Method and system for performing automatic vestibular assessment
The invention provides a headgear apparatus for performing vestibular assessment of a user. The headgear apparatus includes a processor, a memory and a display device for displaying a plurality of visual stimuli and a hardware means to provide a plurality of non-visual stimuli to the user. A plurality of infrared cameras and an inertial measurement device are employed for measuring eye movement and head movement in response to visual stimuli/non-visual stimuli provided to the user and generating one or more graph based patterns. Headgear apparatus further includes a laser source for providing a laser beam to be projected on a surface during presentation of non-visual stimuli. Accordingly, a machine learning model is employed for interpreting the one or more graph based patterns to detect the presence of an abnormality.
1 . A headgear apparatus for performing vestibular assessment of a user, the headgear apparatus comprising:
a processor;
a memory communicatively coupled to the processor;
a display device communicatively coupled to the processor and the memory, wherein the display device is configured to display a plurality of visual stimuli to the user;
a laser source communicatively coupled to the processor and memory, wherein the laser source is configured to project a laser beam providing a head-fixated target onto an external surface or wall;
a plurality of hardware devices communicatively coupled to the processor and the memory, wherein the plurality of hardware devices is configured to provide a plurality of non-visual stimuli to the user;
a plurality of infrared cameras communicatively coupled to the processor and the memory, wherein the plurality of infrared cameras is configured to measure an eye movement of the user in response to the plurality of visual stimuli, the plurality of non-visual stimuli, and the head-fixated target; and
an inertial measurement device communicatively coupled to the processor and the memory, wherein the inertial measurement device is configured to determine a head movement of the user in response to the plurality of visual stimuli, the plurality of non-visual stimuli, and the head-fixated target, wherein the inertial measurement device is further configured to differentiate between a deliberate head movement and an unintentional head movement,
and wherein the plurality of infrared cameras and the inertial measurement device are configured to simultaneously measure eye movement and head movement relative to the head-fixated target;
wherein the processor is configured to:
generate one or more temporally-correlated graph-based patterns based on a combination of the eye movement and the head movement of the user; and
interpret the one or more temporally-correlated graph-based patterns using a machine learning model utilizing at least one of a deep neural network (DNN) and a convolutional neural network (CNN), wherein the machine learning model is trained on labeled vestibular datasets that include VOR suppression responses to the head-fixated target to detect a presence of an abnormality selected from a torsion pattern, a nystagmus pattern, or a vestibulo-ocular reflex (VOR) pattern abnormality.
2 . The headgear apparatus of claim 1 , wherein the plurality of visual stimuli corresponds to one of a smooth pursuit test, saccade test, optokinetic test, gaze test or a combination.
3 . The headgear apparatus of claim 1 , wherein the plurality of non-visual stimuli corresponds to one of mastoid vibration, caloric, Valsalva, Hyperventilation, Vestibulo-Ocular Reflex (VOR) suppression, positional stimulus, head shaking, head thrust or head impulse.
4 . The headgear apparatus of claim 1 , wherein a mixed simulation of visual stimuli and non-visual stimuli is provided to the user.
5 . The headgear apparatus of claim 4 , wherein the mixed simulation includes a real-life projection.
6 . The headgear apparatus of claim 1 , wherein the vestibular assessment includes identifying a vestibular condition associated with the user, wherein the vestibular condition is one of brain diseases, nerve palsy, concussion, Attention-deficit/hyperactivity disorder (ADHD), strokes, early Parkinson's, Multiple Sclerosis or Benign Paroxysmal Positional Vertigo (BPPV).
7 . The headgear apparatus of claim 1 is a pair of augmented reality goggles.
8 . A method for performing vestibular analysis/assessment of a user using a headgear apparatus, the method comprising:
displaying, by one or more processors of the headgear apparatus, a plurality of visual stimuli to the user;
providing, by one or more processors of the headgear apparatus, a plurality of non-visual stimuli to the user;
projecting, by a laser source communicatively coupled to the one or more processors, a laser beam providing a head-fixated target onto an external surface or wall;
measuring, using a plurality of infrared cameras communicatively coupled to the one or more processors of the headgear apparatus, an eye movement of the user in response to the plurality of visual stimuli displayed to the user, the plurality of non-visual stimuli provided to the user, and the head-fixated target; and
determining, using an inertial measurement device communicatively coupled to the one or more processors of the headgear apparatus, a head movement of the user in response to the plurality of visual stimuli, the plurality of non-visual stimuli, and the head-fixated target, wherein the inertial measurement device is further configured to differentiate between a deliberate head movement and an unintentional head movement;
and wherein the plurality of infrared cameras and the inertial measurement device are configured to simultaneously measure eye movement and head movement relative to the head-fixated target;
generating, by one or more processors of the headgear apparatus, one or more temporally-correlated graph-based patterns based a combination of the eye movement and the head movement of the user; and
interpreting, by the one or more processors of the headgear apparatus, the one or more temporally-correlated graph-based patterns using a machine learning model utilizing at least one of a deep neural network (DNN) and a convolutional neural network (CNN), wherein the machine learning model is trained on labeled vestibular datasets that include VOR suppression responses to the head-fixated target to detect a presence of an abnormality selected from a torsion pattern, a nystagmus pattern, or a vestibulo-ocular reflex (VOR) pattern abnormality.
9 . The method of claim 8 , wherein the plurality of visual stimuli corresponds to one of a smooth pursuit test, saccadic test, optokinetic test, gaze test or a combination.
10 . The method of claim 8 , wherein the plurality of non-visual stimuli corresponds to one of mastoid vibration, caloric, Valsalva, Hyperventilation, Vestibulo-Ocular Reflex (VOR) suppression head shaking, head thrust head impulse or positional testing.
11 . The method of claim 8 , wherein a mixed simulation of visual stimuli and non-visual stimuli is provided to the user.
12 . The method of claim 11 , wherein the mixed simulation includes a real-life projection.
13 . The method of claim 8 , wherein the vestibular assessment includes identifying a vestibular condition associated with the user, wherein the vestibular condition is one of brain diseases, nerve palsy, concussion, Attention-deficit/hyperactivity disorder (ADHD), strokes, early Parkinson's, Multiple Sclerosis or Benign Paroxysmal Positional Vertigo (BPPV).
14 . The method of claim 8 , wherein the headgear apparatus is a pair of augmented reality goggles.