Automatic detection of perforators for microsurgical reconstruction.

Area of Science:

• Medical Imaging
• Surgical Planning
• Artificial Intelligence in Medicine

Background:

• Deep inferior epigastric perforator (DIEP) flaps are standard for mastectomy reconstruction.
• Preoperative imaging is crucial for identifying perforator vessels, guiding surgical planning.
• Current manual mapping is time-consuming and subjective.

Purpose of the Study:

• To evaluate the feasibility of computer software for preoperative planning in DIEP flap breast reconstruction.
• To assess the accuracy and efficiency of an automated blood vessel analysis tool compared to manual mapping.
• To determine the impact of AI-assisted planning on surgical preparation time and outcomes.

Main Methods:

• Applied blood vessel centerline extraction and local characterization algorithms to preoperative imaging data of 40 patients.
• Compared automated perforator identification and measurements with manual mapping and intraoperative findings.
• Assessed software accuracy for vessel caliber and location, and estimated time savings.

Main Results:

• Software demonstrated improved caliber estimates for vessels >1.5 mm (P=2e-3) but was less accurate for smaller vessels (P=6e-4).
• Vertical vessel location measurements differed significantly (P=0.02) but were clinically irrelevant (2-3 mm error).
• The automated tool reduced planning time by approximately 2 hours per case.

Conclusions:

• Computer software shows feasibility for assisting preoperative planning in DIEP flap breast reconstruction.
• AI-driven tools can reduce subjectivity and significantly decrease planning time for imaging teams.
• This pilot study supports the integration of AI into clinical practice for improved efficiency and patient care.