Study On Wireless And Portable 3D Ultrasound Imaging System

Compare with other medical imaging modalities, medical ultrasound(US) imaging is relative inexpensive, safe, portable and high resolution. However, two-dimensional(2D) US has a limited field of view, it can only display a cross sectional image, instead of the whole picture of the anatomy being investigated. What is more, 2D US are lack of reproducibility, its diagnosis results are highly depended on the experiences and skills of the clinician. Three-dimensional(3D) US overcomes the shortcomings of 2D US, and has a high value in many medical applications. Portable US play an important role in space-constrained environments such as wards. However, by limitation of the hardware, currently most portable US system does not provide 3D imaging, which has strongly limited the application of portable US.We proposed a wireless and portable 3D US imaging system in this study. This system is composed of a portable US scanner, a remote workstation, a Wi-Fi local network that connects the portable US scanner and the Ethernet and the client and server program running on the portable US scanner and the workstation respectively. The client program running on the portable US scanner captures the screen in real-time, compress the screenshots in JPEG format and transfers the compressed image data to the remote workstation via wireless network. The JPEG compression program is implemented with the algorithm defined in the CImage class of Microsoft foundation class library. JPEG compression can achieve a high image compression ratio and effectively reduce the size of transmitted data, which can help avoid network congestion. The server program running on the remote workstation handles the request from the client running in the portable US scanner, and receives image data. There is an optical character recognition(OCR) module and a B-scan image region detection module embedded in the server program. The OCR module is developed based on tesseract-ocr, an open source project. The OCR module can recognize the probe model and the imaging depth information printed on the screen of the portable US scanner. These processes can be done automatically, without human intervention. The 3D US imaging system proposed in this study use freehand scanning protocol, and estimate the relative position and orientation between adjacent image frames by means of speckle decorrelation tracking. Conventional freehand 3D US imaging systems depend on cumbersome position and orientation sensing device for 3D reconstruction, and requires non-trivial, time-consuming calibration procedure. However, the system proposed in this study does not require any position and orientation sensing device for 3D reconstruction, and is thus more portable.We examined the proposed system with a US resolution phantom and human subject. The volume reconstruction and distance measurement results show that the proposed system is feasible and has a reasonable accuracy.