Rondevoo Tech LLC v. DeepRadiology Corp

I. Executive Summary and Procedural Information

• Parties & Counsel:
  • Plaintiff: Rondevoo Technologies, LLC (California)
  • Defendant: DeepRadiology Corporation (Delaware)
  • Plaintiff’s Counsel: Brandt Law Firm
• Case Identification: 1:20-cv-00592, D. Del., 04/29/2020
• Venue Allegations: Venue is based on Defendant's incorporation in the State of Delaware, which Plaintiff asserts establishes residency in the district pursuant to the Supreme Court's decision in TC Heartland.
• Core Dispute: Plaintiff alleges that Defendant’s "Deep Radiology Imaging Technology" infringes three patents related to systems and methods for generating special-purpose, learning-based image analysis algorithms.
• Technical Context: The technology concerns computer-implemented systems that are trained by users to automatically identify, classify, and quantify specific features or "entities" within digital images, with applications in medical diagnostics such as Alzheimer's disease research.
• Key Procedural History: The complaint does not mention any prior litigation, inter partes review proceedings, or licensing history. The three patents-in-suit are alleged to share a common specification, which may suggest that claim terms should be construed consistently across the patent family.

Case Timeline

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 7,088,854 - Method and apparatus for generating special-purpose image analysis algorithms

The Invention Explained

• Problem Addressed: The patent’s background section describes the difficulty and inconsistency of using computer systems to quantify specific objects, or "entities," within complex images (Compl. ¶¶9-11; ’854 Patent, col. 1:36-47). This task often required manual analysis by experts, which was time-consuming, expensive, and produced subjective, qualitative results rather than standardized, quantitative data (’854 Patent, col. 2:1-26).
• The Patented Solution: The invention is a system that learns from an expert user to automate image analysis. A user provides input to an "evolving algorithm" by identifying examples of target entities within an image, thereby exercising their "judgment" (’854 Patent, col. 6:12-24). The system uses this input to generate a storable "product algorithm" that can then be used to automatically classify and count instances of those entities in new, different images (’854 Patent, col. 6:26-35).
• Technical Importance: This approach was designed to standardize the classification of histological and other scientific image data, enabling reproducible, large-scale quantitative studies that were previously impractical (’854 Patent, col. 3:1-17).

Key Claims at a Glance

• The complaint asserts independent Claim 1 (Compl. ¶17).
• The essential elements of Claim 1 are:
  • A computer program product on a computer usable medium.
  • Obtaining an image with multiple chromatic data points.
  • Generating an "evolving algorithm" that partitions the data points into an entity identified "in accordance with a user's judgment."
  • Storing an instance of this evolving algorithm as a "product algorithm."
  • The product algorithm enables the "automatic classification" of the entity within a second image.

U.S. Patent No. 7,254,266 - Method and apparatus for generating special-purpose image analysis algorithms

The Invention Explained

• Problem Addressed: The patent addresses the same problem as the ’854 Patent, with which it shares a common specification: the need for a reproducible and automated method for expert quantification of image data (’266 Patent, col. 1:36-47, col. 2:1-26).
• The Patented Solution: The ’266 Patent claims a method that emphasizes the process of training and disseminating the learned expertise. The claimed method involves a specific, iterative training mode where a first user provides feedback to refine an evolving algorithm, which is then stored as a product algorithm and provided to a second user for application on new image sets (’266 Patent, Claim 1). This explicitly covers the capture of one user's expertise for use by another.
• Technical Importance: The claimed method provides a framework for creating and distributing expert-trained analytical tools, standardizing analysis across different users and labs (’266 Patent, col. 3:1-17).

Key Claims at a Glance

• The complaint asserts independent Claim 1 (Compl. ¶22).
• The essential elements of Claim 1 are:
  • A method for automating expert quantification of image data.
  • Obtaining a product algorithm via a "training mode" that uses "iterative input" from a "first user" to create an "evolving algorithm."
  • The training mode comprises a multi-step feedback loop: presenting a first set of entities for feedback, obtaining feedback, executing the algorithm, presenting a second set for feedback, obtaining approval, and storing the result as a product algorithm.
  • Providing the product algorithm to a "second user" to apply against a "second set of image data."

U.S. Patent No. 8,687,879 - Method and apparatus for generating special-purpose image analysis algorithms

• Technology Synopsis: As a continuation of the same patent family, the ’879 Patent addresses the same technical problem of automated image quantification (’879 Patent, col. 1:35-46). It claims a "non-transitory computer program product" where a "locked evolving algorithm" is generated through a multi-step, iterative training mode involving user feedback, similar to the method claimed in the ’266 Patent. This claim language appears to update the concepts from the earlier patents to conform with evolving standards in patent law regarding software inventions.
• Asserted Claims: Independent Claim 1 (Compl. ¶27).
• Accused Features: Plaintiff alleges that Defendant’s Accused System is a non-transitory computer program product that generates and stores a "locked evolving algorithm" for subsequent use, thereby infringing the ’879 Patent (Compl. ¶¶55, 58, 64).

