Laser speckle force feedback estimation
Inventors
Oberlin, John • Ashkezari, Hossein Dehghani
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12279851-B2
Publication Date
2025-04-22
Expiration Date
Abstract
Provided herein are systems, methods, and media capable of determining estimated force applied on a target tissue region to enable tactile feedback during interaction with said target tissue region.
Core Innovation
The disclosure describes a computer-implemented method for determining an estimated force applied on a target tissue region. The method obtains a set of images of the target tissue region, and determines a perfusion property, a set of spatial measurements, or both of the target tissue region based at least on the set of images. The method then determines a deformation of the target tissue region based at least on the set of spatial measurements.
The method determines a viscoelastic property of the target tissue region based at least on the deformation of the target tissue region, the perfusion property of the target tissue region, or both. The viscoelastic property is used to determine the estimated force applied on the target tissue region. The framework supports deriving tissue properties from image-derived perfusion and spatial deformation measurements and using those derived properties for force estimation.
The disclosure further characterizes embodiments in which the set of images may include laser speckle, RGB image, and/or RGB-Depth image, including examples using multi-wavelength illumination. Deformation and viscoelastic property determination are supported by real-time simultaneous acquisition during deformation, and by using additional sensed inputs such as depth measurements. The disclosure also describes machine-learning approaches for determining elastic/viscoelastic property determination and force-estimation models based on representations from two speckle images.
Claims Coverage
The document contains one independent claim with multiple dependent claims that specify the imaging inputs, the relationship between images and tissue measurements, and optional extensions to feedback/control and additional outputs. The independent claim includes inventive features centered on deriving perfusion and deformation from images, deriving viscoelastic properties, and mapping those properties to an estimated applied force.
Image-based force estimation via perfusion and deformation
Obtaining a set of images of the target tissue region; determining a perfusion property, a set of spatial measurements, or both based at least on the set of images; determining a deformation of the target tissue region based at least on the set of spatial measurements; determining a viscoelastic property based at least on the deformation and/or the perfusion property; and determining the estimated force applied on the target tissue region based at least on the viscoelastic property.
Multi-modal and multi-wavelength image acquisition
The set of images includes a laser speckle image, an RGB image, an RGB-Depth image, or any combination of these, and the method includes obtaining the set of images while emitting two or more different wavelengths of light at a target tissue region.
Depth-measurement supported deformation determination
Obtaining depth measurements from a depth sensor, and determining the deformation of the target tissue region based on those depth measurements.
Dimensionality-constrained deformation determination
Determining the deformation of the target tissue region includes one-dimensional, two-dimensional, and/or three-dimensional deformation, or combinations thereof.
Machine-learning determination of viscoelastic properties
Determining the viscoelastic property of a target tissue region using a machine learning algorithm.
Across the independent claim and its refinements, the core coverage maps images to perfusion and spatial measurements, derives deformation, derives viscoelastic properties, and then determines an estimated applied force. Dependent refinements specify image modality combinations, multi-wavelength illumination, optional depth-sensor measurements for deformation, constrained deformation dimensionality (1D/2D/3D), and machine-learning determination of viscoelastic properties.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?