Systems and methods for standardization of electrocardiogram signal images

Inventors

BADE, Sairam • Barve, Rakesh • Mishra, Yash • Prasad, Ashim • Kant, Shashi • Sharma, Mayank • Dodle, Durgaprasad

Assignees

Anumana Inc

Abstract

An apparatus for standardization of electrocardiogram signal images, the apparatus having an imaging device, at least a processor and a memory communicatively connected to the at least a processor, the memory containing instructions configuring the at least a processor to receive an overlay image from the imaging device, identify a captured fixed background image and a captured primary image within the overlay image, wherein the captured primary image includes a plurality of electrocardiogram signals, compare the captured fixed background image to one or more image quality thresholds, determine an image quality score of the primary image as a function of the captured fixed background image, and output one or more image modification datum as a function of the image quality score.

Core Innovation

The invention relates to standardization of electrocardiogram signal images using an imaging device that receives an overlay image. Within the overlay image, a captured fixed background image and a captured primary image are identified, where the captured primary image comprises a plurality of electrocardiogram signals. The captured fixed background image is used as the reference basis for quality evaluation of the captured primary image.

The approach compares the captured fixed background image to one or more image quality thresholds. Based on the comparison, an image quality score of the captured primary image is determined as a function of the comparison. This image quality score is then used to drive the production of output data intended to standardize the electrocardiogram signal images.

The invention outputs one or more image modification datum as a function of the image quality score. In particular embodiments, the one or more image modification datum are generated using a modification machine learning model trained on modification training data that correlates image quality scores with image modification data. Iteratively, the image modification datum may be generated to produce an updated overlay image, an updated fixed background image, and an updated primary image, followed by determining an updated image quality score using the updated fixed background image.

Claims Coverage

The independent claims are an apparatus claim and a method claim that share the same core pipeline: receive an overlay image, identify a captured fixed background image and a captured primary image with a plurality of electrocardiogram signals, compare the captured fixed background image to one or more image quality thresholds to determine an image quality score, and output one or more image modification datum as a function of that score. The dependent claims refine the fixed background content and the comparison/processing structure, including additional acquisition structure and ML-based iterative updates.

Across the independent claims, image standardization is achieved by using an overlay image to obtain a captured fixed background image and a captured primary image containing a plurality of electrocardiogram signals, comparing the fixed background to image quality thresholds to compute an image quality score, and outputting image modification datum based on that score. Dependent inventive features further specialize the fixed background content, the threshold-comparison mechanism, and the generation of modification outputs, including modification-machine-learning-model training and iterative updated overlay capture and re-scoring.

Stated Advantages

Documented Applications