Low-frequency ultrasound, built by biohackers and measured with blood tests
An open, reproducible, self-buildable version of the ~33 kHz immersion-ultrasound device from University of Texas mechanobiology research. The goal is a device that other people can build and calibrate themselves, and then use to measure its biological effects on their own bodies.
- It is a non-commercial DIY biohacking project. We document how to design, build, calibrate and self-test a research device, and we share the data openly.
- It is not a medical device, a treatment or a product. It makes no therapeutic claim, and nothing is sold.
- The evidence so far is preclinical: cells and mice from a single lab. The long-term safety of "reversing" senescence in humans is unknown. Read DIY biohacking: benefits & risks before anything else.
The idea in one minute
In 2022–2025 the late Michael Sheetz's group reported something striking. A weak pulse of low-frequency ultrasound, delivered through water, made senescent human cells start dividing again. Their senescence markers dropped (SA-β-gal, p16/p21, SASP cytokines). Old mice given repeated sessions were far more active, and they had fewer senescence-marker-positive cells in kidney and pancreas. The work is now owned by Mechanobiologics Inc. and is being tested in a sham-controlled human "ultrasound bathtub" trial at UT Health San Antonio (XPRIZE Healthspan).
The physics is surprisingly accessible. The stimulus is about 0.5–4 mW/cm², which is thousands of times weaker than an ultrasonic cleaner. The component family is the same as in cleaning baths. The hard part is not generating 33 kHz. It is proving the dose, meaning a known, uniform, non-cavitating pressure field at the body, measured by someone else's hydrophone and not just by the knob setting.
Project goals
- Reproduce the dose, not the box. Write an open, measurable definition of the exposure: frequency, pressure (with its peak/RMS convention), duty cycle, spatial uniformity and cavitation limits. Define an acceptance test that any build must pass.
- Reference prototype. Build a first calibrated unit using the three hardware architectures we are evaluating, with the lowest cost and skill barrier compatible with a verified dose.
- Make self-built kits easy to adopt. Publish a bill of materials with alternates, open PCB/firmware/3D files, a step-by-step build guide, a shared calibration workflow (a circulating reference hydrophone plus DIY transfer hydrophones) and a public registry of verified builds.
- Measure biology, properly. A self-testing protocol with multiple baselines, a sham phase, and blood and functional markers matched to the UT human trial. See Analyze biological feedback.
- Stay safe and legal. Electrical safety for water immersion, oncogenic-risk awareness, exclusion criteria, and a clear boundary on what a non-commercial DIY project may do under the patents. See IP rights.
From the first build to many self-built kits
Every design decision is judged by one question: can the second, tenth and hundredth builder reproduce the result without us?
1 · Reference build
One unit, fully characterised: a 3-D pressure map, a cavitation check, thermal and electrical safety. Every measurement is published as raw data.
2 · Builder kit
Off-the-shelf parts with EU/CN alternates, open PCB (KiCad), MicroPython/C firmware, 3-D-printable fixtures, a build guide and a tool list. No custom machining.
3 · Verify & register
Each new build runs the acceptance test with a calibrated or transfer hydrophone and publishes its calibration file. Only verified builds contribute biological data.
Sections
Evidence
Papers, trials and parameters, with conflicts between sources shown openly (Pa vs kPa, peak vs RMS, 1.5 vs 1.6 s).
Read →Hardware: 3 architectures
Adapted lab bath · open driver + Langevin transducer · multi-channel array tub. Pros and cons, block diagrams, hardware and software for each.
Read →Hand first
Treat one hand, keep the other as control: a randomised, blinded, sham-controlled n-of-1 that a whole-body tub cannot offer. The measurements, the rig, the protocol.
Read →Using your own bathtub
No purpose-built tub needed — but a battery-powered retrofit kit, a mapping protocol, and a foot-bath stage first. The risks that home use adds.
Read →The cancer branch
The same lab, the same 33 kHz, 20–70× the intensity — and tumour cells die instead of senescent cells restarting. The dose ladder, and why we keep this out of scope.
Read →Open questions for Mechanobiologics
Everything we need to know to reproduce the exposure faithfully.
Read →Analyze biological feedback
Measuring senescence from blood: markers, kits, labs, variability and a before/after protocol.
Read →DIY biohacking: benefits & risks
What might help, what could hurt (including oncogenic risk), and who should not try.
Read →Intellectual property
Which patents exist, what they cover, and what a non-commercial DIY project can and cannot do.
Read →Roadmap
| Phase | Deliverable | Exit criterion | Status |
|---|---|---|---|
| 0 · Reverse-spec | Source analysis, dose definition v0, open questions | Parameters consolidated, conflicts listed | in progress |
| 1 · Calibration first | DIY hydrophone and a transfer calibration against a reference hydrophone; field-mapping gantry | Pressure measured with a stated uncertainty (target ±20%) | planned |
| 2 · Bench prototypes | Architecture A and B built on a 4 L beaker / foot bath | 4–8 kPa uniform to ±25% over the target zone, no inertial cavitation, stable 32–35 °C | planned |
| 2b · Biological cross-check | With a partner lab: the published tumour/normal cell assay run against our rig as a biological dosimeter, plus the oncogenic-safety arm | Measured dose agrees with the biology; no priming of transformed cells at the rejuvenation dose | planned |
| 3 · Builder kit v1 | BOM, PCB, firmware, guide, acceptance test | Two independent builders pass the acceptance test | planned |
| 3a · Hand study | Contralateral-controlled, blinded, crossover protocol on one hand with non-invasive endpoints | A pre-registered result, positive or null, with a documented dose per session | next up |
| 3b · Home retrofit | Battery-powered kit for an existing bathtub, mapped per tub, after the foot-bath stage | Mapped field; insulation and hearing records clean; kit fully removable | planned |
| 4 · Self-testing protocol | Baseline + sham + active phases, biomarker panel, data schema for biohack.it | Protocol reviewed by a physician; first n-of-1 datasets | planned |
| 5 · Array tub (optional) | Architecture C, full-body immersion | Safety case, field map, independent electrical check | later |
Research notes behind this site: 01 source analysis · 02 review of prior ChatGPT analysis.