US 10448903

Here is a concise summary of US patent 10448903:

Title: Methods and systems for retrospective internal gating

Assignee: Memorial Sloan Kettering Cancer Center

Inventor: Adam L. Kesner

Filing Date: November 1, 2018 (Application number US16/178,332)

Issue Date: October 22, 2019

Abstract: The invention describes a method for reconstructing respiratory-gated Positron Emission Tomography (PET) images from raw PET data. It uses respiratory motion information derived from signal fluctuations of individual voxels to create usable respiratory phase information. This method allows for the reconstruction of respiratory-gated PET images without external hardware and in a fully automated manner.

Plain-Language Overview of Independent Claims:

• Claim 1 (Method): This claim describes a method for retrospective internal gating that involves:

  1. Acquiring a series of images over time (t1...tn) which include a moving object.
  2. Extracting time-activity information for the individual voxels (3D pixels) within these images.
  3. Determining the phase information of the moving object's motion using this extracted time-activity information.
  4. Generating an updated series of images where corrections have been applied to account for the moving object's motion, utilizing the previously determined phase information.

• Claim 19 (Non-Transitory Computer-Readable Medium): This claim covers a computer-readable storage medium containing a program. When a processor executes this program, it instructs a machine to perform the same steps as outlined in Claim 1: acquire images, extract voxel time-activity information, determine motion phase information, and generate corrected images.

• Claim 20 (System): This claim describes a system that includes one or more processors and a non-transitory computer-readable medium. The medium stores instructions that, when executed by the processors, cause the system to perform the same method as described in Claim 1: acquire images, extract voxel time-activity information, determine motion phase information, and generate corrected images.

CAFC 2026 Dockets: A search of the CAFC 2026 dockets for patent 10448903 did not yield any specific results regarding litigation or other proceedings involving this patent in 2026.

As of April 26, 2026, a search for litigation involving US patent 10448903 across common patent litigation tracking sites and court dockets (CAFC, PACER, and Unified Patents) did not yield any specific results for known litigation.

PACER is a national index for district, bankruptcy, and appellate courts, allowing for nationwide searches to determine if a party is involved in federal litigation. However, access to detailed case information typically costs $0.10 per page. For cases filed on or after March 1, 2012, all case information and documents for the U.S. Court of Appeals for the Federal Circuit are available through PACER.

Therefore, based on the searches conducted, there is no known litigation information publicly available for US patent 10448903 at this time.

Proceedings overview

There are no AIA trial proceedings on file for US patent 10448903. This means the patent has not been challenged through Inter Partes Review (IPR), Post-Grant Review (PGR), or Covered Business Method (CBM) proceedings at the Patent Trial and Appeal Board (PTAB). Consequently, for a defendant, the patent claims remain untested in this forum, and there is no estoppel against future PTAB challenges.

Strategic summary

As there are no PTAB proceedings on file for US patent 10448903, all claims (1-20) remain untested by the PTAB. There are no canceled or sustained claims through this administrative process. This means that a potential defendant is not estopped under 35 U.S.C. § 315(e)(2) from raising any ground that could have been raised in an IPR, PGR, or CBM. The absence of PTAB activity suggests that the patent has not yet faced the scrutiny common for patents that are actively asserted or widely licensed.

Recommended next steps

Since no PTAB activity exists for US patent 10448903, a potential defendant has a full range of prior art grounds available should they choose to file a petition for Inter Partes Review (IPR) or Post-Grant Review (PGR), assuming the timing requirements are met. The absence of PTAB challenges also means there are no Final Written Decisions to consult or specific claims that have been validated or invalidated by the PTAB.

Inventors

The named inventor for US patent 10448903 is Adam L. Kesner. At the time of filing, it is highly probable that Adam L. Kesner was employed by Memorial Sloan Kettering Cancer Center, as they are both the original and current assignee of the patent.

Original assignee

The entity named on the issued patent is Memorial Sloan Kettering Cancer Center. Their primary line of business is cancer treatment, research, and education. As a leading medical and research institution, Memorial Sloan Kettering Cancer Center likely implements the methods and systems described in the patent (e.g., for improved diagnostic imaging in oncology) within their clinical practice, thereby embodying the claims. They are an operating entity and are currently active.

