Deep Learning-Based Phase Unwrapping for Diverse Applications

Publication ID: 24-11857288_0007_PTD
Published: November 07, 2025
Category:New Applications & Use Cases

Legal Citation

pr1or.art Inc., “Deep Learning-Based Phase Unwrapping for Diverse Applications,” Published Technical Disclosure No. 24-11857288_0007_PTD, Published November 07, 2025, available at https://archive.pr1or.art/24-11857288_0007_PTD
This technical disclosure describes improvements that would be readily apparent to a Person Having Ordinary Skill In The Art (PHOSITA) when considered in combination with the foundational architecture disclosed in U.S. Patent No. 11,857,288.

Summary of the Inventive Concept

The inventive concept applies the core technology of deep learning-based phase unwrapping for dense MRI to entirely new industries, including agriculture, construction, industrial processes, pharmaceuticals, and neuroscience.

Background and Problem Solved

The original patent addressed the challenge of myocardial strain imaging in cardiac MRI using deep learning-based phase unwrapping. However, this technology has broader applications beyond cardiac imaging. The inventive concept solves the problem of limited application scope by extending the technology to various fields, enabling new use cases and applications.

Detailed Description of the Inventive Concept

The inventive concept leverages the deep learning-based phase unwrapping methodology to analyze displacement encoded MRI data in diverse contexts. For instance, in agriculture, the technology can monitor and predict crop yields by analyzing soil moisture and crop growth. In construction, it can detect early signs of structural damage in buildings by analyzing displacement encoded MRI data of building materials and structures. Similarly, in industrial processes, it can analyze fluid dynamics, and in pharmaceuticals, it can monitor product consistency and purity. In neuroscience, it can non-invasively monitor brain activity and function.

Novelty and Inventive Step

The inventive concept's novelty lies in its application of deep learning-based phase unwrapping to entirely new fields, which was not anticipated by the original patent. The inventive step is the recognition of the technology's broader applicability and its adaptation to diverse industries, resulting in new and non-obvious use cases.

Alternative Embodiments and Variations

Alternative embodiments of the inventive concept could include variations in the deep learning architecture, the type of MRI data used, or the specific application domains. For example, using different CNN structures or incorporating additional sensors could further enhance the technology's capabilities.

Potential Commercial Applications and Market

The inventive concept has significant commercial potential in various industries, including agriculture, construction, industrial processes, pharmaceuticals, and neuroscience. The target market includes companies and research institutions seeking innovative solutions for monitoring and analysis in these fields.

Field of Art

Medical imaging, machine learning, and signal processing, with expertise in deep learning techniques for medical image analysis, particularly magnetic resonance imaging (MRI) and phase unwrapping algorithms

Person of Ordinary Skill (PHOSITA) Profile

A skilled practitioner with advanced degrees in biomedical engineering, computer science, or medical imaging, possessing expertise in convolutional neural networks, signal processing, and MRI data analysis techniques

Obviousness Rationale

A person having ordinary skill in the art would recognize that the deep learning-based phase unwrapping technique disclosed in the source patent represents a generalizable signal processing methodology that could be readily adapted to various domains requiring displacement or phase analysis. The core technical innovation of using convolutional neural networks to decode complex signal phase information is fundamentally transferable across different application contexts. The PTD demonstrates that the underlying machine learning approach can be systematically applied to analyze displacement-encoded data in multiple industries by leveraging the same core computational principles.

Obvious Combinations & Variations

Source Patent Element
U-Net structured CNN for computing wrapping label map in medical imaging
PTD Variation
Applying U-Net CNN to analyze displacement in agricultural soil moisture monitoring
Obviousness Reasoning
The neural network architecture is a known technique for signal processing, and adapting it to different input domains represents a predictable extension of the original technology
Source Patent Element
Displacement encoding technique for capturing complex motion data
PTD Variation
Using similar displacement encoding to analyze structural deformation in construction materials
Obviousness Reasoning
The fundamental signal processing methodology is transferable across domains, with the core technical challenge of phase unwrapping remaining consistent
Source Patent Element
Deep learning segmentation of image data into distinct classes
PTD Variation
Extending class segmentation to industrial fluid dynamics and pharmaceutical product analysis
Obviousness Reasoning
The machine learning approach of classifying complex signal data is a known technique that can be systematically applied to different input domains
Source Patent Element
Cardiac tissue motion analysis using MRI displacement encoding
PTD Variation
Applying similar displacement analysis techniques to neural tissue and brain activity monitoring
Obviousness Reasoning
The technical principles of motion and displacement tracking are fundamentally similar across biological tissue types, making the extension predictable
35 U.S.C. § 103 Summary: Based on the teachings of US Patent 11857288, a person having ordinary skill in the art would find the variations disclosed in the Published Technical Disclosure to be obvious extensions of the core deep learning-based phase unwrapping methodology. The systematic application of convolutional neural network techniques to analyze displacement-encoded data across multiple domains represents a predictable technological progression that would be readily conceived by an ordinarily skilled practitioner in medical imaging and machine learning.

Original Patent Information

Patent NumberUS 11,857,288
TitleSystems and methods for phase unwrapping for dense MRI using deep learning
Assignee(s)University of Virginia Patent Foundation