חניאל גבאי

תלמיד המחלקה לתואר שני ירצה בנושא:

Simplified Noised-Reduced Optical Interferometric Imaging
for Biomedical and Biometrical Applications

In interferometric imaging, both the quantitative phase and the amplitude maps of the light interacted with a sample can be numerically reconstructed from the interference pattern recorded by a digital camera. From the reconstructed phase data, it is possible to produce high sensitivity thickness profile of the sample without scanning or external labeling. These capabilities make interferometry a highly attractive tool for both macroscopic and microscopic imaging applications. However, this sensitivity also makes these imaging systems very vulnerable to environmental noises including mechanical vibrations and spatial imaging noises. This vulnerability increases when trying to image deep samples with complex surface topography (i.e. topographies containing rapidly changing thickness values). Moreover, in areas where the sample is deeper than the illuminating wavelength, the final images will contain 2π phase unwrapping ambiguities. These two issues make the imaging of deep and complex topographies a challenging task.

To address these problems, we have designed and constructed a novel interferometric imaging system, which is capable of handling complex and deep samples by illuminating the sample with two wavelengths to elegantly solve 2π phase unwrapping ambiguities. Since this technique is highly sensitive to optical noise, which may finally result in corruption of the reconstructed data, we obtained low-noise level by combining several engineering approaches. First, a single element (beamsplitter), common-path interferometer was designed and constructed for the proposed imaging system. This design significantly reduces environmental noise factors and wavefront distortions, while minimizing the need for additional optical elements that may add temporal noise due to system stability issues. In addition, a simultaneous dual-channel interferometric imaging of the sample is obtained. A simple subtraction between the two images further reduces the spatial noise due to image averaging in the digital reconstruction process performed in the Fourier domain. Moreover, we have used a tunable low-coherence light source, which lowers coherent noise and enables us to implement the two-wavelengths phase unwrapping method in an easy way. In contrast to traditional interferometric systems, our setup is easy to align and use and requires no special optical expertise. Moreover, no time-consuming beam path matching is required in order to achieve interference with the low-coherence light source.

To demonstrate the usefulness of our system, we imaged fingerprint patterns in three dimensions, with nanometer scale thickness accuracy. This type of imaging has great advantages for biometrical uses in protecting against identity theft and spoofing, as well as for biomedical applications serving as a potential medical diagnostic tool. For example, fingerprints may serve as a simple, non-invasive biomarker for Alzheimer's and Celiac diseases due to significant changes in the fingerprint morphology during the disease.

We will also present microscopic application of interferometric imaging for cancer diagnosis and monitoring by measuring the optical thickness fluctuations of the cells over time. We found that cancer cells fluctuate significantly more than healthy cells taken from the same organ of the same individual, and that metastatic cancer cells fluctuate significantly more than primary cancer cells, a fact that that supports the hypothesis that metastatic cells are more elastic and can easily squeeze through narrow body capillaries, while spreading themselves through the body.

העבודה נעשתה בהנחיית ד"ר נתן שקד, המחלקה להנדסה ביו-רפואית, אוניברסיטת תל-אביב

ההרצאה תתקיים ביום ראשון 27.1.2013, בשעה 14:45,

בחדר 315, הבניין הרב תחומי, אוניברסיטת תל אביב