Assignment timeline

• 2017-05-30 (executed) / recorded 2024-05-30 — Reel 067574/0953
  • Conveyance: ASSIGNMENT OF ASSIGNORS INTEREST
  • Assignor: KESNER, ADAM L.
  • Assignee: MEMORIAL SLOAN KETTERING CANCER CENTER
  • Correspondent: MEMORIAL SLOAN KETTERING CANCER CENTER, OFFICE OF TECHNOLOGY DEVELOPMENT, 1275 YORK AVENUE, NEW YORK, NY 10065
  • Context: Inventor to original assignee transfer of interest in the invention.

The USPTO Assignment Center shows only one recorded assignment for US10448903, which is the assignment from the inventor to Memorial Sloan Kettering Cancer Center.

Timeline diagram

timeline
    title Ownership of US 10448903
    2007 : Priority date
    2017 : Inventor assigned rights to MSKCC
    2018 : Application filed
    2019 : Patent issued
    2024 : Assignment recorded

NPE / troll-pattern signals

1. Shell-entity transfer — Not present. The assignee, Memorial Sloan Kettering Cancer Center, is a well-known, operating medical and research institution, not a shell entity.
2. Known asserter in the chain — Not present. Memorial Sloan Kettering Cancer Center is not listed as a known patent asserter or NPE.
3. Repeat correspondent across the chain — Not present. Only one assignment is recorded (Reel 067574/0953), with Memorial Sloan Kettering Cancer Center's Office of Technology Development listed as the correspondent. There is no recurrence of this correspondent across a chain of transfers.
4. Cascading transfers — Not present. There is only one recorded assignment in the chain (Reel 067574/0953).
5. Pre-litigation transfer — Not present. No litigation has been identified for this patent, and the recorded assignment (executed 2017-05-30, recorded 2024-05-30, Reel 067574/0953) precedes the patent's issuance (2019-10-22) and is a transfer from the inventor to the initial assignee.
6. Bankruptcy fire-sale — Not present. Memorial Sloan Kettering Cancer Center is an active and operating institution.
7. Privateering — Not present. There is no evidence to suggest privateering, as the patent is held by a research institution rather than a typical operating company.
8. Defensive aggregator (anti-NPE) — Not present. The patent is currently assigned to Memorial Sloan Kettering Cancer Center, which is not a known defensive aggregator.

Verdict

Insufficient data. The sole recorded assignment for US10448903 is from the inventor, Adam L. Kesner, to Memorial Sloan Kettering Cancer Center (executed 2017-05-30, recorded 2024-05-30, Reel 067574/0953). This represents the initial transfer of rights from the inventor to his likely employer/institution, which is a standard patent prosecution practice. There are no subsequent transfers or any other signals indicating NPE activity.

USPTO Patent Assignment Search for US10448903

Most Relevant Prior Art for US Patent 10448903

Based on the "Patent Citations" section of US patent 10448903B2, excluding related applications within the same patent family (US9814431B2 and US10117625B2), the following patents are considered prior art. For each, a full citation, publication/filing date, brief description, and potential anticipation under 35 U.S.C. § 102 are provided.

The independent claims of US10448903 focus on:

1. Acquiring a series of images of a moving object.
2. Extracting time-activity information for voxels.
3. Determining motion phase information from this time-activity information.
4. Generating motion-corrected images using the phase information.

The novelty of US10448903 largely lies in the "retrospective internal gating" where motion information is derived from individual voxel signal fluctuations within the acquired image data itself, without external hardware, and in an automated manner, to create a time-varying object motion function.

Here's the analysis of the cited prior art:

1. US6298260B1

• Full Citation: US6298260B1, "Respiration responsive gating means and apparatus and methods using the same"
• Publication/Filing Date: Priority: February 25, 1998; Publication: October 2, 2001
• Brief Description: This patent describes a system and method for generating a respiration-responsive gating signal for use in medical imaging, particularly for PET scans. It primarily focuses on using external respiratory sensors (e.g., pressure transducers, strain gauges, motion detectors) to generate a gating signal that can then be used to correct for respiratory motion. It can involve storing scan data with corresponding respiratory phase information and reconstructing images for specific phases.
• Potential Anticipation (35 U.S.C. § 102): This patent does not anticipate the independent claims of US10448903. While it addresses respiratory gating and generating corrected images, it explicitly relies on external hardware (respiration sensors) to determine motion, which is contrary to the "internal" aspect of US10448903 that derives motion information from individual voxel signal fluctuations of the acquired images themselves. Therefore, elements 2 and 3 of Claim 1, regarding extracting time-activity information for voxels and determining phase information based on that voxel information, are not present.

