Jayaseelan Marimuthu

In this studies a prototype low cost portable handheld microwave system (MiS) designed and tested for back fat depth measurements using non-invasive techniques on lamb carcase in commercial abattoirs. For this application, two different type antennas was designed and tested. These results demonstrate the capacity of this prototype MiS system together with proposed antennas to estimate fat depth at the C site in lamb carcasses non-invasively in commercial abattoir at rate 100 carcase in 30 minutes.

Non-invasive and non-destructive measurements of fat depth are sought after within the beef industries, as overfat carcases cause significant economic loss and wastage for processors. Within the Australian beef industry fat depth is measured manually, however this has the disadvantage of being destructive, subjective and time-consuming [1]. Since biological tissues in animals, feature a high contrast in their dielectric properties (skin, fat, and muscle) at microwave frequencies [2], this represents an opportunity to different fat from lean and thus estimate fat depth and body composition of carcases and live animals. This concept has been tested in lamb carcases where a low-cost portable Microwave System has been developed for measuring C-site back fat depth [3]. This paper details the testing of a similar device in beef, where we hypothesised that it would accurately predict P8 and rib fat depth in beef carcases.

In the meat industries, overfat carcases cause significant economic loss due to the labour required for trimming fat, and the waste that it represents. Fat is the most variable component, both in its amount and distribution in the carcase, and on this basis the measurement of carcase fat depth is the cornerstone of most carcase classification schemes for beef, lamb and pork worldwide [1,2]. Fat depth is often measured manually, however this has the disadvantage of being destructive, subjective and time-consuming. The ability to estimate fat depth accurately via a non-invasive and non-destructive technique is therefore highly sought after [3]. One solution showing considerable promise for determining carcase fatness is a Microwave System (MiS) using low power non-ionizing electromagnetic waves [4]. Since biological tissues in animals, feature a high contrast in the dielectric properties (skin, fat, muscle and bone) at microwave frequencies [4], MiS can accurately evaluate the fat depth and body composition of carcases and live animals. As an illustration of tissue parameters, the range of contrast available in X-ray imagery within soft tissues is less than 2%; whereas within microwave electromagnetic fields the range in relative dielectric constant goes from a minimum of about 4 in fat to a maximum of about 70 in muscle [5]. Furthermore, the microwave devices required to produce and measure these fields are low cost and highly portable. Working within these constraints a low-cost portable Microwave System has been developed for measuring back fat depth in lamb carcases. This paper details early testing of this device, testing the hypothesis that it will provide a reliable estimate of back-fat depth in lamb carcases.

Microwave biomedical imaging has the potential to be a future complementary diagnostic technique for cost-critical situations. The accessibility and portability of diagnostic imaging can mean accurate information is available where it is needed. The feasibility of microwave imaging has been demonstrated by using lab equipment such as vector network analyzers, which, whilst they are accurate, they are also large and expensive. Capturing the required signal responses can instead be performed by using software defined radio (SDR) technology, which due to its high manufacturing volumes can have a cost which is several orders of magnitude lower than a VNA. Although the performance of such a system might be lower than that of a full-sized VNA, the performance is still adequate for biomedical imaging applications. In this paper, an SDR-based system's performance is analyzed and shown to accurately detect an abnormality within a head phantom.

Microwave medical imaging is an attractive complement to current diagnostic tools for medical applications due to its low-cost, portability and non-ionization radiation. In this paper, generic software defined radio (SDR) technology is used to perform biomedical measurements, paving the way for low-cost, compact and reconfigurable imaging systems. SDR technology is capable of wide band instantaneous spectrum capture, in our case 20 MHz at a time in the 0.3 - 3.8 GHz band. To provide the resolution required in medical imaging, it needs to combine multiple frequency spectrums together. Using this as the basis of the system, a cylindrical scanning system and directional antenna are used to measure a biological phantom with multiple targets inside. The system is able to detect the two targets with varying spacing.

A low-cost microwave medical imaging system, which uses software defined radio (SDR) technology, is presented. SDR technology has long been used to quickly develop and prototype new communication system applications. Due to its generic nature, it has the potential to be mass-produced lowering the cost. Here, we re-purpose SDR technology to perform near-field biomedical radar with the use of a switching networking and an array of antennas. To verify its operation, two different antenna arrays using the frequency range from 1-4 GHz are used. The results show that using low-cost technology it is possible to successfully image a bio-mimicking phantom.

A compact bandpass filter with wide stopband and multiple transmission zero is presented. The device uses a low impedance feeding network, folded parallel coupled microstrip lines (FPCML) with stepped impedance and two capacitive coupled open stubs. The low impedance feeding network is utilized to enhance the coupling of the PCML structure and to achieve sharp lower and upper cut-off frequencies, while the folded open stub embedded with high/low impedance at end of PCML structure able to extend the upper stopband. Meanwhile, the two capacitive coupled open-ended stubs are able to produce multiple transmission zeros at the stopband to enhance the rejection response and achieve harmonic suppression. The proposed filter with a centre frequency of 2.45 GHz is designed and optimized using ADS Momentum. The filter shows excellent performance with 0.1 dB insertion loss and more than 15 dB attenuation in the wide stopband, with 20% fractional passband.

A low-cost, portable and reconfigurable Software Defined Radar (SDRadar) based is proposed for medical imaging. The proposed system can replace the commercial VNA, which is currently used in experimental microwave imaging systems. The proposed SDRadar has the capability to produce synthetic stepped-frequency continuous wave, frequency modulated continuous waveforms (FMCW) and pseudo random noise waveforms that can be used for non-linear and time variant medium. This will lead to a vital opportunity to improve the time, resolutions and accuracy in microwave-based medical imaging. Furthermore, the proposed device with hybrid waveforms can be designed for better imaging of complex human tissues.

Software defined radar (SDRadar) is investigated as a possible transceiver for microwave-based medical imaging applications. In radar-based microwave imaging, wideband pulses are synthetically created using a stepped frequency continuous wave (SFCW) varying over the entire wideband frequency range. Although the resolution is typically low compared with other tools, such as X-ray, microwave imaging is gaining popularity due to low health risk, and low-cost compared to conventional medical imaging systems. Recent works show the feasibility of this technique by employing commercial vector network analysers (VNA); however, using VNA results in a rigid, bulky and expensive system. To realize a mass screening diagnostic tool, a low-cost portable system based on SDRadar as a transceiver is proposed. The SFCW-SDRadar prototype is implemented using both open source software and hardware. The software part of the radar is realized using GNU Radio, whilst the hardware part is implemented using bladeRF.

Broadband bandpass filter with compact size, wide stopband and sharp upper cut-off band is presented. The filter is designed to cover the band that can be used in microwave-based head imaging for stroke detection. The proposed filter uses a low impedance feeding network, a pair of folded parallel coupled microstrip line, a rectangular microstrip resonator, and cross coupled open-ended stubs. A prototype is designed with bandwidth 1.9 GHz (centered at 2.45 GHz) and more than 15 dB attenuation at the stopband that extends up to 20 GHz. The rate of the upper cutoff frequency is more than 140 dB/GHz.