patent.fyi
Applications  /  US 2026/0354680
United States Patent Application Publication

US 2026/0354680 A1

Published
Nov 12, 2026
Filed
Sep 4, 2024
Earliest date on the face
Sep 7, 2023
Classification
A61B 3/10
First published here
Nov 19, 2026

Generated automatically. Not reviewed by a person. Not legal advice or an opinion.

Section 1 · machine-generated

What if?

Questions about how individual parts of claim 1 might be changed. A question may describe something different from what the claim language recites, or a version of it; no view is offered on which.

1.1One frequency swept beam

What ifWhat if a single terahertz beam illuminates one corneal spot while the source sweeps in frequency?
Claim language“multiple Terahertz beams”

A tunable source and a focusing lens deliver one beam along a fixed optical path. A coherent receiver records the returned amplitude and phase at each frequency, and a calibrated dielectric model estimates water fraction from the frequency response.

1.1Through transmission of an explant

What ifWhat if a receiver array measures terahertz beams transmitted through an excised cornea?
Claim language“receive backscattered versions”

A sealed sample holder supports the excised cornea between thin polymer windows, with emitters on one side and detectors on the other. An empty holder provides a transmission reference, and a dielectric model uses measured corneal thickness and frequency-dependent attenuation to estimate water fraction.

1.2Pulse timing per channel

What ifWhat if the processor measures the separation between interface echoes independently in each return?
Claim language“determine joint statistics”

A pulsed terahertz receiver records the front and rear interface echoes from a corneal specimen thin enough for both echoes to resolve. An optical thickness measurement and the echo separation give the terahertz group index for that location. A calibration maps each group index separately to a local water fraction.

1.2Fixed reference for each detector

What ifWhat if each detector channel produces its own calibrated reflectance reading from a fixed reference?
Claim language“determine joint statistics”

A polished metal reflector supplies an amplitude reference for each channel during calibration. During measurement, the processor divides each return amplitude by its stored reference amplitude and converts the resulting reflectance to a local water fraction using that channel's calibration curve.

1.3Water fraction by mass loss

What ifWhat if the instrument determines water fraction from the mass lost when an excised cornea dries?
Claim language“based on the joint statistics”

A microbalance records the specimen mass after surface liquid is blotted and again after vacuum drying to constant mass. The processor subtracts the dry mass from the initial mass and divides the difference by the initial mass to report water fraction.

Section 2 · machine-generated

Variations

Further arrangements of the same subject matter, described here as of the publication date at the foot of this page.

Other field of use

Moisture measurement on a paper web

A row of terahertz heads projects beams onto adjacent spots across a moving paper web and collects backscatter from each spot. A roller encoder triggers the heads together so that each set of returns represents one transverse strip. A processor computes the covariance of the frequency-resolved returns across beams over a moving window of strips. A calibration based on gravimetric measurements and paper basis weight maps those statistics to web moisture content.

Component substitution

Electro-optic sampling of corneal backscatter

A zinc telluride electro-optic sampling assembly replaces the terahertz receiver. A pulsed terahertz beam illuminates successive corneal spots, and an off-axis parabolic mirror directs each return onto the crystal. A synchronized optical probe samples the electric field at stepped delays, and balanced photodiodes record the field-induced polarization change. A processor calculates cross-beam spectral covariance from the reconstructed waveforms and maps it to corneal water fraction using calibrated phantoms.

Parameter or scale variation

Eight-beam acquisition on two concentric rings

A beam-steering mirror directs a 300 GHz beam to eight corneal spots in sequence. The spot centers are spaced at 90-degree intervals on rings with radii of 1 mm and 2 mm in the plane perpendicular to the instrument axis. The receiver acquires each backscatter return in a 10 ms window and repeats the eight-position scan 16 times. A processor computes cross-beam covariance from the repeated scans and applies a calibration derived from phantoms with known water fractions.

Combination with existing systems

Corneal measurement with optical coherence tomography

A corneal optical coherence tomography scanner and a terahertz backscatter head share a patient support and fixation target. A curved calibration target establishes the coordinate transformation between the optical volume and the terahertz beam positions. The processor uses the optical surface map to assign a local incidence angle to each terahertz beam before calculating covariance across the corrected returns. It then uses the measured corneal thickness as a separate input when mapping the joint statistics to tissue water fraction.

