Autonomous Robotic Catheter Navigation

Lesion Target

Magnetic Steering

Field strength (steering gain k) 4.0

Insertion speed 0.55 units/s

Blood-flow turbulence 0.30

Navigating…

Live Telemetry

Distance to target

— mm

Lateral deviation

— mm

Wall contacts

How it works

The catheter tip is steered by an external magnetic field, exactly as in real robotic navigation systems (e.g. magnetic EP/neurovascular catheters). The autopilot always knows the correct sequence of branches to the chosen lesion — the physics being simulated is how precisely the tip can be held on that centerline while blood flow tries to push it into the vessel wall.

At every instant the lateral offset x from the vessel centerline obeys a driven spring–damper equation:

a = -k·x - c·v + F_turbulence(t)
v += a·dt
x += v·dt

where k is the magnetic steering gain (your slider — a stiffer field pulls the tip back to center faster), c is fixed viscous damping from blood and vessel friction, and F_turbulence is a pulsatile disturbance scaled by the turbulence slider. When |x| reaches the local vessel radius, the tip strikes the wall: velocity along the wall normal is zeroed (inelastic contact), the contact force k·|x| is logged, and the strike is counted.

Field strength — raise it for tighter centerline tracking and fewer wall strikes; too low and turbulence pushes the tip into the wall, especially through bifurcations.
Insertion speed — faster advance covers the branch tree quicker but gives the steering loop less time per bifurcation to damp out lateral drift.
Turbulence — models pulsatile blood flow disturbing the tip; a "perfect" run keeps deviation near zero and wall contacts at zero all the way to the target.

Real-world relevance: this is the same idea behind magnetically-navigated robotic catheter systems used in electrophysiology and neurovascular intervention — an external field bends a soft catheter tip through tortuous vasculature to reach a lesion with minimal trauma to the vessel wall.

Catheter tip & path Target lesion Vascular tree

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.