusion.
• Blood perfusion serves as an indicator of tumor growth rate, potentially aiding in the assessment of tumor malignancy.
• This computational approach offers a promising non-invasive tool for breast tumor characterization.