Application of a maximum likelihood estimator in an experimental study in ultrasonic diffraction tomography

Area of Science:

• Diffraction Tomography
• Inverse Problems
• Computational Imaging

Background:

• Inverse problems are crucial for reconstructing object properties from indirect measurements.
• Diffraction tomography offers a powerful, non-invasive imaging modality.
• Accurate object detection and localization are fundamental in various imaging applications.

Purpose of the Study:

• To evaluate a novel approach for detecting and locating known objects using diffraction tomographic data.
• To assess the performance and robustness of filtered backpropagation algorithms in this context.
• To explore the potential of this method in medical imaging, specifically for soft tissues.

Main Methods:

• Utilized experimental data to validate filtered backpropagation algorithms.
• Assessed algorithm performance under conditions of modeling errors.
• Tested robustness against limited angular coverage in tomographic data acquisition.

Main Results:

• Experimental validation confirmed the efficacy of the proposed inverse problem approach.
• Filtered backpropagation algorithms demonstrated significant robustness to modeling inaccuracies.
• The method proved effective even with severely limited angular data.

Conclusions:

• The evaluated approach provides a reliable method for object detection and localization in diffraction tomography.
• The demonstrated robustness makes it suitable for real-world applications with imperfect data.
• This technique holds potential for advancing soft tissue medical imaging.