Automated or data-driven variant

Eye tracking gates terahertz beam acquisition

An infrared camera measures the pupil center and corneal reflection during each terahertz sweep. A controller compares both positions with a reference frame, discards sweeps that exceed a calibrated motion threshold, and reacquires the affected beam positions. A processor calculates cross-beam spectral covariance after each beam position has the required number of accepted repeats. A calibration model converts the resulting statistics into a corneal water fraction estimate.

Regression estimates corneal tissue water fraction

A multibeam terahertz head records frequency-resolved backscatter from each corneal spot and periodically records a reference reflector to correct gain drift. Software extracts cross-beam covariance of log power and pairwise spectral coherence from repeated sweeps. A gradient-boosted regression model trains on those features using curved phantoms with gravimetrically measured water fractions and varied thicknesses. During measurement, the model maps the same features from the corneal returns to an estimated water fraction.

Specific version of a part

Four beams at separate angles

A terahertz beam splitter and four steering mirrors direct four beam paths toward the same corneal region at separate incidence angles. A mirror in each path collects the return and focuses it onto a dedicated receiver.

Covariance of return amplitudes

The processor groups synchronized amplitude readings from repeated acquisitions into one vector per acquisition. It subtracts the mean for each channel and averages the products of paired deviations to form a covariance matrix.

Regression calibrated by tissue mass

A ridge regression uses the variances and covariances of return amplitudes as inputs and the gravimetrically measured water fractions of calibration specimens as targets. The processor applies the fitted coefficients to statistics from a measured cornea and reports an estimated water mass fraction.

Section 3 · quoted from USPTO records

Other work

Passages reproduced word for word from two earlier-published U.S. patent documents, placed beside the language of claim 1. No characterization is offered. Each passage is marked A or B for the document it comes from; ¶ numbers are the paragraph numbers in the USPTO’s published full text of that document.

Document AUS 8,267,517 B2
Published
Sep 18, 2012
Filed
Nov 7, 2008
Class
A61B 3/14
Document BUS 8,690,331 B2
Published
Apr 8, 2014
Filed
Aug 29, 2012
Class
A61B 3/10
Claim 1
Document A
Document B
1.1illuminate a cornea using multiple Terahertz beams and receive backscattered versions of said multiple Terahertz beams;
By the lighting unit 2 irradiating the cornea 101 with infrared light L1, the infrared light L1 is reflected on the cornea 101 to generate a corneal reflection light image L2.
A · US 8,267,517 B2 · ¶ 28
The corneal hydration sensing system 100 includes an illumination system 102 configured to provide an illumination beam of terahertz radiation 104, and an optical system 106 arranged in an optical path of the illumination system 102 to relay and direct at least a portion of the illumination beam of terahertz radiation 104 onto a cornea 108 of a subject and to receive at least a portion of terahertz radiation reflected from the cornea 108 to provide a return beam of terahertz
B · US 8,690,331 B2 · ¶ 19
1.2determine joint statistics of the backscattered versions of said multiple Terahertz beams; and
1.3determine information about water content of the cornea based on the joint statistics.
The signal processing system 116 processes the detection signal 114 to provide a measure of an amount of hydration sensed in the cornea 108 of the subject.
B · US 8,690,331 B2 · ¶ 19

An empty cell means only that no passage was reproduced there.

Statements in these documents
However, since tremor is very small high-frequency oscillation with an amplitude on the order of 1 μm, it is difficult to accurately detect tremor as compared with drift and flick (microsaccade).
A · US 8,267,517 B2 · ¶ 8
The present invention has been made in view of the above-mentioned problems, and an object thereof is to provide an eye movement measurement apparatus capable of accurately detecting a tremor component.
A · US 8,267,517 B2 · ¶ 9
Therefore, there remains a need for improved systems for sensing corneal hydration.
B · US 8,690,331 B2 · ¶ 8
The signal processing system processes the detection signal to provide a measure of an amount of hydration sensed in the cornea of the subject.
B · US 8,690,331 B2 · ¶ 9
Cite aspatent.fyi, US 2026/0354680 A1, claim 1, first published 2026-11-19T02:03:35-05:00, https://patent.fyi/US20260354680/, SHA-256 b7612eefcf96813c6a79e177cfedc5e716a4287ad112e7c5c80151f07a0d8ce8
First published
Permanent addresshttps://patent.fyi/US20260354680/
Content fingerprint (SHA-256)b7612eefcf96813c6a79e177cfedc5e716a4287ad112e7c5c80151f07a0d8ce8
Methodv2026-11-18.2 · automated, not reviewed by a person
Correctionscorrections@patent.fyi