<p>The visible wavelength range has proven to be a useful spectral window for observing biophotonic events such as absorption in materials (oxy-hemoglobin and deoxy-hemoglobin), light scattering in biological tissue, and biochemical and fluorescence reactions. Diffuse reflectance spectroscopy (DRS) is a technique that utilizes the diffuse reflectance spectra from turbid media (e.g. biological tissue) to quantify the optical properties (e.g. absorption and scattering) of those media. DRS in the visible wavelength range can be utilized to optically differentiate between healthy and cancerous tissue, and thus has applications in intra-operative tumor margin assessment. </p><p>The footprint of conventional DRS systems used for intra-operative tissue margin assessment prohibits their widespread use inside the surgical suite, where space is at a premium. Conventional quantitative DRS imaging systems utilize unwieldy fiber probes, cooled CCD cameras, and imaging spectrographs for imaging tissue margins. These system components not only increase system size, limiting their use inside the surgical suite, but also limit imaging resolution, imaging speed, and increase overall system cost. </p><p>Silicon is an attractive candidate for the development of compact, customized photodetector elements for biophotonic imaging applications such as intra-operative tumor margin assessment using DRS. This thesis deals with the design and development of a customized DRS imaging probe composed of custom silicon imaging arrays for intra-operative breast tumor margin assessment. The first generation of the customized imaging probe consisted of a 4x4 array of annular epitaxial Si pn junction photodiodes (PDs) with a measured responsivity of 0.28 A/W - 0.37 A/W for λ= 470 nm - 600 nm, and a measured dark current density of 1.456 nA/cm2 - 4.48 nA/cm2. The imaging array was used to detect diffuse reflectance when placed in direct contact with tissue. A quartz light delivery tube coupled to a xenon lamp was optimized to deliver light to the tissue through the holes of the annular imaging array across a 256 mm2 imaging area. The pixel-to-pixel spacing in the imaging array was 4.5 mm, the highest resolution reported to date for a multi-pixel DRS probe. This resolution was limited by pixel-to-pixel optical crosstalk, which was theoretically calculated and experimentally characterized, to validate the theoretical model for future designs. This first generation probe was successfully tested on diffuse reflectance standards, tissue-mimicking phantoms, animal tissue, and human breast tissue, and yielded an SNR of 30 dB - 55 dB on all measured specimens. </p><p>The next generation of the customized imaging probe consisted of a 4x4 array of annular thin-film Si pn junction PDs heterogeneously bonded to a transparent Pyrex substrate, to enable integration with a guided wave light delivery system. The 4x4 thin-film PD array design and development was prototyped using a 1x2 thin-film PD array heterogeneously bonded to a Pyrex substrate. The responsivity and dark current of the thin-film PDs in the 1x2 array were measured to be 0.19 A/W - 0.34 A/W for λ= 470 nm - 600 nm and 0.63 nA/cm2, respectively. The process for the 1x2 thin-film PD array was scaled to fabricate a 4x4 array of thin-film PDs for DRS, and the 4x4 array was optically and electrically characterized. These heterogeneously bonded thin-film single crystal Si PDs have the highest uncooled responsivity to dark current density ratio (greater than 0.30 - 0.54 cm2/nW for λ= 470 nm - 600 nm) reported to date, to the best of our knowledge. The 1x2 array of thin-film PDs were also heterogeneously bonded to a flexible substrate without any degradation in PD optical and electrical characteristics, opening the door towards conformal tissue imaging.</p> / Dissertation
| Identifer | oai:union.ndltd.org:DUKE/oai:dukespace.lib.duke.edu:10161/6153 |
| Date | January 2012 |
| Creators | Dhar, Sulochana |
| Contributors | Jokerst, Nan M |
| Source Sets | Duke University |
| Detected Language | English |
| Type | Dissertation |
Page generated in 0.0137 seconds