2. US6144874A

• Full Citation: US6144874A, "Respiratory gating method for MR imaging"
• Publication/Filing Date: Priority: October 15, 1998; Publication: November 7, 2000
• Brief Description: This patent discloses a method for respiratory gating in Magnetic Resonance (MR) imaging. It involves acquiring MR data during a free-breathing cycle and identifying a respiration signal from a portion of the acquired MR data itself. The respiration signal is used to group data into different respiratory phases for image reconstruction. This patent explicitly mentions deriving motion information "directly from the MR data itself," without external sensors.
• Potential Anticipation (35 U.S.C. § 102): This patent potentially anticipates elements of Claim 1 of US10448903, particularly the concept of deriving motion information internally from the image data. It describes:
  • Acquiring images (MR data).
  • Extracting a respiration signal from a "portion of the acquired MR data." This is analogous to "extracting time-activity information for voxels" (Claim 1, element 2) and determining "phase information for motion" (Claim 1, element 3).
  • Using this signal to reconstruct images corrected for motion (Claim 1, element 4).
    The key distinction, if any, for US10448903 might lie in the specific "time-activity information for voxels" and how it's combined to form a "time varying object motion function" which is not explicitly detailed in US6144874A, which focuses more broadly on a "respiration signal." However, the principle of internal, data-driven gating to correct motion is present. It could anticipate Claim 1, 19, and 20.

3. US6556695B1

• Full Citation: US6556695B1, "Method for producing high resolution real-time images, of structure and function during medical procedures"
• Publication/Filing Date: Priority: February 5, 1999; Publication: April 29, 2003
• Brief Description: This patent describes a method for producing high-resolution real-time images by acquiring a series of images, identifying and tracking moving regions or objects within the images, and then using this motion information to re-register the images or select specific images for display or further processing. It focuses on motion tracking and correction for various medical procedures, potentially using image-based methods.
• Potential Anticipation (35 U.S.C. § 102): This patent potentially anticipates general aspects of motion correction from image data. It includes:
  • Acquiring a series of images with a moving object (Claim 1, element 1).
  • Identifying and tracking motion within the images to "generate motion information." This broadly covers extracting time-activity information and determining phase (Claim 1, elements 2 and 3).
  • Using this motion information to create higher resolution images (Claim 1, element 4).
    The description is broad enough that it could potentially cover the underlying principles of deriving motion information internally from images. It could anticipate Claim 1, 19, and 20.

4. US6501981B1

• Full Citation: US6501981B1, "Apparatus and method for compensating for respiratory and patient motions during treatment"
• Publication/Filing Date: Priority: March 16, 1999; Publication: December 31, 2002
• Brief Description: This patent describes a system and method for compensating for respiratory and patient motions during medical treatment, particularly radiotherapy. It uses a motion sensor (e.g., optical tracking system) to detect patient motion and then uses this information to adjust the treatment delivery (e.g., move the treatment couch, or gate the radiation beam).
• Potential Anticipation (35 U.S.C. § 102): This patent does not anticipate the independent claims of US10448903. Similar to US6298260B1, it relies on an external motion sensor to detect and compensate for motion, failing to include the "internal" derivation of motion information from voxel time-activity as claimed in US10448903 (Claim 1, elements 2 and 3).

5. US6539074B1

• Full Citation: US6539074B1, "Reconstruction of multislice tomographic images from four-dimensional data"
• Publication/Filing Date: Priority: August 25, 2000; Publication: March 25, 2003
• Brief Description: This patent describes a method for reconstructing 4D tomographic images (3D images over time), specifically addressing cardiac or respiratory motion. It involves acquiring image data over time, estimating a motion field (e.g., using a reference image and registration techniques), and then applying motion compensation during reconstruction to produce a motion-corrected image.
• Potential Anticipation (35 U.S.C. § 102): This patent potentially anticipates elements of US10448903 related to 4D imaging and motion correction. It involves:
  • Acquiring images over time (Claim 1, element 1).
  • Estimating a motion field, which is a form of determining motion information from the acquired data. While not explicitly "voxel time-activity," it's data-driven motion estimation (analogous to Claim 1, elements 2 and 3).
  • Reconstructing motion-compensated images (Claim 1, element 4).
    The specific method of deriving the motion function from "individual voxel signal fluctuations" as emphasized in US10448903's abstract might be a distinguishing feature, but the overall concept of retrospective, data-driven motion correction in tomographic imaging is present. It could anticipate Claim 1, 19, and 20.

