Abstract
Purpose: :
For the development of DAISY, an instrument to accurately, fastly and objectively assess strabismus in young children, recording of head pose (position and rotation), is imperative. We identified requirements for the measurement of the head pose, with particular emphasis on its relation with various kinds of incomitance.
Methods: :
DAISY should measure the angles of strabismus with a minimal accuracy of 1.0° (95% CI) within 30s. Clinical requirements for assessing head pose in incomitant strabismus were identified. Various incomitances were simulated to delineate these requirements further: V- and A-patterns, up- and downshoot in adduction, trochlear and abducens palsy and Duane and Brown syndrome. Technological methods for remote estimation of head pose were investigated and evaluated.
Results: :
DAISY measured the angles of strabismus in gaze ahead with accuracies varying between 0.4°- 3.0°, depending on the angle of strabismus. The required accuracy of the estimation of the head pose to detect a unidirectional, linear form of incomitant strabismus (e.g. long-standing N.VI palsy) was ±3°. Additional accuracy of head pose estimation is required in case of nonlinear unidirectional (e.g. recently acquired N.VI palsy, Duane syndromes), linear skewed incomitance (e.g. long-standing N.IVpalsy), and non linear skewed incomitances (e.g. recently acquired N.IV palsy or Brown syndrome). The resulting clinical requirements were: min. accuracy: ±3°, range: 50° horizontally and vertically, continuous measurement, non invasive, > 15Hz. Potential technological methods that were evaluated, like search coils, marker based video tracking and feature based video tracking, all met these requirements.
Conclusions: :
The technical methods investigated are adequate to assess the head pose. Even unrestrained, video based measurement with a single camera could be adequate.
Keywords: eye movements • strabismus • strabismus: diagnosis and detection