AuraSonics
02 · The Solution

We turned the chest
into an instrument.

AuraSonics is not a better stethoscope. It is a new sensing modality — a soft wearable that listens to the heart from three thousand points at once, captures the energy a microphone can't feel, and hands a physician an objective, spatially mapped, AI-readable picture of the cardiac field.

A new category: Wearable Cardiac Tomography.

Acoustic + mechanical 3,000 sensors 1–3,000 Hz 3D source localization Patient digital twin
The Moat
The Core Inventive Idea

The heart makes sound.
It also makes motion.

Every auscultation tool before it captured acoustic pressure, typically at a single point. AuraSonics captures two physically different energies at the same instant — through two complementary sensor types in one belt.

The failing heart often moves before it sounds — and a microphone on the chest can't feel that motion, because it rides along with it. We capture both.

Type A

Fine acoustic

Heart sounds, murmurs, valve vibration, turbulent-flow signatures — high-resolution pressure waves.

Type B

Gross mechanical

Systolic thrust, chest-wall recoil, ventricular filling, heart-rocking — the physical movement a mic cannot feel.

Captured simultaneously. Across the whole torso.
Density Is the Advantage

From one point
to three thousand.

A stethoscope hears a single local projection of the heart. AuraSonics wraps the torso in a soft belt of 20 rings × 150 patches = 3,000 individually addressable sensors, each reporting where, when, and how the chest wall moves and sounds.

Each ring: 75 Type A · 75 Type B  |  alternating, coordinate-aligned, individually addressed
Stethoscope
1 point
vs
AuraSonics
3,000 points
3,000
Addressable sensors across the torso
1–3,000 Hz
Mechano-acoustic sensing range
1–5 µm
Displacement sensitivity per patch
>70 dB
Signal-to-noise — engineering target
baseline f₀shifted f₀ + Δf →
a femtofarad change in capacitance  →  a kilohertz shift in frequency
Amplification at the Physics Level

We don't fight the noise.
We outrun it.

Each patch is a tuned LC resonator. A faint physiological event produces a vanishingly small capacitance change that ordinary electronics would lose in noise. Our resonator converts it into a large, clean frequency shift — gain built into the physics, before a single line of software runs.

We never amplify a sound — there is nothing to amplify. Motion shifts each resonator's frequency; a digital counter simply counts the shift, resolving every event into a compact 16-bit signature.

We don't listen to the heart.
We measure it.

Every patch is tuned to its own frequency, so the entire array is read at once — wirelessly, crosstalk-free. One sweep. Three thousand answers.

It Doesn't Just Detect — It Locates

We can tell you where
the abnormality lives.

With sound arriving at 3,000 known coordinates, the AI solves the inverse problem: it triangulates the origin of a signature through tissue-aware propagation and renders it on a patient-specific thoracic map — to a modeled accuracy of 1–3 mm, enough to distinguish mitral from tricuspid, aortic from pulmonic.

A signal can say something is wrong. A field is designed to say where.

orientation inferred source extent
The Personal Benchmark

Every patient, measured against a personalized healthy-reference baseline.

AuraSonics builds a digital twin — an age-, sex-, and anatomy-matched simulation of a healthy thorax, computed patch-by-patch across all 3,000 locations. The patient's live reading is subtracted from it.

What remains is a pure deviation map — the exact signature of what is different, where, and by how much. This is the objective, reproducible reference the stethoscope never had.

Healthy digital twin This patient ▨ deviation = the finding
The Intelligence Layer

Clean signal in.
Physician-reviewable insight out.

AuraSonics pairs networks that read the texture of each signal with networks that model how energy propagates across the array — then fuses it with labs and history into a probability-weighted, physician-reviewable assessment.

Capture

L · C · R streams

3,000 patches, two energy types, microsecond-resolved.

CNN

Texture

Spectrogram features — S1–S4, murmurs, crackles.

GNN

Propagation

Array as a graph — spatial spread & source paths.

Fusion

Context

+ labs, history, prior baselines → probability.

Output

Physician map

Spatio-spectral hotspots, confidence, review.

Every score is traceable to the patches, frequencies, and timing that produced it — and a clinician always reviews and approves. Explainable by design, which is exactly what regulators want to see.

Across a Lifetime

Not a snapshot. A trajectory.

Coordinate-aligned, precisely time-stamped scans let the same heart be compared against itself across months and years — tracking disease progression, treatment response, and post-procedure healing. The stethoscope forgets every visit. AuraSonics is built to remember all of them.

Therapy-Response Monitoring

The body cannot
be deceived.

