patent.fyi
Applications  /  US 2026/0354716
United States Patent Application Publication

US 2026/0354716 A1

Published
Nov 12, 2026
Filed
Feb 13, 2026
Earliest date on the face
May 9, 2025
Classification
A61B 5/1455
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.

No entries.

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

Raman monitoring in fermentation lines

A fiber-coupled Raman probe focuses through a sapphire window into a flow cell on a fermentation recirculation loop. The spectrometer sends each spectrum to a processor for dark subtraction, wavenumber alignment, and fluorescence baseline correction. A regression model trained on spectra paired with reference assays estimates glucose and ethanol concentrations and logs them with the batch identifier and acquisition time.

Component substitution

Partial least squares biomarker estimation

A partial least squares regression model replaces the general-purpose analysis engine for concentration estimation. The processor resamples baseline-corrected spectra onto a common wavenumber grid and applies the centering values stored during calibration. Separate models map the processed spectra to individual analyte concentrations using calibration samples with independently measured concentrations. The software flags spectra whose reconstruction residual exceeds a threshold set on a held-out calibration set.

Parameter or scale variation

Depth resolved acquisition in an eye phantom

A motorized focus group moves the Raman sampling volume through five axial positions spaced 0.25 mm apart in a model anterior chamber filled with a biomarker solution. The control software records three 200 ms exposures and one shutter-closed background frame at each position. It subtracts the background frame, takes the median of the three corrected exposures at each detector pixel, and stores each spectrum with its calibrated focal depth. The resulting depth profile shows how the fitted analyte signal changes near the chamber window.

Combination with existing systems

Raman targeting with optical coherence tomography

An anterior-segment optical coherence tomography scanner and a confocal Raman probe share a front objective through a wavelength-selective beam combiner. A model eye with fiducials at known depths establishes the transformation between tomography coordinates and the Raman focus. Segmentation software locates the posterior corneal surface and the anterior lens surface in the tomographic image. The controller uses those boundaries to place the Raman focus at a selected point within the anterior chamber and stores the focus coordinates with the spectrum.

Automated or data-driven variant

Eye motion gating for raman acquisition

A near-infrared camera shares the probe axis through a beam combiner and supplies time-stamped images to a pupil and corneal reflection tracker. A calibration target maps image displacement to lateral displacement at the Raman sampling plane. The acquisition controller enables illumination and starts a detector frame when tracking confidence and position meet configured limits. It discards any frame in which the tracker reports motion beyond the position limit during integration.

Automated correction of spectrometer drift

An optical switch selects either the eye probe or a sealed reference channel containing a polystyrene coupon, with both channels feeding the same spectrometer. The controller records a reference spectrum before each measurement session and fits a wavelength correction from the measured positions of multiple reference peaks. The processor applies the correction to subsequent sample spectra before interpolation onto the analysis model's wavenumber grid. The software requests recalibration when the reference peak residuals exceed a stored tolerance.

Cite aspatent.fyi, US 2026/0354716 A1, claim 1, first published 2026-11-19T02:47:19-05:00, https://patent.fyi/US20260354716/, SHA-256 193ee98ef0eafac9899deaa43a4fc33b3fcdcf92ff7e3b1f7a624873eda757b9
First published
Permanent addresshttps://patent.fyi/US20260354716/
Content fingerprint (SHA-256)193ee98ef0eafac9899deaa43a4fc33b3fcdcf92ff7e3b1f7a624873eda757b9
Methodv2026-11-18.2 · automated, not reviewed by a person
Correctionscorrections@patent.fyi