Lukas Winter1, Haopeng Han1, and Thoralf Niendorf1,2,3
1Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrück Center for Molecular Medicine, Berlin, Germany, 2Experimental and Clinical Research Center (ECRC), a joint cooperation between the Charité Medical Faculty and the Max Delbrück Center for Molecular Medicine, Berlin, Germany, 3MRI.TOOLS GmbH, Berlin, Germany
Synopsis
Cost effective open source
imaging (COSI) is a collaborative initiative currently building an affordable
low field open source MR scanner with the technical documentation available at www.opensourceimaging.org.
As part of this initiative COSI Measure has been developed in order to
automatically map the static magnetic field. COSI Measure is an open source
multipurpose 3-axis system for ~3000€, which can be equipped with other field
mapping probes like electromagnetic field sensors, used for 3D printing / CNC
machinery application or other applications, that require programmable
submillimeter movement and sensor readouts in space and time.Purpose
MRI is the most powerful
diagnostic imaging modality available. The major drawbacks are cost and
complexity, limiting its use mainly to industrialized countries and larger
hospitals
1 . Multiple studies showed that there are several clinical
applications where the diagnostic value gathered from low (B0=0.2T) field
MRI is equivalent to costly “high” field (B0≥1.0T) systems
2-6. Cost effective open source
imaging (COSI) is a collaborative initiative currently building an affordable
low field open source MR scanner with the technical documentation freely
available at
www.opensourceimaging.org.
As part of this initiative COSI Measure has been developed in order to
automatically map the static magnetic field. COSI Measure is an open source multipurpose
3-axis system for ~3000€, which can be equipped with other field mapping probes
like electromagnetic field sensors, used for 3D printing / CNC machinery
application or other applications, that require precise programmable movement
and sensor readouts in space and time.
Mechanical Subsystem
The mechanical subsystem has been
designed with the free version of SketchUp (Trimble Navigation Ltd.) and is
displayed in Figure 1. A strong and robust setup was intended, to support positioning
the measurement system in safe distance to the magnetic field of an MRI, while
being able to use sensor probes that cover the distance between the driver and
the target. The base frame consists of (40x40)mm² aluminum profiles and has the
dimensions of (800x800x800)mm³. Four ball screws have been used with a 5mm thread
distance. Floating bearings have been used to support the mechanical movement.
Ball screws and floated bearings are mounted on aluminum (10mm thickness)
plates. Limit switch holders have been designed and 3D printed to control position
of the sensors and to keep cabling distances short (Figure 2).
Electronic Subsystem
The schematic of the COSI
multipurpose system is displayed in Figure 2. The heart of COSI Measure is a
Beagle Bone Black (BBB)with an 1GHz ARM cortex A-8 processor, 512MB DDR3 RAM
and embedded Linux. In addition the BBB is equipped with 2x PRU 32-bit
microcontrollers for real time control of the stepper motors. The BBB has 65 GPIO
pins allowing for flexible application, HDMI and USB port for standalone
application (monitor, keyboard, mouse) and an Ethernet port for remote
communication. In addition to the BBB we used a BeBoPr++ and a J5-BOB. The
BeBoPr++ is designed to work with the BBB, which can be mounted on top. It is
equipped with interfaces to connect up to 5 motor drivers, 6 limit switches and
thermos sensors for a possible CNC/3D-printing application. We used 4x NEMA23
stepper motors (3A@36V, 3NM holding torque, 1.8° per step at 200 steps per
evolution) together with 4x DM542A motor drivers, which allow 2-128x
microstepping leading to a theoretical precision limit of 200nm and half
holding current operation for power saving. The motors and drivers are supplied
by 2 AC/DC power modules (36V, 9.7A). Six inductive limit switches (4mm
detection range) are used to detect the working range of COSI Measure. An
emergency power off button has been included. Control of the system can be either
performed over a GUI or G-Code. For magnetostatic field measurements a 3-axis
Gaussmeter (460, Lakeshore) was used.
Results
The working volume of COSI
measure is (50x50x60)mm³. It can be easily adjusted changing the dimensions of
the aluminum profiles and ball screws. In order to measure backlash and precision,
a ball pen was attached to the probe holder (Figure 1). The system supports
submillimeter movement as shown in Figure 4a-b with no visible backlash (Figure
4b). The submillimeter vibrations are due to duct tape fixation of the ball
pen, which can be conveniently addressed with a customized probe holder. The
current maximum speed of the system is 0.4m/s is more than sufficient for field
mapping applications. The limiting factors for the speed are the mechanical
motor couplers connecting the stepper motors to the ball screws, which will be
exchanged for future applications since the electronic subsystem supports much
higher. Furthermore the motors and current system are strong enough to easily
lift >3kg, which is more than sufficient for field probes. The maximum
strength is yet to be determined after the motor couplers have been exchanged.
Conclusion
COSI Measure is a programmable
multipurpose 3-axis robotic system, which can be utilized for magnetostatic
field measurements, electromagnetic field measurements, 3D printing, CNC
machinery and various other applications, that require reproducible
submillimeter movement and/or measurements. Design files, software, technical
documentation, part list etc. will be made available according to the values of
open source hardware at
www.opensourceimaging.org.
References
[1] Rylands-Monk F, Aunt Minnie Europe, 2015 [2] Pääkkö E, Eur Radiol, 2005; 15(7):1361-8 [3] Ejbjerg B, Ann Rheumat Dis, 2005; 64(9):1280-7 [4] Kersting-Sommerhoff B, Eur Radiol, 1996; 6(4):561-5 [5] Parizel PM, Eur J Radiol, 1995; 19(2):132-8 [6] Jhaveri K, JMRI, 2015; 41(4):866-9