Every medicine is a molecule, and every molecule that reaches the bloodstream changes the heart — the force of its contraction, the behavior of its valves, the timing of its rhythm. Each change leaves a mark. Every mark can be read.

So we hold each patient against their perfect twin — the same heart, rendered in flawless health by the same three-thousand-point field — and we measure the distance. We call it the delta. Then we watch which way it moves. No marketing claim, no placebo, no wishful thinking survives that comparison.

"Whatever enters the bloodstream returns a verdict."

PERFECT TWIN · Δ = 0 DISTANCE FROM IDEAL WEEKS OF THERAPY Δ WIDENS — IT IS DOING HARM Δ CLOSES — IT IS WORKING FIRST SCAN · BASELINE Δ
1
Everything leaves a signature

A drug, a supplement, a lungful of clean air or of smog — nothing the body takes in leaves the heart unchanged, and nothing escapes the field.

2
The twin sets the standard

Before any compound has time to act, the first scan already shows how far the real heart stands from its own ideal — the baseline delta.

3
The delta returns the verdict

Week by week the portrait moves. A therapy that helps closes the delta; one that harms widens it. From that motion alone: benefit or harm.

Illustrative and hypothetical — development-stage, not clinically validated. The same twin-and-delta principle extends beyond the acoustic field: every modality we add carries its own standard of perfect health.

The Clinical Wedge

The interval
no one is watching.

Most aortic dissections strike below the size guidelines operate on — and the scans that watch these patients come months to years apart. Nothing observes the space between them. The degeneration is silent; too often, the first symptom is the catastrophe itself.

AuraSonics is designed for exactly this interval: continuous, self-referenced surveillance where a trajectory beats a snapshot — a rising deviation against the patient's own baseline, crossing an action threshold in time to act.

“The belt earns the CT; the CT earns the surgery.”

1
Detect

A rising, non-transient deviation trajectory crosses the action threshold — automated and probability-weighted.

2
Flag

A clinician-reviewable alert: “high, and rising — recommend confirmatory imaging.”

3
Confirm

Definitive testing is ordered — CT angiography, transesophageal echo, and/or MRI.

4
Decide

Imaging confirms or refutes a high-risk aneurysm. No one is asked to act on an acoustic signal alone.

5
Repair

If confirmed and indicated — an elective repair, before rupture.

Illustrative and hypothetical — development-stage, not clinically validated. AuraSonics is a triage-and-timing layer that routes the right patient to definitive imaging; it never replaces it. Our standard is validation-first.

Why the Incumbents Can't Follow

A new category —
not a better version.

The best-funded names still optimize the same narrow signal: acoustic energy, at one or a few points. Eko raised over $150M to connect a single sensor to AI — one energy, where AuraSonics captures two.

And they share one blind spot by physics: a sensor that rides the chest cannot feel the heart's motion. AuraSonics resolves the motion they ride over.

Acoustic stethoscope
"Smart" mic patches
(Eko, et al.)
AuraSonics
Energy captured
Acoustic only
Acoustic only
Acoustic + mechanical
Capture points
1
1–few
3,000
Spectrum
~20–200 Hz
~20–1,000 Hz
1–3,000 Hz
Spatial map
None
None
3D · 1–3 mm target
Healthy baseline
Human memory
Population model
Patient digital twin
Output
An opinion
A flag
A mapped, explainable field

Comparison reflects AuraSonics' engineering targets against publicly available specifications of commercially deployed devices. AuraSonics is development-stage and not yet clinically validated.

They are improving auscultation. We are replacing it with a measurement.

The Defensible Core

Five patent applications.
Three families. One platform.

The advantage isn't a single trick — it's an interlocking architecture, each layer hard to invent and harder to design around. Cardiology is the beachhead; the same sensing field extends to pulmonary, vascular, and beyond.

Family 01

Multi-modal sensing architecture

Dual-channel Type A / Type B LC-resonant patches, frequency-division readout, and high-density spatial sensing across the torso.

Family 02

Real-time AI diagnostic pipeline

CNN + GNN analysis of heart-sound components and propagation, generating spatio-spectral maps and abnormality inference.

Family 03

Digital-twin data fusion

Patient-specific healthy baseline, deviation mapping, and synthesis with clinical data into a longitudinal, treatment-aware record.

The Verdict

A defensible new category.
Now we build it.

AuraSonics is a development-stage company. The technology described is the subject of pending provisional patent applications; performance figures such as >70 dB SNR and 1–3 mm localization are engineering and modeled targets, not yet clinically validated. AuraSonics products are not FDA-cleared and are not available for clinical use.