Self-Guided Algorithm for Fast Image Reconstruction in Photo-Magnetic Imaging: Artificial Intelligence-Assisted...

Area of Science:

• Biomedical Imaging
• Medical Physics
• Artificial Intelligence in Medicine

Background:

• Photomagnetic imaging (PMI) combines laser-induced heating with magnetic resonance thermometry (MRT) for temperature and absorption mapping.
• Tumor detection in PMI relies on temperature contrasts caused by higher hemoglobin levels in abnormal tissues.
• Existing PMI reconstruction algorithms can be limited in accuracy, spatial resolution, and speed.

Purpose of the Study:

• To develop and evaluate a novel artificial intelligence-based image reconstruction algorithm for PMI.
• To improve the accuracy, spatial resolution, and reduce the recovery time of absorption maps in PMI.
• To leverage machine learning for enhanced tumor boundary detection and functional a priori information in PMI.

Main Methods:

• A supervised machine learning approach was employed to detect tumor boundaries directly from MRT temperature maps.
• The detected tumor information was integrated as a soft functional a priori into the standard PMI reconstruction algorithm.
• The enhanced PMI algorithm was validated using a tissue-like phantom containing inclusions simulating tumors.

Main Results:

• The AI-enhanced PMI algorithm significantly improved spatial resolution compared to standard methods.
• Absorption recovery accuracy was enhanced, achieving a low percentage error of 2%.
• Image artifacts were reduced by 15%, and the reconstruction process was accelerated approximately 9-fold.

Conclusions:

• The developed AI-based image reconstruction algorithm substantially improves PMI performance.
• This AI-driven approach offers a more accurate, higher-resolution, and faster method for absorption mapping in biomedical imaging.
• The findings demonstrate the potential of integrating AI with PMI for improved diagnostic capabilities.