Apparatus and method for generating output data as a function of digital image data and an analysis module
Inventors
Ranganathan, Mohan Krishna • Ip, Wui • AGARWAL, Animesh • Kant, Shashi • Barve, Rakesh • SINGH, SHIV PRATAP
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-12507936-B1
Publication Date
2025-12-30
Expiration Date
Abstract
An apparatus and method for generating output data as a function of digital image data and an analysis module. The apparatus includes at least a processor and a memory communicatively connected to the at least a processor. The memory instructs the processor to present a graphical user interface comprising medical time-series data, receive a capture event as a function of a first user command, generate, as a function of the capture event, digital image data, wherein the digital image data comprises a window of the graphical user interface, verify, using user confirmation, the window of the graphical user interface, process, using one or more analysis models, the digital image data, wherein processing the digital image data comprises extracting at least a cardiac metric from the digital image data, and generate, using the analysis module, output data based on the at least a cardiac metric extracted from the digital image data.
Core Innovation
The disclosed apparatus and method present medical time-series data in a graphical user interface, generate digital image data representing an interface window upon a capture event, and verify the window using user confirmation. Analysis models extract cardiac metrics from the digital image data. An analysis module then generates a classification using cardiac metric labels associated with diagnostic conditions, alongside confidence scores derived from comparison with historical labeled digital image data containing annotated cardiac images.
The disclosure depicts machine-learning processes including a data classifier, a feedforward neural network, a weighted-sum neuron, and an electrocardiogram algorithm. Because background, abstract, and summary texts are omitted from the supplied content, a specific technical problem or deficiency in prior systems is not explicitly described in patent. The system functions primarily to generate cardiac classifications and confidence scores directly from captured interface images.
Claims Coverage
The claims define two inventive features embodied in two independent claims directed to an apparatus and a method.
Apparatus for image-based cardiac classification
A computing device presents medical time-series data in a graphical user interface, generates digital image data of an interface window in response to a capture event, and verifies the window using user confirmation. Analysis models extract cardiac metrics from the image data, and an analysis module generates a classification using cardiac metric labels associated with diagnostic conditions and produces confidence scores by comparing the labels with historical labeled digital image data containing annotated cardiac images with verified classifications.
Method for image-based cardiac classification
The method presents medical time-series data in a graphical user interface, generates and verifies digital image data of an interface window in response to a capture event, and extracts cardiac metrics from the image data using analysis models. The method generates a classification using cardiac metric labels associated with diagnostic conditions and produces confidence scores by comparing the labels with historical labeled digital image data containing annotated cardiac images with verified classifications.
The two independent claims cover apparatus and method implementations for capturing and user-verifying an interface window displaying medical time-series data, extracting cardiac metrics, and producing diagnostic classifications with confidence scores based on historically labeled images.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Generating classifications associated with diagnostic conditions from extracted cardiac metrics, with confidence scores based on comparison to historical labeled cardiac images with verified classifications.
Representing cardiac indicators related to heart rate variability, arrhythmia patterns, or ischemic event detection.
Generating a remote-device alert when a cardiac metric deviates from a predefined threshold.
Storing image data, extracted cardiac metrics, and classifications in a structured database, and transmitting image data and corresponding metrics to a third-party system.
Interested in licensing this patent?