6. US7734078B2

• Full Citation: US7734078B2, "System and method for generating composite subtraction images for magnetic resonance imaging"
• Publication/Filing Date: Priority: September 18, 2002; Publication: June 8, 2010
• Brief Description: This patent focuses on generating composite subtraction images in MRI, particularly for perfusion studies, by performing image registration to correct for motion between a baseline image and subsequent images. It aims to reduce motion artifacts in subtraction images.
• Potential Anticipation (35 U.S.C. § 102): This patent describes motion correction between images for subtraction purposes rather than explicitly gating based on an internally derived physiological motion phase. While it involves image acquisition (Claim 1, element 1) and motion correction (Claim 1, element 4) using image registration, it does not explicitly disclose extracting "time-activity information for voxels" to "determine phase information for motion" of a cyclical nature in the same way as US10448903. Therefore, it is less likely to anticipate the specific method of US10448903's core claims.

7. US20040218794A1

• Full Citation: US20040218794A1, "Method for processing perfusion images"
• Publication/Filing Date: Priority: May 1, 2003; Publication: November 4, 2004
• Brief Description: This patent application describes methods for processing perfusion images, including steps for motion correction between acquired image frames. It mentions calculating motion vectors and performing image registration to align images acquired at different times.
• Potential Anticipation (35 U.S.C. § 102): Similar to US7734078B2, this patent focuses on motion correction via image registration rather than deriving a physiological gating signal from voxel-level fluctuations. While it involves acquiring images (Claim 1, element 1) and correcting for motion (Claim 1, element 4), it does not explicitly detail the extraction of time-activity information for voxels to determine cyclical phase information (Claim 1, elements 2 and 3) in the manner described by US10448903.

8. US7359535B2

• Full Citation: US7359535B2, "Systems and methods for retrospective internal gating"
• Publication/Filing Date: Priority: June 20, 2003; Publication: April 15, 2008
• Brief Description: This patent describes methods for retrospective internal gating in imaging systems, particularly CT. It involves acquiring images, identifying temporally cyclical signals from the images themselves (e.g., from a region of interest), and using these signals to create a time-varying object motion function that correlates with phases of periodic motion. This function is then used to gate the images.
• Potential Anticipation (35 U.S.C. § 102): This patent highly anticipates the independent claims of US10448903. The title itself ("Systems and methods for retrospective internal gating") is very similar. It explicitly details:
  • Acquiring images at different times (Claim 1, element 1).
  • Identifying temporally cyclical signals from the images themselves (analogous to extracting time-activity information for voxels and determining phase information) (Claim 1, elements 2 and 3).
  • Creating a time-varying object motion function correlating times and phases of periodic motion.
  • Using this function to perform gating.
    The abstract of US10448903 states "respiratory motion information derived from individual voxel signal fluctuations, is used in combination to create usable respiratory phase information." US7359535B2 also describes deriving internal signals from image data for gating. The primary difference may lie in the specific algorithmic details of how "voxel time-activity information" is extracted, prioritized, and combined to generate the time-varying object motion function, as detailed in the dependent claims and description of US10448903 (e.g., weighting factors, filtering, iterative summation based on standard deviation). However, the core concept of retrospective internal gating is clearly present. It could anticipate Claim 1, 19, and 20.

9. US20050123183A1

• Full Citation: US20050123183A1, "Data driven motion correction for nuclear imaging"
• Publication/Filing Date: Priority: September 2, 2003; Publication: June 9, 2005
• Brief Description: This patent application describes methods for motion correction in nuclear imaging (e.g., PET, SPECT) by extracting motion information directly from the acquired emission data (list mode data). It outlines techniques to reconstruct motion fields or motion correction factors from the raw data itself, without external sensors, and then use these to correct for motion artifacts.
• Potential Anticipation (35 U.S.C. § 102): This patent highly anticipates the independent claims of US10448903. It specifically addresses "data driven motion correction for nuclear imaging," which aligns directly with US10448903's focus on PET. It describes:
  • Acquiring image data (list mode data for nuclear imaging).
  • Extracting motion information directly from the emission data itself (analogous to extracting time-activity information for voxels and determining phase information).
  • Using this information to correct for motion artifacts (generating updated images correcting for motion).
    The description details deriving motion from the data itself and correcting for it, encompassing the core elements of Claim 1 of US10448903. It could anticipate Claim 1, 19, and 20.

