Beamforming

Steering light
you cannot see.

Ultrasound has no lens. Instead, an array of elements fires and listens with tiny time delays — and constructive interference creates a focused beam. The image is not captured; it is computed from wave mathematics.

Δt = d · sinθ / c
Phase Delay per Element
A(x,t) = Σ aₙ · e^i(ωt−kxₙ)
Array Element Sum
X(ω) = ∫ x(t) e^−iωt dt
Fourier Transform
Beamforming (left) — Interference Pattern (right)
Number of elements 12
More elements → finer lateral resolution
Steering angle 0°
Phase delays steer the beam without moving the probe
Focus depth 3.5 cm
Delays curved to converge at focal point
Element spacing 1.5 λ
λ/2 spacing minimizes grating lobes
🎯
Constructive vs Destructive

Where wavefronts arrive in phase, they amplify. Where they arrive out of phase, they cancel. Beamforming is simply engineering the delays so that constructive interference happens exactly where you want to look.

📡
Lateral vs Axial Resolution

Axial resolution (along the beam) is determined by pulse length — shorter pulses = better. Lateral resolution (side-to-side) is determined by beam width — which is where focusing and element count matter most.

🔢
Fourier at Every Step

The Fourier transform converts between time-domain echo signals and frequency-domain information. It's used in Doppler processing, noise filtering, and the reconstruction mathematics that turn raw signals into beamformed images.

📡
SONAR → RADAR → Medicine

Beamforming was developed for military SONAR and RADAR during WWII. The mathematics translating these to medical ultrasound in the mid-20th century is one of the clearest examples of dual-use physics in history.