Reconstruction of optical vector-fields with applications in endoscopic imaging

Milana Gataric,George S. D. Gordon,Francesco Renna,Alberto Gil C. P. Ramos,Maria P. Alcolea,Sarah E. Bohndiek

DOI: https://doi.org/10.1109/TMI.2018.2875875

2018-07-18

Abstract:We introduce a framework for the reconstruction of the amplitude, phase and polarisation of an optical vector-field using calibration measurements acquired by an imaging device with an unknown linear transformation. By incorporating effective regularisation terms, this new approach is able to recover an optical vector-field with respect to an arbitrary representation system, which may be different from the one used in calibration. In particular, it enables the recovery of an optical vector-field with respect to a Fourier basis, which is shown to yield indicative features of increased scattering associated with tissue abnormalities. We demonstrate the effectiveness of our approach using synthetic holographic images as well as biological tissue samples in an experimental setting where measurements of an optical vector-field are acquired by a fibre endoscope, and observe that indeed the recovered Fourier coefficients are useful in distinguishing healthy tissues from lesions in early stages of oesophageal cancer.

Numerical Analysis,Image and Video Processing,Medical Physics,Applications

What problem does this paper attempt to address?

The key problem that this paper attempts to solve is to develop a framework for reconstructing the amplitude, phase and polarization information of the optical vector field reflected from tissues. Specifically, the authors propose a new method to recover this information through calibration measurements, even if there are unknown linear transformations in the imaging device. The following is a specific description of this problem:

1. **Problem Background**:
- Traditional white - light endoscopes only record the amplitude information of the optical vector field, ignoring the phase and polarization information that is potentially valuable for early - cancer diagnosis.
- Phase and polarization information are very important for detecting early - stage cancer because they can indicate an increase in scattering due to microstructure changes (such as esophageal cancer).

2. **Research Objectives**:
- Develop a framework that can recover the amplitude, phase and polarization information of the optical vector field from calibration measurements.
- Research whether healthy and diseased tissues can be effectively distinguished through a specific representation system (such as the Fourier basis).
- Provide an efficient and reliable algorithm to recover this information directly from the original endoscope measurement data.

3. **Specific Problems**:
- How to achieve phase and polarization imaging in fiber - optic endoscopes, considering the complexity and frequent changes in fiber - optic transmission?
- Is there a particularly useful method for representing the optical vector field that can be used to detect optical distortions related to early - stage cancer?
- How to efficiently and stably recover this representation from the original endoscope measurement data?

4. **Solutions**:
- A general reconstruction framework is proposed, which allows the use of different regularization terms and can use different or non - orthogonal bases in image representation and device calibration.
- The effectiveness of the Fourier representation has been verified experimentally, especially the potential of its coefficients in distinguishing healthy and diseased tissues.
- Use ℓ1 regularization to directly recover the Fourier coefficients from the original measurement data, thereby reducing the reconstruction time and improving real - time performance.

5. **Application Prospects**:
- This method can be applied to the detection of early - stage cancer, especially in diseases such as esophageal cancer, by analyzing the Fourier coefficients to identify diseased tissues.
- Provide directions for future research, including further optimizing the reconstruction algorithm and exploring other possible representation systems.

Through these efforts, this paper aims to promote the development of medical imaging technology, especially to improve the accuracy and efficiency of early - cancer detection.