10. US7756307B2

• Full Citation: US7756307B2, "Method of, and software for, conducting motion correction for a tomographic scanner"
• Publication/Filing Date: Priority: October 17, 2003; Publication: July 13, 2010
• Brief Description: This patent describes a method for motion correction in tomographic scanning, particularly PET. It involves acquiring data, detecting motion from the raw data (e.g., from the sinogram or projections), and then using this detected motion information to correct the reconstructed images. It emphasizes "data-derived motion correction" without external markers.
• Potential Anticipation (35 U.S.C. § 102): This patent highly anticipates the independent claims of US10448903. It explicitly describes "data-derived motion correction" for tomographic scanners like PET, which directly aligns with US10448903's approach. It involves:
  • Acquiring image data.
  • Detecting motion from the raw data (analogous to extracting voxel time-activity and determining phase information).
  • Using this motion information to correct reconstructed images.
    The concept of deriving motion internally from the data for correction is central to both patents. It could anticipate Claim 1, 19, and 20.

11. US7574249B2

• Full Citation: US7574249B2, "Device-less gating of physiological movement for improved image detection"
• Publication/Filing Date: Priority: February 8, 2005; Publication: August 11, 2009
• Brief Description: This patent describes a method for "device-less gating" of physiological movement (e.g., respiration, cardiac activity) for improved medical imaging. It involves acquiring image data, identifying physiological movement from the image data itself (e.g., by tracking features or changes in intensity over time), and using this to gate the images.
• Potential Anticipation (35 U.S.C. § 102): This patent highly anticipates the independent claims of US10448903. The term "device-less gating" strongly correlates with "retrospective internal gating" and the explicit statement in US10448903's abstract about not using external hardware. It includes:
  • Acquiring image data with physiological movement (Claim 1, element 1).
  • Identifying physiological movement from the image data itself (analogous to extracting voxel time-activity and determining phase information) (Claim 1, elements 2 and 3).
  • Using this information for gating to improve image detection (generating updated images correcting for motion) (Claim 1, element 4).
    This patent teaches the core concept of extracting physiological motion signals directly from the image data for gating purposes. It could anticipate Claim 1, 19, and 20.

12. US20070081704A1

• Full Citation: US20070081704A1, "System, program product, and methods for attenuation correction of emission data on PET/CT and SPECT/CT"
• Publication/Filing Date: Priority: March 17, 2005; Publication: April 12, 2007
• Brief Description: This patent application focuses on methods for attenuation correction in hybrid PET/CT and SPECT/CT imaging, addressing motion artifacts during the acquisition of attenuation maps (from CT) and emission data (from PET/SPECT). It suggests correcting for misalignment due to respiratory motion, potentially by incorporating motion models.
• Potential Anticipation (35 U.S.C. § 102): While this patent addresses motion and its impact on image quality (specifically attenuation correction in PET/CT), its primary focus is on the correction of attenuation maps rather than the retrospective internal gating method of deriving physiological phase information from voxel time-activity. It may broadly relate to motion-corrected imaging but does not specifically teach the steps of US10448903's independent claims.

13. US20080226149A1

• Full Citation: US20080226149A1, "Motion Compensation in Functional Imaging"
• Publication/Filing Date: Priority: August 4, 2005; Publication: September 18, 2008
• Brief Description: This patent application describes methods for motion compensation in functional imaging (e.g., PET, SPECT, fMRI). It involves acquiring dynamic image data, detecting motion from the acquired image data, determining a motion vector field, and applying this to correct the functional images. It can involve generating a motion curve.
• Potential Anticipation (35 U.S.C. § 102): This patent highly anticipates the independent claims of US10448903. It specifically mentions detecting motion from the acquired image data for functional imaging, which is a key aspect of US10448903. It describes:
  • Acquiring dynamic image data (Claim 1, element 1).
  • Detecting motion from the acquired image data (analogous to extracting voxel time-activity and determining phase information) (Claim 1, elements 2 and 3).
  • Applying motion compensation to correct images (generating updated images correcting for motion) (Claim 1, element 4).
    The concept of internal, data-driven motion compensation in functional imaging directly addresses the core innovation of US10448903. It could anticipate Claim 1, 19, and 20.

