Technology

The Science Behind Gait Analysis: How PAWSCHECK Detects What You Can't See

January 25, 20267 min readDr. Alastair Greenway, BVM&S, MRCVS, ABVA

Your eyes can't see a 3mm difference in stride length. But that difference matters. Here's how gait analysis turns video into objective data about how your dog moves.

Watch a dog walk past you. Can you tell if their left front leg takes slightly shorter steps than their right? Can you detect a 50-millisecond difference in how long each paw contacts the ground? Can you measure whether their hip drops 2 degrees more on one side?

No. Neither can anyone else. Human observation has limits.

That's not a criticism. It's biology. Our eyes and brains aren't designed to process movement at that level of precision. We're good at noticing obvious limping. We're terrible at detecting subtle asymmetries.

This is where gait analysis comes in. It measures what we can't see, quantifies what we can only sense, and tracks changes too gradual for memory to register.

What Is Gait Analysis?

Gait analysis is the systematic study of movement. In veterinary medicine, it's used to objectively assess how animals walk, trot, and run.

The goal isn't just to say "this dog limps." It's to quantify exactly how they move, measure specific parameters, and detect deviations from normal patterns.

Think of it like this. You can look at a dog and say "something seems off." Gait analysis tells you the left hind limb stride is 12% shorter than the right, stance phase is reduced by 80 milliseconds, and vertical displacement shows compensation in the front limbs.

That's the difference between intuition and data.

Why Objective Measurement Matters

Veterinary assessments of lameness, even by experienced clinicians, show surprising variability. Studies have found that vets disagree on lameness severity in up to 30% of cases during visual examination alone.

That's not because vets aren't skilled. It's because subjective assessment has inherent limitations.

Factors that affect human observation:

  • • Lighting conditions in the exam room
  • • How fast the dog moves
  • • Whether the dog is nervous or relaxed
  • • The observer's experience and attention
  • • Subtle compensations that mask the primary problem
  • • Day-to-day variations in the dog's movement

Objective measurement removes these variables. The same dog, analysed on the same system, produces consistent data regardless of who's watching or where they're examined.

The Biomechanics of Normal Gait

Before we can detect abnormal movement, we need to understand what normal looks like.

The Gait Cycle

When a dog walks or trots, each limb goes through a repeating cycle:

Stance Phase

The paw is on the ground, bearing weight and propelling the body forward.

Swing Phase

The paw is in the air, moving forward to the next contact point.

The ratio between stance and swing time tells us a lot. In normal movement, this ratio is consistent across all four limbs (accounting for slight natural front-to-rear differences).

When a limb hurts, dogs reduce stance phase. Why bear weight on a painful leg longer than necessary? They rush through the stance phase and extend the swing phase. This shows up as measurable timing changes.

Symmetry as a Health Indicator

Healthy dogs move symmetrically. Left and right limbs mirror each other. Front limbs work as a pair. Hind limbs work as a pair.

Temporal symmetry

The timing of left vs. right limb movements should match.

Spatial symmetry

The distance covered by left vs. right strides should be equal.

When symmetry breaks down, something is affecting one side more than the other. This could be pain, weakness, reduced range of motion, or compensation for problems elsewhere.

How Video-Based Gait Analysis Works

PAWSCHECK uses your smartphone video to analyse movement. Here's what happens behind the scenes.

Step 1: Video Capture

You record your dog walking and trotting from multiple angles. Front view, side view, rear view. This gives comprehensive data about movement in three-dimensional space.

The video quality from modern smartphones is more than sufficient. We don't need specialised high-speed cameras. Standard 30 or 60 frames per second captures what we need.

Step 2: Movement Tracking

Computer vision algorithms analyse each frame of the video, identifying and tracking specific points on the dog's body.

