18581848. ULTRASOUND SYSTEM WITH A NEURAL NETWORK FOR PRODUCING IMAGES FROM UNDERSAMPLED ULTRASOUND DATA simplified abstract (KONINKLIJKE PHILIPS N.V.)
Contents
ULTRASOUND SYSTEM WITH A NEURAL NETWORK FOR PRODUCING IMAGES FROM UNDERSAMPLED ULTRASOUND DATA
Organization Name
Inventor(s)
Christine Menking Swisher of San Diego CA (US)
Jean-Luc Francois-Marie Robert of Cambridge MA (US)
ULTRASOUND SYSTEM WITH A NEURAL NETWORK FOR PRODUCING IMAGES FROM UNDERSAMPLED ULTRASOUND DATA - A simplified explanation of the abstract
This abstract first appeared for US patent application 18581848 titled 'ULTRASOUND SYSTEM WITH A NEURAL NETWORK FOR PRODUCING IMAGES FROM UNDERSAMPLED ULTRASOUND DATA
The present disclosure describes medical imaging systems and methods configured to generate medical images based on undersampled imaging data. The images are generated using a neural network trained with samples of known fully sampled data and undersampled data derived from the known fully sampled data applied to acquired sparsely sampled data.
- Neural network trained with known fully sampled data and undersampled data
- Training involves adversarial generative network with a generator and discriminator
- Generator trained with known undersampled data to generate estimated image data
- Generator capable of generating data that the classifier cannot differentiate as real or fake
- Trained generator applied to unknown undersampled data
Potential Applications: - Improving image quality in medical imaging - Enhancing diagnostic capabilities - Reducing the need for extensive imaging data collection
Problems Solved: - Addressing challenges of generating high-quality images from undersampled data - Improving efficiency in medical imaging processes
Benefits: - Enhanced image reconstruction accuracy - Faster imaging processes - Reduced data collection requirements
Commercial Applications: Medical imaging equipment manufacturers can integrate this technology into their systems to offer more efficient and accurate imaging solutions to healthcare providers.
Questions about Medical Imaging Systems with Undersampled Data:
1. How does the neural network training process improve image reconstruction accuracy?
- The neural network training process enhances image reconstruction accuracy by utilizing known fully sampled data and undersampled data to generate high-quality images.
2. What are the potential implications of using this technology in medical imaging practices?
- The use of this technology can lead to faster and more accurate diagnostic processes, ultimately improving patient care outcomes.
Original Abstract Submitted
The present disclosure describes medical imaging systems and methods configured to generate medical images based on undersampled imaging data. The images may be generated by applying a neural network trained with samples of known fully sampled data and undersampled data derived from the known fully sampled data to a acquired sparsely sampled data. The training of the neural network may involve training adversarial generative network including a generator and a discriminator. The generator is trained with sets of known undersampled data until the generator is capable of generating estimated image data, which the classifier is incapable of differentiation as either real or fake, and the trained generator may then be applied to unknown undersampled data.