14. US20070127797A1

• Full Citation: US20070127797A1, "Methods and systems to facilitate reducing banding artifacts in images"
• Publication/Filing Date: Priority: November 23, 2005; Publication: June 7, 2007
• Brief Description: This patent application addresses reducing banding artifacts in images, which can be caused by various factors, including motion. It describes methods for identifying and correcting for such artifacts, possibly through data processing or reconstruction adjustments.
• Potential Anticipation (35 U.S.C. § 102): This patent focuses on reducing specific image artifacts (banding) that might be caused by motion, but it does not disclose the specific method of deriving physiological phase information from voxel time-activity to gate images for motion correction as claimed in US10448903.

15. US20070237372A1

• Full Citation: US20070237372A1, "Cross-time and cross-modality inspection for medical image diagnosis"
• Publication/Filing Date: Priority: December 29, 2005; Publication: October 11, 2007
• Brief Description: This patent application describes methods for medical image diagnosis involving comparing images acquired at different times or from different modalities. It may involve image registration techniques to align images for comparison, but the primary focus is on diagnostic comparison rather than motion gating.
• Potential Anticipation (35 U.S.C. § 102): This patent focuses on image comparison and registration for diagnostic purposes. While image registration can correct for some motion, it does not explicitly teach the specific steps of extracting voxel time-activity to determine cyclical phase information for motion gating as found in US10448903's independent claims.

16. US20090290774A1

• Full Citation: US20090290774A1, "Dynamic computed tomography imaging"
• Publication/Filing Date: Priority: May 26, 2006; Publication: November 26, 2009
• Brief Description: This patent application describes methods for dynamic computed tomography imaging, particularly for reducing motion artifacts. It involves acquiring data over time, estimating a motion model or motion vectors from the acquired data, and then using this to reconstruct motion-corrected dynamic images.
• Potential Anticipation (35 U.S.C. § 102): This patent potentially anticipates elements of US10448903, especially in the context of CT. It involves:
  • Acquiring dynamic image data (Claim 1, element 1).
  • Estimating motion from the acquired data (analogous to extracting voxel time-activity and determining phase information) (Claim 1, elements 2 and 3).
  • Reconstructing motion-corrected images (Claim 1, element 4).
    The general principle of deriving motion information internally from image data for correction is present. It could anticipate Claim 1, 19, and 20.

17. US20100183206A1

• Full Citation: US20100183206A1, "Adjusting acquisition protocols for dynamic medical imaging using dynamic models"
• Publication/Filing Date: Priority: June 21, 2007; Publication: July 22, 2010
• Brief Description: This patent application describes adjusting imaging acquisition protocols based on dynamic models, potentially including patient motion. It focuses on optimizing the acquisition process rather than retrospectively correcting images using internally derived phase information.
• Potential Anticipation (35 U.S.C. § 102): This patent deals with adjusting acquisition protocols and dynamic models, which is a different inventive concept than the retrospective internal gating method claimed in US10448903. While it relates to dynamic imaging and motion, it does not disclose the specific steps of deriving phase information from voxel time-activity and then using it to generate corrected images in a retrospective manner.

18. US20090076369A1

• Full Citation: US20090076369A1, "Method For Reducing Motion Artifacts In Highly Constrained Medical Images"
• Publication/Filing Date: Priority: September 17, 2007; Publication: March 19, 2009
• Brief Description: This patent application describes methods for reducing motion artifacts in medical images by exploiting "highly constrained" image data (e.g., sparsity). It focuses on advanced reconstruction techniques that inherently deal with motion, rather than explicitly deriving a motion phase signal from voxel time-activity for gating.
• Potential Anticipation (35 U.S.C. § 102): This patent addresses motion artifact reduction through reconstruction algorithms that leverage image constraints. It does not explicitly teach the steps of extracting time-activity information from individual voxels to determine phase information for gating and then generating corrected images based on that phase information, which is central to US10448903.

19. US20090299184A1

• Full Citation: US20090299184A1, "Imaging or communications system utilizing multisample apodization and method"
• Publication/Filing Date: Priority: February 16, 2008; Publication: December 3, 2009
• Brief Description: This patent application describes imaging systems and methods utilizing multisample apodization, primarily focusing on improving image quality by suppressing artifacts or enhancing resolution through signal processing techniques, which could include handling motion-induced blurring.
• Potential Anticipation (35 U.S.C. § 102): This patent focuses on signal processing (apodization) for image quality. While it might address aspects of motion-induced blurring in a general sense, it does not explicitly disclose the specific method of deriving physiological phase information from voxel time-activity to gate images for motion correction as claimed in US10448903.

