Using NI LabvIEW State Diagram Toolkit to Create Integrated Motion Control and Data Acquisition for Ultrasound Critical-Angle Reflectometry

Author(s):
Matthew A. Lewis - UT Southwestern Medical Center at Dallas
Peter P. Antich - UT Southwestern Medical Center at Dallas
Edmond Richer - UT Southwestern Medical Center at Dallas

Industry:
Medical/ Medical Instrumentation, Life Science

Products:
Motion Control, LabVIEW

The Challenge:
Developing an integrated solution that combines motion control of a six-axis positioning robot with multichannel ultrasound data acquisition in a system for evaluating bone quality in the human calcaneus (heel bone).

The Solution:
Using the National Instruments LabVIEW State Diagram Toolkit to interface motion control and data acquisition components of the ultrasound critical-angle reflectometry system.

Ultrasound Critical-Angle Reflectometry

The current gold-standard technique for assessing bone in osteoporosis and a variety of other disease states is dual photon x-ray absorptiometry (DEXA). Aside from using ionizing radiation, this technique is sensitive only to the mineral content in bone and provides no information on the anisotropy of bone: the fact that bone is a solid with mechanical properties that differ depending on direction.

In comparison, the reflection ultrasound technique known as ultrasound critical-angle reflectometry (UCR – a technique developed in the laboratory of P. Antich, UT Southwestern) is capable of noninvasively assessing the anisotropy of human bone, and we hypothesize that this technique will contribute to a more complete, quantitative understanding of bone quality (the ability of bone to resist fracture).

Unlike clinical ultrasound, which uses echolocation of reflections from tissue boundaries, UCR is based on the analysis of the reflection of high-frequency sound waves from the surface of bone at varying angles of incidence. In general, an ultrasonic wave focused on the surface of a solid at some angle θ generates two refracted waves: one pressure and one shear. As the angle of incidence θ increases, the refracted pressure wave in the solid eventually is total-internally reflected, a phenomenon that can be observed as a critical-angle in the reflection spectrum from the solid. Using this angular information and Snell’s law for the refraction of waves at material boundaries, we can calculate the speed of sound in the solid. It is well-established that the elasticity of a solid is closely related to the speed of sound, and bone elasticity is well-correlated with bone breaking strength (bone quality).

System Overview

Our latest UCR devices are based on a novel 48-channel receiver array encased in a water-filled sensor with a cylindrically focused transmitter. The 48-channel sensor is coupled to a chassis of custom amplifiers and 40 MHz digitization channels that are accessible with LabVIEW using the Call DLL Library VI. On the transmitter side, an NI PCI-5411 arbitrary waveform generator and appropriate NI-FGEN VIs drive the pulse generation through a high-power RF amplifier. Precise frequency and pulse control is crucial when working with multilayer materials such as bone, which is comprised of a dense, cortical outer layer and spongy, trabecular core. By varying sampling time, input frequency, and pulse shape, it is possible to measure the mechanical properties of bone as they vary in depth.

A key step in the UCR method is the alignment of the sensor head at a right angle to a plane tangent to the bone surface. To accomplish this, we developed a six-degree-of-freedom robotic platform for precise kinematics about the bone surface. Three linear axes controlled by an NI PCI-7344 motion controller provide translation of the focal point of the ultrasonic sensor to the anatomical site and in depth to the bone. Two of these axes provide arbitrary tilt and pitch of the UCR sensor head to find normal alignment. The third axis allows the normally aligned sensor to rotate in place on the bone surface. This latter procedure facilitates measurement of mechanical properties in different planes of interest, and using this information we can quantitate the anisotropy of bone in the human heel bone (calcaneus).

Creating a Single-Application System

In our previous generation UCR system, a general-purpose motion control program written in LabVIEW 6 with Flexmotion VIs was controlled by a separate, custom data acquisition LabVIEW VI. Using the VI Server mechanism, the data acquisition program simulated user interaction with the motion control program interface to carry out certain data acquisition routines such as those associated with analysis of anisotropy. The performance of this approach was especially sensitive to tight inner While loops in either program and also to the Execution settings of either VI.

In developing our latest, site-specific (heel only) system, we recognized the need for optimization of the application software, especially because we needed to build multiple systems for testing at other institutions. We needed a more rigorous course of software engineering. Available solutions for this problem included a client-server model with message passing. However, in the interest of simplification and code documentation, we used the LabVIEW State Diagram Toolkit to organize and simplify the functionality of our primary UCR application.

Using the LabVIEW State Diagram Toolkit, we rapidly developed an application that integrated motion control and data acquisition components. After initialization of NI-Motion software, NI-FGEN, and proprietary data acquisition components, the application enters a default loop in the state diagram in which data acquisition and motion control oversight receive equal time sharing. Although many acquisition sequences are automated, this default loop in the state diagram is important because initial localization of the UCR sensor focus to a point in the neighborhood of the bone surface is performed under guidance of an ultrasound technician. The operator can simultaneously view the reflection data from the bone while manually controlling the translations and rotation of the UCR sensor.

Once the bony surface is manually identified, the operator can direct the system into a variety of other states in the state diagram, including a routine for automatically finding the surface normal, a routine for collecting data, a variety of specialty motions for fine positioning, and a variety of maintenance states. Users can implement new functionality as new states in the LabVIEW State Diagram Toolkit. For example, we implemented coupling of linear and angular axes so that depth could be varied in arbitrary directions in a special motion state that calculates the relevant coordinates. Complex rotations using custom quaternion VIs are also generated in specific states.

LabVIEW Toolkit Eases Development

In summary, the LabVIEW State Diagram Toolkit facilitated and eased the development of a biomedical application that included complex motion control and data acquisition.

For more information, conact:

Matthew A. Lewis

Assistant Professor

UT Southwestern Medical Center at Dallas

5323 Harry Hines Blvd. MC 9058

Dallas, TX 75390

Tel: (214) 648-3659

Fax: (214) 648-2991

E-mail: mal11@po.cwru.edu

Browse All Case Studies »