Federal Circuit Remands AI Image Registration Patent Case to USPTO: Key Insights for AI Patent Prosecution
In a significant procedural development for artificial intelligence patent litigation, the U.S. Court of Appeals for the Federal Circuit issued a decisive ruling in *Wentao Zhu et al. v. Katherine K. Vidal* (Case No. 24-1558), granting the plaintiffs’ motion and remanding the case to the USPTO for further proceedings. Closed on July 26, 2024, just 136 days after filing, the case centered on U.S. Patent Application No. 16/540717 — covering a neural network system for image registration and image segmentation trained using a registration simulator — a technology with broad implications across medical imaging, autonomous systems, and computer vision.
The defendant, Katherine K. Vidal, named in her official capacity as Director of the USPTO, faced a patentability challenge that ultimately resulted in agency remand rather than a final merits determination. For patent attorneys navigating AI patent prosecution, in-house IP counsel monitoring USPTO examination practices, and R&D leaders developing neural network-based imaging systems, this case offers critical strategic insights into how patentability disputes over machine learning inventions are increasingly resolved at the appellate level.
What would you like to do next?
Choose your path based on your current needs:
📋 Case Summary
| Case Name | Wentao Zhu et al. v. Katherine K. Vidal |
| Case Number | 2024-1558 (Fed. Cir.) |
| Court | Federal Circuit, Appeal from USPTO |
| Duration | March 12, 2024 – July 26, 2024 136 days |
| Outcome | Agency Remand — No Final Patentability Ruling |
| Patent at Issue | |
| Technology at Issue | Neural Network System for Image Registration |
Case Overview
The Parties
⚖️ Plaintiffs
Named inventors on patent application US16/540717, representing a technically sophisticated team working at the intersection of deep learning and medical or computational imaging.
🛡️ Defendant
Named in her official capacity as Director of the United States Patent and Trademark Office, this dispute arose from a USPTO patentability determination.
The Patent at Issue
This case centered on U.S. Patent Application No. 16/540717 (Publication No. US20210049757A1) — covering a neural network system for image registration and image segmentation trained using a registration simulator — a technology with broad implications across medical imaging, autonomous systems, and computer vision.
- • US16/540717 — Neural network system for image registration and segmentation
Developing AI-driven imaging tech?
Ensure your neural network systems for image registration are clear of patent challenges before launch.
Litigation Timeline & Procedural History
| Milestone | Date |
| Case Filed | March 12, 2024 |
| Court | U.S. Court of Appeals for the Federal Circuit |
| Case Closed | July 26, 2024 |
| Total Duration | 136 days |
Filed on March 12, 2024, and resolved in approximately 136 days, this case moved at a notably expedited pace for Federal Circuit appellate proceedings. The venue — the Court of Appeals for the Federal Circuit in the District of Columbia — is the exclusive appellate forum for USPTO patentability disputes, making it the definitive battleground for challenges to patent examination outcomes.
The case was categorized as an Invalidity/Cancellation Action with Patentability as the core verdict cause, indicating the dispute originated from a USPTO rejection or adverse determination during prosecution or post-grant proceedings. The appeal was resolved by motion practice rather than full merits briefing and oral argument, with the court granting the motion and remanding — a procedural outcome that suggests the parties or the court identified a specific agency error requiring correction without full appellate adjudication. No chief judge was designated in the case record.
The Verdict & Legal Analysis
Outcome
The Federal Circuit granted the plaintiffs’ motion and remanded the case to the USPTO for further proceedings consistent with its order. Critically, the court ordered that each side bear its own costs — a neutral cost allocation suggesting the remand was procedural or consent-based rather than a clear victory on the merits for either party. No damages were awarded, as this was a patentability dispute against the USPTO rather than a private infringement action.
Verdict Cause Analysis
The case was classified under Patentability with an Invalidity/Cancellation framework, placing it squarely within the category of disputes where inventors challenge adverse USPTO determinations — typically arising from:
- Final rejections during examination (e.g., obviousness rejections under 35 U.S.C. § 103 citing prior art combinations)
- Inter partes review (IPR) or ex parte reexamination outcomes
- Patent Trial and Appeal Board (PTAB) decisions unfavorable to the patent applicant or owner
The remand outcome — without a final ruling on patentability — indicates the Federal Circuit identified a procedural deficiency, inadequate agency reasoning, or an error in the USPTO’s application of patentability standards that required the agency to reconsider its determination. This type of outcome is significant: it neither confirms nor invalidates the patent claims, but compels the USPTO to conduct fresh analysis under corrected legal standards.
