This article is more than 5 years old. Information may no longer be current.
Eye tracking improves accuracy, but still has limitations
Current systems cannot compensate for eye movements exceeding 10°, a physicist said.
Click Here to Manage Email Alerts
ROME – The technology for measuring and tracking eye movements during surgery has reached acceptable standards, but there is still ground for improvement, said a speaker here at the European Society of Cataract and Refractive Surgeons winter meeting.
“At present, eye tracking systems can improve accuracy in refractive surgery procedures, but cannot compensate for eye movements exceeding 10° (1.5 mm). With that amount of rotation, they become unreliable and the only thing we can do is stop lasering,” said Lo J. Bour, PhD, of Amsterdam University Neurophysiology Clinic.
The eye, he explained, rotates in three directions: on the horizontal, vertical and torsional plane. These movements are expressed in degrees of eye rotation, which have an equivalent in the amount of displacement of corneal surface (figure 1).
“A displacement of 1 mm is about 7° of eye rotation,” Dr. Bour said. “During fixation, the involuntary movements of a normal eye usually remain within the range of 1° to 2° of rotation angle, which is about 0.3 mm at the corneal surface level. Within that range, especially when using fast lasers, refractive surgery can be done without an eye tracker. If involuntary movements are greater, centration is lost and eye trackers are needed to avoid disappointing results.”
Analyzing eye movements
Based on Dr. Bour’s project, the clinical neurophysiology laboratory of the University Hospital in Amsterdam has developed a magnetic double induction method for analyzing eye movements. The system, which is routinely used for neurological and psychiatric patients, can also be used for preoperative evaluation of eye movements in refractive surgery patients (figure 2).
“This method can give a fair idea of what kind of eye movements the subject may have during surgery,” Dr. Bour said. “It cannot be used during surgery because the magnetic field on which the system is based would be disturbed by the surgical microscope.”
When undergoing this examination, the subject sits in a magnetic field, head supported by a chin rest and a forehead band. Two circular magnetic field sensors are placed in front of the eyes (figure 4). After instillation of one drop of topical anesthetic, two golden rings are placed on the limbus of the eye (figure 3).
“When the eye moves, the sensors measure very accurately the type, amplitude, duration and direction of movements, which are then recorded on diagrams,” Dr. Bour explained. “What we normally observe during fixation are microsaccades (small, fast eye movements) of about 30 to 40 msec duration. But we also have small ocular deviations, both in the vertical and horizontal direction, when the eye blinks.”
|
Limited reliability
“Once we have analyzed the way each patient behaves during fixation, we more or less know what to expect,” Dr. Bour said. “However, the emotional stress of surgery might produce some unexpected reactions, like sudden fast eye movements (saccades). To avoid being taken by surprise by these reactions, and at the same time optimize results, we have to use eye trackers.”
But are the current eye tracking systems up to the task of compensating for these fast eye movements intraoperatively?
Dr. Bour said video-based systems, which are the most popular, have a good resolution of less than 20 miniarcs, but sample frequency, which is less than 60 Hz, is too low to have a good representation of fast eye movements.
“The video systems should be improved by increasing sample frequency to at least 250 Hz, then video-based systems would be effective and reliable eye trackers,” he said.
For Your Information:
- Lo J. Bour, PhD, can be reached at Department of Neurology, Clinical Neurophysiology, Academic Medical Center H2-222 Meibergdreef 9, 1105 AZ Amsterdam, Netherlands; (31) 20-56-63-515; fax: (31) 20-56-69-187.