patent.fyi
Applications  /  US 2026/0354911
United States Patent Application Publication

US 2026/0354911 A1

Published
Nov 12, 2026
Filed
Feb 6, 2026
Earliest date on the face
Mar 17, 2025
Classification
A61B 34/30
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.

1.1Factory calibration and fixed rules

What ifWhat if the system uses factory calibration and fixed geometric rules with no training dataset?
Claim language“obtain and store training data”

A technician measures the tool center point and joint offsets with a calibration fixture and stores the resulting coefficients. The controller uses those coefficients and a rigid transform from current fiducial measurements to map a preset path into robot coordinates.

1.2Hand motion through passive linkage

What ifWhat if the apparatus uses a hand-driven linkage with no actuator?
Claim language“at least one actuator”

The operator moves a handle that transmits force through a rigid parallelogram linkage to the instrument holder. Mechanical stops limit travel, and a passive counterweight balances the instrument arm.

1.2Fixture guided tool travel

What ifWhat if a rigid fixture constrains instrument travel and the apparatus uses no sensors during the procedure?
Claim language“sensors configured to generate real-time procedural data”

A subject-specific fixture seats against a rigid anatomical surface and aligns a guide sleeve with a preplanned entry axis. The sleeve constrains lateral motion, and a collar on the tool shaft contacts the sleeve at the preset depth.

1.3Direct sensor signal path

What ifWhat if live measurements pass directly into the motion circuit with no real-time data pool?
Claim language“a real-time data pool”

A strain-gauge bridge feeds a conditioned voltage directly to an analog comparator. When the voltage crosses a preset threshold, the comparator opens the motor-enable circuit without recording the measurement.

1.4Live landmark geometry

What ifWhat if the controller computes the procedure path from current landmarks and instrument geometry alone?
Claim language“comparisons between the real-time procedural data and the training data”

A stereo camera measures three noncollinear fiducials to establish a local coordinate frame. The controller transforms predefined waypoints into robot coordinates and checks the resulting path against joint and travel limits.

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

Robotic finishing of cast housings

A fixture holds a cast housing while a robot carries a rotary abrasive tool and a wrist force sensor. The controller compares a live surface scan with archived scans paired with successful finishing paths for the same casting geometry. It aligns the selected path to the current housing and adjusts tool pressure using the wrist force measurements.

Component substitution

Case retrieval with dynamic time warping

A dynamic time warping matcher replaces the trained model used to select a procedure segment. The archive stores synchronized force and instrument displacement traces paired with procedure segments, and the controller compares an initial live trace with archived traces from the same procedure phase. It ranks the cases by alignment cost, checks the selected segment against current workspace limits, and sends the segment to the motion controller.

Parameter or scale variation

Separate planning and force control rates

A bench system uses a motorized stage with 25 mm of travel, a 0 to 5 N load cell, and an elastomer specimen. The controller samples the load cell and updates motor commands at 1 kHz, while a planner compares the latest 200 ms of force and displacement samples with archived bench traces at 10 Hz. The planner updates the displacement target, and the motor controller retracts the instrument when the measured force exceeds 2 N.

Combination with existing systems

Procedure planning with intraoperative ultrasound

An intraoperative ultrasound console streams frames from a probe mounted on a second robotic arm, and the controller timestamps each frame with the corresponding arm position. A calibration transform maps image coordinates into the instrument arm's base frame. The planner compares live tissue boundaries with archived image and instrument trajectory pairs, then uses the mapped boundaries to update the instrument path.

Automated or data-driven variant

Automatic recovery from sensor disagreement

A stereo camera and actuator encoders independently estimate the instrument tip position, and the controller stores their time-aligned estimates in the live data pool. A monitoring process compares the difference between those estimates with archived differences for the current motion phase. Persistent disagreement pauses instrument motion and triggers a fresh image acquisition, after which the controller checks agreement before issuing the next motion segment.

Automatic indexing by procedure phase

A hidden Markov model labels archived sensor sequences by procedure phase using instrument motion, contact force, and image features. The archive indexes each labeled interval together with its starting conditions and recorded motion segment. During execution, the controller estimates the current phase from the live stream, retrieves intervals with that label, and ranks their motion segments by similarity to the current measurements.

Specific version of a part

Recorded instrument pose regression

The training subsystem stores labeled image frames and synchronized six-axis instrument poses in a persistent archive. An offline regression model fits instrument position and orientation from image features and is selected using position and angular errors on held-out recordings.