III. The Accused Instrumentality

Product Identification

• The "Deep Radiology Imaging Technology" (the "Accused System") (Compl. ¶39).

Functionality and Market Context

• The complaint alleges the Accused System "enables image analysis based on product algorithms" (Compl. ¶39). The infringement allegations suggest the system is a computer program that is trained using user input to generate an algorithm, which is then used to automatically analyze subsequent images (Compl. ¶¶41-45, 47-53).
• The complaint does not provide sufficient detail for analysis of the Accused System's specific technical operation or its market context beyond these general allegations. Plaintiff makes its allegations based on Defendant's use of the system, "at least in internal testing" (Compl. ¶¶41, 47).
• No probative visual evidence provided in complaint.

IV. Analysis of Infringement Allegations

’854 Patent Infringement Allegations

• Identified Points of Contention:
  • Scope Questions: A central question may be whether Defendant’s "Deep Radiology Imaging Technology," likely a modern machine learning or AI system, constitutes an "evolving algorithm" as that term is used in the patent. The definition of this term will be critical to determining the scope of the claims.
  • Technical Questions: The complaint alleges infringement occurs "at least in internal testing and usage" (Compl. ¶44). This suggests a potential dispute over how the commercially available product actually operates and whether its training process meets the "in accordance with a user's judgment" limitation, which requires a specific form of user interaction.

’266 Patent Infringement Allegations

• Identified Points of Contention:
  • Scope Questions: Claim 1 recites a specific, multi-step "training mode." A potential point of contention is whether the Accused System's training process includes every recited step (e.g., presenting a first set, getting feedback, presenting a second set, getting approval).
  • Technical Questions: A key factual question may be whether the Accused System performs the step of "providing said product algorithm to at least one second user." This raises the question of whether a model trained by one user can be transferred to and used by a separate, distinct user within the defendant's system architecture.

V. Key Claim Terms for Construction

• The Term: "evolving algorithm"

  • Context and Importance: This term is central to all asserted claims and defines the core inventive concept of a learning-based system. Its construction will determine whether the claims read on modern AI and machine learning systems. Practitioners may focus on this term because its meaning in a patent with a 2001 priority date may not directly map to contemporary AI technology.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification suggests the term is not limited to a single implementation, stating the system may utilize "a set of evolving algorithms (e.g., Bayes' Theorem, a neural network, or any other image classification algorithm)" (’854 Patent, col. 6:7-10).
    • Evidence for a Narrower Interpretation: The specification consistently describes the algorithm's evolution as a direct result of iterative user feedback. For example, "User input during the evaluation can modify the evolving product algorithm" by being "used by the system to change the parameters defining a certain class of entities" (’854 Patent, col. 6:22-26). This could support a narrower definition requiring a specific user-driven, iterative refinement process.

• The Term: "user's judgment"

  • Context and Importance: This term qualifies how the "evolving algorithm" is generated and trained. The level of human interaction required by this term will be critical, particularly if the Accused System is highly automated or relies on pre-trained models.
  • Intrinsic Evidence for Interpretation:
    • Evidence for a Broader Interpretation: The specification describes user input that could be interpreted as minimal, such as when a user "select[s] an area of the image that contains an entity to be counted or classified" to define a "sample set of chromatic data points" (’854 Patent, col. 8:40-46).
    • Evidence for a Narrower Interpretation: The specification also describes a more substantive, corrective role for the user, where input "reflects the user's judgment about the accuracy of a classification" and is used to refine the algorithm over multiple iterations (’854 Patent, col. 10:16-19). This may imply a higher standard of cognitive input than simply providing initial training examples.

VI. Other Allegations

• Willful Infringement: The complaint does not contain an explicit count for willful infringement. It alleges that Defendant has had knowledge of the patents-in-suit "at least as of the service of the present Complaint" (Compl. ¶67). This allegation may form the basis for a claim of post-filing willful infringement but does not allege pre-suit knowledge.

VII. Analyst’s Conclusion: Key Questions for the Case

• A core issue will be one of definitional scope: can the term "evolving algorithm," as described in the context of the patents' 2001 priority date and their specific embodiments, be construed broadly enough to encompass the potentially more advanced AI or machine learning architecture of the "Deep Radiology Imaging Technology"?
• A key evidentiary question will be one of operational correspondence: what are the specific, factual steps involved in training and using the Accused System? Discovery will likely focus on whether this process maps onto the detailed, multi-step training modes and user interactions—including the transfer of an algorithm from a "first user" to a "second user"—as required by the asserted claims.