Formation Algorithms for Low-Cost Freehand Ultrasound Scanner Based on Ego-Motion Estimation and Unsupervised
ClusteringConventional ultrasound devices use multi-element piezoelectric transducer arrays to produce 2D or 3D images.
- Philips Lumify – 200 USD/month per probe, plus 75 USD/month for guarantee.
- Sonosite IVIZ–USD 10,000.
- GE VSAN Extend–starts at USD 2995.
- Clarius–starts at USD 6900.
- Butterfly IQ–USD 1999 plus 420 USD/year for the cloud-based user license.
There are a variety of firms (such as GE Healthcare, Koninklijke Philips N.V., Canon Inc., Hitachi, Siemens Healthcare, Samsung Medison Co., Ltd., and Esaote) that are continuously looking to introduce new ultrasound systems that incorporate the latest technology in the market. GE Healthcare dominated the ultrasound industry and had the most significant market share in 2020. 8 ].
Many big companies are shifting their focus to developing portable ultrasound equipment that one person can easily transport from one location to another. On the 15th of November, 2021, TELEMED Medical Systems presented MicrUs Pro Handheld USB Ultrasound, their most recent ultrasound diagnostic system based on PC, at the MEDICA 2021 conference. The MEDICA conference in 2021 was an excellent opportunity for TELEMED to showcase its 10 ]. The mechanical sensors that are typically used to monitor the ultrasound probe’s location create a device that is bulky and difficult to operate. So, these sensors weren’t used for this particular probe. Thus, producing a high-quality and well-formed ultrasound image with a lower cost depends on using sophisticated algorithmic techniques for processing signals. This article thoroughly describes the image-making algorithms needed to create a correctly geometric 2-D B-mode ultrasound image using a standard ultrasound probe. It does not comprehensively describe the probe’s design for hardware.
Please refer to 9 to understand the theories utilized to minimize the hardware complexity within the probe prototype. The ego-motion estimation strategies used in the earlier work for understanding some of the concepts used to reduce hardware complexity 11 Together with modern signal processing methods explained in the paper, use advanced signal processing algorithms to determine the linear motion of the probe. The probe’s position estimations help create a logically correct 2D B-mode ultrasound image. A new ego-motion estimation algorithm is suggested based on the correlation measurements between the successive line echo scans (1D records). This provides information on the probe’s direction during a linear scan as it passes through the skin. The velocity estimates assist in mapping images into a series of geometrically correct pixels.
The ego-motion estimation method depends on the focus parameters of the transducer and the speckle or texture detail, as well as how echogenic the tissues are. J. F. Chen et al., in [ 18 We have calculated the decorrelation measurements based on the speckle pattern underneath when the transducer is moved through the elevation focus to measure the transducer’s movement within the elevation plane. A. Krupa et al. have also demonstrated using speckle information in the images for tracking both in-plane and out-of-plane motions. The authors have also demonstrated using speckle information to track both out-of-plan 19 ].