Questions: Optic Flow and Navigation in Medial Superior Temporal Area

Question 1

A person walks forward down a hallway. How does MST compute their heading direction from the visual input?

• A. It tracks the fastest-moving object in the visual field and uses its trajectory as a heading cue
• B. It locates the focus of expansion — the central point from which all visual elements radiate outward — in the global optic flow field
• C. It averages the individual motion vectors detected by MT neurons across the visual field
• D. It relies primarily on vestibular signals from the inner ear, using visual input only as a cross-check

Answer: (B)

The geometric key to heading perception is the focus of expansion: when you move forward, the entire visual field expands outward from a single point that corresponds precisely to where you are heading. MST neurons have large receptive fields and respond selectively to global flow patterns (expansion, contraction, rotation), allowing them to locate this focus. Simply averaging MT vectors (option C) would not isolate the focus; tracking individual objects (option A) is a different strategy; vestibular input (option D) is supplementary, not primary.

Question 2

A patient with damage to the right MST has impaired smooth pursuit eye movements. In which direction is pursuit most affected?

• A. Leftward — contralateral to the lesion, consistent with crossed projections
• B. Rightward — ipsilateral to the lesion, consistent with MST's role in monitoring visual motion during gaze
• C. Both horizontal directions equally, since pursuit requires bilateral MST coordination
• D. Vertical directions only, since horizontal pursuit relies on the frontal eye fields

Answer: (B)

MST lesion studies show that smooth pursuit is impaired specifically in the ipsilateral direction — rightward for a right MST lesion. This differs from many other visual cortical areas that show contralateral deficits, and reflects MST's role in processing ipsilateral motion during gaze. This finding confirms MST's dual role in both self-motion perception and gaze control.

Question 3

MST neurons process mainly visual motion signals and receive no input from non-visual sensory systems.

• T. True
• F. False

Answer: False

MST also receives vestibular input from the inner ear. This multisensory convergence is functionally significant: when you actually move through the world, optic flow and vestibular signals should agree. When they conflict — such as when watching a large-screen movie where your body is stationary but your visual field mimics forward motion — the mismatch between visual and vestibular signals can produce motion sickness. MST's integration of both modalities is central to coherent self-motion perception.

Question 4

The focus of expansion in optic flow is located at the point in the visual field toward which the observer is heading.

• T. True
• F. False

Answer: True

This is the core geometric relationship between optic flow and heading. When you move forward, visual elements in every direction stream away from a central point — this is the focus of expansion, and its location in the visual field directly specifies heading direction. MST neurons compute global optic flow patterns over large receptive fields to identify this focus, enabling the brain to derive heading without requiring knowledge of absolute speed or distance traveled.

Question 5

Why can MT alone not provide information about the observer's heading direction, and what computation does MST add?

Model answer: MT neurons respond to local motion — direction and speed in small patches of the visual field. Local motion signals cannot distinguish self-motion from world-motion, and do not reveal the global geometric structure of optic flow. MST integrates MT's local signals across large spatial scales, responding to global flow patterns like radial expansion and contraction. By identifying where the focus of expansion falls in the visual field, MST extracts heading direction — information that no local motion signal can provide on its own.

This is the core hierarchical computation: MT extracts local velocity; MST extracts global structure from those local signals. The analogy is how V1 edge detectors feed into higher areas that extract shapes — each stage derives increasingly abstract properties. Heading perception requires the global level because the diagnostic signal (focus of expansion) is a property of the entire flow field, not any single point in it.