The researchers of Indian Institute of Technology (IIT) Madras, studying microwave inverse imaging, said the technique has wide applications in early breast cancer detection, medical imaging, non-destructive testing and evaluation, and concealed weapon detection at security checkpoints.
Microwave imaging is an advanced version of radar technology, and the technique is used to image hidden or embedded objects in a structure, using electromagnetic waves in the microwave frequency range, according to an official statement issued by IIT Madras. The findings of the research team comprising Dr Uday K Khankhoje, Associate Professor of Department of Electrical Engineering, IIT Madras, and Chandan Bhat, a research scholar, IIT Madras have been published in the reputed peer-reviewed journal ‘IEEE Geoscience and Remote Sensing Letters’.
In microwave imaging, traditionally, a reference set of measurements are taken without the object in the imaging setup, then the object or patient is introduced and a new set of measurements are taken. By taking the difference between these sets of measurements, a Physics-based algorithm computes the physical properties of the object, and this final step produces a microwave ‘image’ of the object. However, the need for two sets of measurements increases the overall imaging time. In some situations, during the detection of objects buried under the ground — using a ground penetrating radar — it is not possible to get a reference set of measurements, simply because extracting a buried object can be impractical, said researchers.
Explaining the ideas underlying this research, Dr Khankhoje said, “…Since we can carry the ‘image’ process without a prior set of reference measurements without the object, it becomes possible to extend the techniques of microwave imaging to new applications, such as underground imaging using ground penetrating radars, and to speed up applications like breast cancer imaging.”
Subscriber Only Stories
Bhat said, “Many applications, such as indoor scattering and breast cancer detection, require two sets of measurements. The proposed approach provides a solution to reduce the data needed for a general scenario…”