Encoded lead screw stage

A brushless motor turns a preloaded lead screw that translates the instrument carriage along a linear rail. A linear encoder reports carriage position, and an inline strain-gauge load cell measures axial force during travel.

Time indexed subject buffer

A memory ring buffer stores timestamped force samples, joint positions, and image frame identifiers under the active subject session. The controller retrieves the newest entries using a shared clock to align measurements across sensors.

Landmark matched trajectory playback

The controller compares distances and angles between live landmarks with descriptors from archived training cases and selects the closest match. It registers the associated trajectory to the current landmarks and autonomously issues joint setpoints at a fixed update rate.

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 2007/0244724 A1
Published
Oct 18, 2007
Filed
Apr 13, 2006
Class
A61B 5/00
Document BUS 2020/0146754 A1
Published
May 14, 2020
Filed
Oct 5, 2018
Class
A61B 34/20
Claim 1
Document A
Document B
1.1(a) a training subsystem configured to obtain and store training data in an archived data pool and to train or select one or more models;
As shown in FIG. 2, the historical reference database includes a series of historic patient records 34 that can be accessed by the monitoring device 16 through the communication channel 36. The historic patient records are stored within the historic reference database 32 and include all of the medical record data, monitoring parameters, treatments and outcome information for a large number of patients.
A · US 2007/0244724 A1 · ¶ 19
1.2(b) a procedure-execution subsystem including a robotic apparatus having at least one actuator configured to move at least one instrument relative to the subject and one or more sensors configured to generate real-time procedural data during the physical medical procedure;
More specifically, the present invention relates to a monitoring device that obtains real-time monitoring data from a patient and accesses a historic reference database to receive a predicted outcome for the patient based upon real-time monitoring data and other medical record data for the patient.
A · US 2007/0244724 A1 · ¶ 1
The system 100 in this embodiment includes an imaging device 103, a motion tracking system 105 and a robotic arm 101 for performing a robotically-assisted surgical procedure. The robotic arm 101 may comprise a multi joint arm that includes a plurality of linkages connected by joints having actuator(s) and optional encoder(s) to enable the linkages to rotate, bend and/or translate relative to one another in response to control signals from a robot control system.
B · US 2020/0146754 A1 · ¶ 29
1.3(c) a real-time data pool associated with the subject; and
As shown in FIG. 1, the monitoring device 16 includes a real-time medical record 30 for the patient 12 being treated by the physician.
A · US 2007/0244724 A1 · ¶ 16
1.4(d) a controller comprising at least one processor and memory storing instructions that, when executed, cause the controller to: (i) access the archived data pool and the real-time data pool; (ii) determine a procedure plan based at least in part on comparisons between the real-time procedural data and the training data in the archived data pool; and (iii) execute the procedure plan by autonomously generating and issuing control commands to the robotic apparatus, wherein the physical medical procedure is executed without real-time guidance or intervention by a human operator.
In embodiments, the motion tracking system 105 may track the position of the robotic arm 101 (e.g., via marker device 202 on or proximate to end effector 102 as shown in FIG. 1) to determine the position of the end effector 102 within the common coordinate system of the patient.
B · US 2020/0146754 A1 · ¶ 59

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

Statements in these documents
Due to the vast amount of real-time information and monitoring parameters presented to the physician at one time, it is often difficult for the physician to recognize certain medical conditions that the physician may not see on a regular basis.
A · US 2007/0244724 A1 · ¶ 3
Based upon the outcome information, treatment information and diagnosis information, the physician can then more quickly and easily diagnose the patient, develop a treatment schedule and provide outcome information to the patient based upon the sampling of historic patients that had similar medical record data and monitoring parameters.
A · US 2007/0244724 A1 · ¶ 8
These procedures may be used to aid the surgeon in planning a surgical procedure and may also provide the surgeon with relevant feedback during the course of surgical procedure.
B · US 2020/0146754 A1 · ¶ 2
There is a continuing need to improve the safety and ease-of-use of computer-assisted surgical systems.
B · US 2020/0146754 A1 · ¶ 2
Cite aspatent.fyi, US 2026/0354911 A1, claim 1, first published 2026-11-19T05:29:29-05:00, https://patent.fyi/US20260354911/, SHA-256 13a4cfb98a16f16a40a784002481f891b099cafa729cc3cf76b62165c36da809
First published
Permanent addresshttps://patent.fyi/US20260354911/
Content fingerprint (SHA-256)13a4cfb98a16f16a40a784002481f891b099cafa729cc3cf76b62165c36da809
Methodv2026-11-18.2 · automated, not reviewed by a person
Correctionscorrections@patent.fyi