Applications of Continuous Wavelet Transforms to Optical Coherence Tomography

  • 01



Name of the Speaker: P. Naveen kumar (EE19D015)
Name of the Guide: Dr. Shanti Bhattacharya
Date/Time: September 1st, 2.30p.m

Optical coherence tomography (OCT) is a well-known imaging technology in the biomedical field and a popular commercial technology in ophthalmology. In terms of image resolution, it is superior to ultrasound imaging but has limited access to objects well below the surface. Given that it has a few micrometers of longitudinal and lateral resolution, it can also be used in non-biomedical applications such as sub-surfacing imaging of fabricated semiconductor chips, historical art conservation, agriculture sciences, etc. There are always certain trade-offs to be kept in mind when using a technology for a specific application, as the design should meet the imaging requirements.

OCT can be used in different forms. We look at Spectral Domain or SDOCT rather than time domain OCT because of its high speed imaging and sensitivity. This made it suitable for non-invasive imaging applications which require faster data acquisitions. However, SDOCT has constraints such as restricted imaging depth, complex conjugate ambiguity (CCA), and longitudinal spatial resolution associated with the spectrometer detection method. Fourier-based discrete signal processing is the cause of the aforementioned issues. Since the post-processing techniques for reconstructing the image from the OCT signal are essential in establishing the picture resolution, spectral domain OCT demands new data processing techniques to make the best use of the detector. To address some of these issues, we explore the use of continuous wavelet transforms (CWT) for the OCT image reconstruction process.

In this talk, we will discuss how the wide wavelet theory and its characteristics have provided us with a means to investigate a filter bank technique for image reconstruction in OCT. The adjustable frequency characteristics of the mother wavelet and the ease of designing a reconstruction filter bank over the required bandwidth have made us utilize the full detector resolution in SDOCT. If an N pixel detector is chosen to generate the axial scan (a-scan) of the sample being imaged, the Fourier technique provides us with useful data over N/2 pixels, whereas the CWT method outputs the same data over N pixels. The simulations with CWT algorithms have shown an improvement in longitudinal spatial resolution by a factor of 2 when compared to Fourier domain methods. CWT doesn't compute the negative frequencies and DC content in the analysed signal. This eliminates redundant data in the output image and also improves the contrast ratio between the actual layers in the sample. This could enhance the visual information in the compound image constructed from a set of clustered a-scans. We intend to design an SDOCT system at a 1310 nm wavelength, which could meet the imaging requirements for disease identification in agriculture.