Steering sound with math: how Open-LIFU's 2D array focuses ultrasound
A single ultrasound transducer emits a cone of sound that spreads in every direction. Useful for imaging, but too diffuse for neuromodulation — if you want to stimulate a precise structure deep in the brain, you need energy concentrated at an exact point in three-dimensional space. That's the core problem Open-LIFU is built to solve.
At the heart of the Open-LIFU platform sits a wearable headset housing one or two 64-element 2D matrix transmit modules. Each of those 64 elements is independently addressable. Each fires at a calculated delay. Together, they do something no single transducer can: bend sound to a point.
The physics of constructive interference
When two wavefronts arrive at the same point at the same time, they add together. When they arrive half a wavelength out of phase, they cancel. Open-LIFU exploits this: by firing all 64 elements with individually tuned delays, every wavefront is arranged to arrive at the target simultaneously. Amplitudes stack — constructive interference — producing a hot spot of acoustic pressure far more intense than any individual element could generate alone. Everywhere else, the waves cancel or disperse harmlessly through tissue.
From a line to a grid
A 1D array can steer and focus a beam, but only within a single plane — you get control in two dimensions, not three. Open-LIFU's 2D matrix layout changes that entirely. With elements arranged in a full grid, the focal point can be positioned anywhere in the volume in front of the transducer — left, right, up, down, near, far — without moving the headset at all.
Steering by delay
The key variable is when each element fires. To focus at a target, the beamformer calculates the distance from that target to every element in the 8×8 grid. Elements that are farther away fire first; elements that are closer fire later. The delays are chosen so that every wavefront arrives at the target simultaneously. Open-LIFU's beamformer runs on a 10 MHz clock for fine-grained phase control, giving sub-microsecond timing precision across all 64 channels. .svg)
Want to move the focus deeper? Increase delays uniformly around the perimeter and decrease them toward the center — the classic converging lens pattern. Want to steer laterally? Shift the delay map in that direction. These profiles are computed and updated in software, meaning the focal point can be repositioned electronically between pulses without touching the hardware.
Every parameter, fully programmable
Every parameter in a sonication sequence is fully programmable — pulse frequency, focal pressure, duration, repetition interval, and train count — giving researchers exact reproducibility across sessions and sites. The system supports a programmable focal pressure range of 0–1200 kPa, and in dual-module configuration achieves a peak negative pressure of 2.16 MPa with an axial steering range of 3–11 cm.
Why 2D matters clinically
With a 1D array, a researcher must physically reposition the transducer to reach targets outside its focal plane. Open-LIFU's 2D matrix eliminates that constraint. The focal point can be steered across a volume electronically — useful for targeting irregular anatomical structures or adapting to subject-to-subject variation without re-seating the headset.
This capability pairs directly with Open-LIFU's spatial localization system. An embossed faceplate pattern enables photogrammetric 3D localization via a standard Android phone, and that mesh feeds directly into the treatment-planning software, letting you target with anatomical precision before the first pulse is fired. The result is a closed loop: anatomy in, delay profile out, focused energy on target.
Open, documented, and ready to extend
Every hardware schematic, firmware file, and software layer is published and community-licensed — so the beamforming logic itself is transparent and auditable, not a black box. Whether you're validating a protocol, building a novel neuromodulation study, or extending the platform for a new application, the 2D array's delay architecture is yours to understand and modify.
The result is a tool that behaves less like a blunt instrument and more like a scalpel made of sound: precise, steerable, reproducible, and entirely non-invasive.