Accuracy of computer-assisted navigation: significant augmentation by facial recognition software.

Area of Science:

• Neurosurgery
• Otolaryngology
• Medical Imaging

Background:

• Image guidance navigation is increasingly used in sinus and skull base surgery.
• Conventional surface registration methods have limitations, including pressure variability and loss of contact, leading to inaccuracies.
• The limited number of registration points in surface registration can also impact accuracy.

Purpose of the Study:

• To evaluate the efficacy of novel facial recognition software for image guidance registration.
• To compare the accuracy of facial recognition software against conventional surface registration.

Main Methods:

• Prospective study of consecutive adult patients undergoing endoscopic sinus surgery (ESS).
• Patients underwent registration using both conventional surface registration and facial recognition software.
• Accuracy was measured at specific anatomical landmarks: head of the middle turbinate (MTH), middle turbinate axilla (MTA), anterior wall of sphenoid sinus (SS), and nasal tip (NT).

Main Results:

• Facial recognition demonstrated high accuracy, with mean errors of 0.47 mm (MTH), 0.33 mm (MTA), 0.39 mm (SS), and 0.36 mm (NT).
• Facial recognition was significantly more accurate than surface registration at MTH (p=0.002), MTA (p<0.001), and SS (p<0.001).
• The integration of facial recognition software did not increase registration time.

Conclusions:

• Automated facial recognition software significantly enhances the accuracy of image guidance registration.
• This technology offers a more precise alternative to conventional surface registration in surgical navigation.