Key points tracked:

• Shoulder joints
• Hip joints
• Elbow joints
• Knee (stifle) joints
• Wrist (carpus) joints
• Ankle (hock) joints
• Paw placement points
• Spine position

Step 3: Biomechanical Calculations

Once we know where each body part is in each frame, we can calculate movement parameters:

  • Stride length: Distance between consecutive paw placements
  • Stride time: Duration of a complete stride cycle
  • Stance phase duration: How long the paw stays on the ground
  • Joint angles: The angles at shoulder, elbow, hip, and knee
  • Velocity: How fast the dog is moving
  • Vertical displacement: How much the body rises and falls

Step 4: Symmetry Analysis

Now comes the comparison. Left vs. right. Front vs. hind.

Small differences are normal. Perfect symmetry doesn't exist, even in healthy dogs. But meaningful asymmetries, consistent across multiple strides, indicate something affecting movement.

Step 5: Pattern Recognition

The system doesn't just measure individual parameters. It looks at patterns across all measurements.

Machine learning models, trained on thousands of gait videos correlated with veterinary diagnoses, help identify these patterns. But the system doesn't diagnose. It detects altered movement and quantifies the changes.

What Gets Measured (And Why It Matters)

Stride Length Asymmetry

This is one of the most revealing measurements. When a limb hurts, dogs shorten the stride on that leg.

Normal variation
Within 5%
Mild asymmetry
5-10% difference
Moderate asymmetry
10-20% difference
Severe asymmetry
>20% difference

Example: A dog with a right front limb stride of 45cm and a left front limb stride of 40cm shows 11% asymmetry. That's significant, even though both strides look roughly similar to the human eye.

Your Dog as Their Own Baseline

Here's a crucial concept in gait analysis. We're not comparing your Labrador to an "average" Labrador. We're comparing your Labrador to themselves.

Breed variations in normal gait are enormous. A Dachshund moves completely differently from a Greyhound. Even within breeds, individual variation is substantial.

So the first video you record establishes YOUR dog's baseline. Their normal movement pattern.

Future videos get compared to that baseline. We're looking for changes from THEIR normal, not deviations from some universal standard.

This makes gait analysis incredibly powerful for longitudinal tracking. A dog who scores "4" on their first video might just have naturally asymmetric movement (old injury, anatomical quirk). That's their baseline 4. If they later score 7, that's a meaningful change.

The Limits of Gait Analysis

Let's be clear about what gait analysis can and cannot do.

What it can do:

  • Detect movement asymmetries objectively
  • Quantify severity of gait changes
  • Track changes over time
  • Identify which limb(s) show altered movement
  • Monitor response to treatment

What it cannot do:

  • Diagnose specific conditions
  • Replace physical examination
  • Determine the cause of lameness
  • Substitute for diagnostic imaging
  • Tell you what treatment is needed

Think of gait analysis as a highly sensitive thermometer. It tells you the temperature has changed. It doesn't tell you why, or what medication to take. That's what vets are for.

Practical Applications in Everyday Life

How does this actually help you and your dog?

Establishing a Baseline

Record your dog while they're healthy. This creates a reference point for detecting future changes.

Monitoring Chronic Conditions

Regular gait analysis tracks whether management strategies for arthritis, hip dysplasia, or other conditions are working.

Post-Surgical Recovery

Objective monitoring shows whether recovery is progressing normally or if problems are developing.

Early Detection

Catch problems early. Before you notice limping. When intervention is most effective and least costly.

The Bottom Line

Gait analysis transforms subjective observation into objective measurement. It catches what eyes miss, quantifies what intuition senses, and tracks what memory forgets.

It's not magic. It's applied biomechanics and computer vision, combined with veterinary science.

The technology measures stride length in millimeters, timing in milliseconds, and angles in degrees. It compares thousands of data points to establish patterns. It detects asymmetries too subtle for human perception.

But at the end of all that sophisticated analysis, it answers simple questions:

  • • Is my dog moving normally for them?
  • • Has something changed?
  • • Is treatment working?
  • • Should I be concerned?

Those are the questions that matter. Gait analysis helps answer them with data instead of guesswork.