System and methods for reducing anomalies in ultrasound images

A system, comprising:

• an ultrasound probe configured to emit an ultrasonic pulse and generate ultrasound image data corresponding to reflections of the ultrasonic pulse;
• memory storing instructions;
• and
• a processor electronically coupled to the memory and when executing the instructions is configured to:
• filter the ultrasound image data into a plurality of sub-band components;
• combine the plurality of sub-band components to attenuate anomalies in a first ultrasound image formed by the combination of the plurality of sub-band components, wherein the plurality of sub-band components are combined based on a plurality of adaptive weights output from a machine learning model and wherein the plurality of adaptive weights are associated with the plurality of sub-band components;
• and
• output for a display device, the first ultrasound image.

A method comprising:

• filtering ultrasound image data into a plurality of sub-band components;
• in a machine learning model, adaptively scoring image quality of the plurality of sub-band components and weighting the plurality of the sub-band components using the image quality scoring;
• compounding the plurality of weighted sub-band components to attenuate anomalies in a first ultrasound image formed from the plurality of weighted sub-band components;
• and
• outputting for a display device, the first ultrasound image.