Summary of Most Relevant Prior Art:

The most relevant prior art documents that potentially anticipate the independent claims (Claim 1, 19, and 20) of US10448903, due to their disclosure of "internal" or "device-less" data-driven motion detection and correction for medical imaging, are:

• US6144874A: "Respiratory gating method for MR imaging" - Teaches deriving a respiration signal directly from MR data itself for gating.
• US6556695B1: "Method for producing high resolution real-time images, of structure and function during medical procedures" - Broadly covers identifying and tracking motion from images for correction.
• US6539074B1: "Reconstruction of multislice tomographic images from four-dimensional data" - Describes estimating motion fields from acquired data for motion compensation.
• US7359535B2: "Systems and methods for retrospective internal gating" - Extremely similar in title and concept, explicitly teaching retrospective internal gating by identifying cyclical signals from images to create a time-varying object motion function.
• US20050123183A1: "Data driven motion correction for nuclear imaging" - Specifically teaches extracting motion information directly from raw emission data for nuclear imaging.
• US7756307B2: "Method of, and software for, conducting motion correction for a tomographic scanner" - Describes data-derived motion correction from raw data in tomographic scanning.
• US7574249B2: "Device-less gating of physiological movement for improved image detection" - Explicitly teaches "device-less gating" by identifying physiological movement from image data itself.
• US20080226149A1: "Motion Compensation in Functional Imaging" - Describes detecting motion from acquired image data for functional imaging motion compensation.
• US20090290774A1: "Dynamic computed tomography imaging" - Teaches estimating motion from acquired data for motion-corrected dynamic CT images.

These patents all demonstrate some form of acquiring image data, deriving motion information from that data internally (as opposed to external sensors), and then using that derived motion information to correct or gate the images. The specific advancements or distinctions of US10448903 would likely reside in the detailed methodology of how the voxel time-activity information is processed and combined to form the time-varying object motion function, which might be found in its dependent claims and detailed description.

Obviousness Analysis of US10448903 Under 35 U.S.C. § 103

This analysis assesses the obviousness of US patent 10448903 under 35 U.S.C. § 103, considering combinations of prior art references cited within the patent and readily available through public search. The independent claims of US10448903 focus on a method, computer-readable medium, and system for retrospective internal gating by extracting time-activity information from image voxels to determine motion phase and generate corrected images.

Independent Claims of US10448903:

• Claim 1 (Method):
  1. Acquiring a series of images at times t1...tn including a moving object;
  2. Extracting time-activity information for voxels of the images;
  3. Determining phase information for motion of the moving object based on the time-activity information for voxels of the images; and
  4. Generating an updated series of images correcting for the motion of the moving object using the determined phase information for motion of the moving object.
• Claim 19 (Non-Transitory Computer-Readable Medium): A computer-readable medium encoded with a program to perform the method of Claim 1.
• Claim 20 (System): A system comprising one or more processors and a non-transitory computer-readable medium with instructions to perform the method of Claim 1.

Identified Prior Art References:

Several prior art references cited in US10448903 are particularly relevant to establishing obviousness. All cited patents below have priority dates earlier than the earliest priority date of US10448903 (May 4, 2007).

1. US7359535B2 (GE Medical Systems Global Technology Company, Llc): "Systems and methods for retrospective internal gating" (Priority: June 20, 2003; Publication: April 15, 2008)
   • Disclosure: This patent describes acquiring projection data for a subject over time, determining a motion function using this data, sorting the data into phases, and reconstructing an image for at least one phase. It explicitly teaches "retrospective internal gating" using acquired data to derive a motion function for phase-based image reconstruction.
2. US7574249B2 (General Electric Company): "Device-less gating of physiological movement for improved image detection" (Priority: February 8, 2005; Publication: August 11, 2009)
   • Disclosure: This patent teaches methods for physiological movement gating without a physical sensor. It includes acquiring scan data, generating a physiological gating signal from this data, sorting the scan data into phases, and reconstructing an image for at least one phase.
3. US20050123183A1 (Paul Schleyer): "Data driven motion correction for nuclear imaging" (Priority: September 2, 2003; Publication: June 9, 2005)
   • Disclosure: This application describes deriving motion information from acquired nuclear imaging data and reconstructing a motion-corrected image based on this information. It also discusses monitoring "differences in count rates for individual detector elements or regions of an image... related to physiological motion."
4. US8229187B2 (General Electric Company): "Respiratory motion extraction from tomographic projection and image data" (Priority: April 11, 2007; Publication: July 24, 2012)
   • Disclosure: This patent provides methods for detecting a respiratory motion signal from tomographic projection data and image data, characterizing the motion, and reconstructing respiratory-gated image data. It discusses using time-varying signals from regions of interest (ROIs) and filtering these signals to isolate respiratory frequencies.

