FPM-R²Net: Fused Photoacoustic and operating Microscopic imaging with cross-modality Representation and Registration...

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Neurosurgery

Background:

• Robot-assisted microsurgery requires precise vision guidance for visualizing microvasculature.
• Conventional microscopy faces limitations in resolution, contrast, and susceptibility to lighting conditions.
• Photoacoustic microscopy (PAM) offers high resolution and contrast for vascular imaging.

Purpose of the Study:

• To develop an integrated imaging system for improved intraoperative visualization in microsurgery.
• To enhance guidance in robot-assisted procedures by fusing standard microscopy with PAM.
• To introduce the FPM-R2Net for cross-modality representation and registration.

Main Methods:

• Proposed the FPM-R2Net, comprising MORNet for unified feature extraction and HIRNet for hierarchical registration.
• Utilized a synthetic dataset with ground truth and an in vivo mouse brain vasculature dataset.
• Evaluated the network's performance using intersection over union and Dice scores.

Main Results:

• The fused imaging scheme significantly improved visualization compared to existing methods.
• FPM-R2Net achieved substantial gains in intersection over union (10.3% synthetic, 15.9% in vivo) and Dice scores (6.6% synthetic, 11.8% in vivo).
• Demonstrated superior performance on both synthetic and in vivo datasets.

Conclusions:

• The proposed fused imaging system enhances intraoperative guidance for robot-assisted microsurgery.
• FPM-R2Net effectively integrates white-light and photoacoustic microscopy for superior vascular delineation.
• This approach holds significant potential for advancing surgical precision and patient outcomes.