For AI and machine learning patents, common grounds for such remands include improper application of the Alice/Mayo framework for patent-eligible subject matter (35 U.S.C. § 101), insufficient consideration of claim limitations in prior art analysis, or failure to give adequate weight to applicant arguments during examination.
Legal Significance
The Federal Circuit’s decision to remand — rather than affirm or reverse outright — carries meaningful precedential weight for AI patent prosecution:
- USPTO examination quality for AI patents remains under judicial scrutiny. The court’s willingness to remand signals ongoing concern about how examiners and the PTAB evaluate neural network and machine learning claims.
- Procedural correctness in agency proceedings is essential. Even technically sound rejections can be undone on appeal if the USPTO’s reasoning is inadequate or fails to address all claim limitations.
- The 136-day resolution timeline demonstrates that Federal Circuit remands can be obtained efficiently through motion practice, without exhaustive briefing — a tactical option worth considering for patent applicants facing adverse USPTO decisions.
Strategic Takeaways
For Patent Attorneys:
- Build detailed prosecution records that anticipate Federal Circuit review standards
- Challenge USPTO rejections that lack claim-by-claim analysis or fail to address applicant arguments
- Consider motion-based appellate strategies for procedurally deficient agency decisions
For IP Professionals:
- Monitor remand outcomes as indicators of USPTO compliance with evolving AI patentability standards
- Evaluate pending AI patent applications for similar procedural vulnerabilities that could support appeals
For R&D Teams:
- Understand that a remand does not resolve patentability — freedom-to-operate analyses for image registration AI should account for the continued pendency of US16/540717
Freedom to Operate (FTO) Analysis & Implications
The remand in *Zhu v. Vidal* has significant FTO implications for AI-driven imaging technologies. Choose your next step:
📋 Understand This Case’s Implications
Learn how Federal Circuit remands impact AI patentability standards and your competitive landscape.
- Monitor the USPTO’s reconsideration post-remand
- Identify trends in Federal Circuit’s review of AI patent rejections
- Access comprehensive legal analysis and precedents
🔍 Check My AI Product’s Risk
Perform an FTO analysis to assess potential infringement risks for your neural network-based imaging systems.
- Input your AI model’s architecture or functional description
- AI identifies potentially blocking patents (like US16/540717)
- Receive a detailed, actionable risk assessment
High Risk Area
AI-driven image registration and segmentation
USPTO Scrutiny
Ongoing oversight of AI patent examination
Strategic Remand
Opportunity for refined claim prosecution
✅ Key Takeaways
Federal Circuit remands in patentability disputes can be obtained via motion practice in as few as 136 days.
Search related Federal Circuit cases →Inadequate USPTO reasoning — particularly on AI/ML patent claims — remains a viable appellate ground.
Explore appellate strategies →Cost-neutral remand orders suggest procedural rather than substantive appellate victories are increasingly common in USPTO appeal practice.
Analyze USPTO appeal trends →*Zhu v. Vidal* (24-1558) is a case to monitor as USPTO issues its remand decision.
Track this case on PatSnap →Frequently Asked Questions
The case involves U.S. Patent Application No. 16/540717 (Publication No. US20210049757A1), covering a neural network system for image registration and image segmentation trained using a registration simulator.
The court granted the plaintiffs’ motion and remanded the case to the USPTO for further proceedings, with each side bearing its own costs. No final patentability determination was issued.
The remand signals continued Federal Circuit oversight of USPTO AI patent examination quality, reinforcing that inadequate agency reasoning in neural network patent rejections remains appealable — a strategic consideration for all AI patent holders and applicants.
Ready to Strengthen Your Patent Strategy?
Join 18,000+ IP professionals using PatSnap Eureka to conduct prior art searches, draft patents, and analyse competitive landscapes with AI-powered precision.
References
- United States Court of Appeals for the Federal Circuit — Case 2024-1558
- U.S. Patent and Trademark Office — Patent Center Application US16/540717
- PACER Case No. 24-1558
- PatSnap — AI Patent Intelligence Solutions
This article is for informational purposes only and does not constitute legal advice. All case information is drawn from publicly available court records. For platform capabilities, visit PatSnap.