Obviousness Combination and Motivation:

A person having ordinary skill in the art (PHOSITA) in medical imaging, particularly in areas like PET or CT reconstruction and motion correction, would have been motivated to combine the teachings of these prior art references to arrive at the invention claimed in US10448903.

Combination: US7359535B2 in view of US20050123183A1 and US8229187B2.

Motivation for Claim 1 Steps:

1. Acquiring a series of images at times t1...tn including a moving object: This step is a fundamental prerequisite for any motion correction technique and is implicitly or explicitly taught by all cited prior art. US7359535B2, US7574249B2, US20050123183A1, and US8229187B2 all involve acquiring imaging data over time to address motion.
2. Extracting time-activity information for voxels of the images:
   • Motivation for "Internal" or "Device-less" Gating: US7359535B2 and US7574249B2 explicitly teach "retrospective internal gating" and "device-less gating" by deriving a motion function or physiological gating signal directly from acquired scan or projection data, rather than external sensors. The background of US10448903 itself highlights the advantages of "software based methods" that are "image based, and thus machine independent," providing a clear motivation for a PHOSITA to pursue such approaches.
   • Motivation for Voxel-level Information: US20050123183A1 teaches "deriving motion information from the acquired nuclear imaging data" and suggests monitoring "differences in count rates for individual detector elements or regions of an image... related to physiological motion." A PHOSITA would readily understand that a "region of an image" can encompass individual voxels, and that monitoring activity fluctuations at the voxel level offers fine-grained spatial information about motion. US8229187B2 further strengthens this by explicitly teaching "respiratory motion extraction from tomographic projection and image data," and describing the use of "time-varying signal from these ROIs [regions of interest]" to characterize motion. Given that images are composed of voxels, analyzing the time-varying signal of individual or groups of voxels (i.e., time-activity information) to detect motion would be an obvious application of these teachings to achieve a more precise internal gating signal.
3. Determining phase information for motion of the moving object based on the time-activity information for voxels of the images: Once a time-varying signal related to motion (such as respiratory or cardiac motion) is derived from image data or voxels (as motivated above), determining phase information is a standard practice in motion gating. US7359535B2 and US7574249B2 both teach using the derived motion function/gating signal to sort data into "phases." This is the natural and expected next step for any gating process.
4. Generating an updated series of images correcting for the motion of the moving object using the determined phase information for motion of the moving object: This is the ultimate goal of all the prior art references dealing with motion correction and gating. US7359535B2, US7574249B2, US20050123183A1, and US8229187B2 all disclose the reconstruction of motion-corrected or gated images as the objective of their methods. Therefore, once phase information is determined from voxel time-activity, applying this information to generate motion-corrected images would be an obvious consequence and desired outcome for a PHOSITA.

Dependent Claims:
The dependent claims of US10448903, such as those related to assigning weighting factors (Claims 3, 4, 6, 7), filtering time-activity signals (Claims 8, 9, 10), and the iterative process for developing a time-varying object motion function (Claims 2, 11-14), would also be obvious. Weighting factors based on mean activity, spatial gradients, or signal variation are routine signal processing techniques for prioritizing data in motion detection. Filtering time-varying signals to isolate specific frequencies (e.g., respiratory) is explicitly taught by US8229187B2. The iterative combination of individual signals, accounting for phase differences, is a known approach in signal processing to enhance a composite signal and improve robustness, which a PHOSITA would apply to optimize the derived motion function.

Conclusion:

Based on the combination of US7359535B2, US7574249B2, US20050123183A1, and US8229157B2, the independent claims (1, 19, and 20) of US10448903 would be obvious to a person having ordinary skill in the art. The motivation to combine these references stems from the recognized advantages of device-less and image-based internal gating for more accurate motion correction in medical imaging, as well as the logical extension of existing motion detection techniques to voxel-level analysis for improved spatial resolution of motion information.