Amanda Dayton

Light-guided lumpectomy: device and case report

We describe the development, design, fabrication, and testing of an optical wire to assist in the surgical removal of small lesions during breast-conserving surgery. We modify a standard localization wire by adding a 200-μm optical fiber alongside it; the resulting optical wire fit through an 18 gauge needle for insertion in the breast. The optical wire is anchored in the lesion by a radiologist under ultrasonic and mammographic guidance. At surgery, the tip is illuminated with an eye-safe, red, HeNe laser, and the resulting glowball of light in the breast tissue surrounds the lesion. The surgeon readily visualizes the glowball in the operating room. This glowball provides sufficient feedback to the surgeon that it is used (1) to find the lesion and (2) as a guide during resection. Light-guided lumpectomy is a simple enhancement to traditional wire localization that could improve the current standard of care for surgical treatment of small, nonpalpable breast lesions.

Differential interference contrast microscopy for the quantitative assessment of tissue organization

The propagation of light through complex structures, such as biological tissue, is a poorly understood phenomenon. Typically the tissue is modeled as an effective medium, and Monte Carlo techniques are used to solve the radiative transport equation. In such an approach the medium is characterized in terms of a limited number of physical scatter and absorption parameters, but is otherwise considered homogeneous. For exploration of propagation phenomena such as spatial coherence, however, a physical model of the tissue medium that allows multiscale structure is required. We present a particularly simple means of establishing such a multiscale tissue characterization based on measurements using a differential interference contrast (DIC) microscope. This characterization is in terms of spatially resolved maps of the (polar and azimuthal) angular ray deviations. With such data, tissues can be characterized in terms of their first and second order scatter properties. We discuss a simple means of calibrating a DIC microscope, the measurement procedure and quantitative interpretation of the ensuing data. These characterizations are in terms of the scatter phase function and the spatial power spectral density

Polyurethane Phantoms with Homogeneous and Nearly Homogeneous Optical Properties

Phantoms with controlled optical properties are often used for calibration and standardization. The phantoms are typically prepared by adding absorbers and scatterers to a clear host material. It is usually assumed that the scatterers and absorbers are uniformly dispersed within the medium. To explore the effects of this assumption, we prepared paired sets of polyurethane phantoms (both with identical masses of absorber, India ink and scatterer, titanium dioxide). Polyurethane phantoms were made by mixing two polyurethane parts (a and b) together and letting them cure in a polypropylene container. The mixture was degassed before curing to ensure a sample without bubbles. The optical properties were controlled by mixing titanium dioxide or India ink into polyurethane part (a or b) before blending the parts together. By changing the mixing sequence, we could change the aggregation of the scattering and absorbing particles. Each set had one sample with homogeneously dispersed scatterers and absorbers, and a second sample with slightly aggregated scatterers or absorbers. We found that the measured transmittance could easily vary by a factor of twenty. The estimated optical properties (using the inverse adding-doubling method) indicate that when aggregation is present, the optical properties are no longer proportional to the concentrations of absorbers or scatterers.

Light-guided lumpectomy: first clinical experience

Despite numerous advances, lumpectomy remains a challenging procedure. We report on the early use of light-guided lumpectomy. Eight patients with non-palpable breast cancer undergoing lumpectomy for biopsy-proven and radiographically identifiable cancer were enrolled in the study. An optical wire was designed that incorporated a standard hook-wire with an optical fiber. The optical wire was placed in the same manner as a standard hook-wire. During light-guided lumpectomy, an eye-safe laser illuminated the optical wire and created a sphere of light surrounding the cancer. The light was visible at the beginning of each surgery and facilitated approaching the cancer without using the wire. Dissection around the sphere of light kept the wire tip within the surgical specimen. Three of eight initial surgical specimens had focally positive margins. Additional cavity shaves were performed during five lumpectomies and resulted in negative margins in seven of eight patients. Light-guided lumpectomy is a minor change to breast conserving surgery that can be easily incorporated into clinical practice. Further investigation into the clinical benefit of light-guided lumpectomy is warranted.

Experimental validation of phase using Nomarski microscopy with an extended Fried algorithm

Reconstruction of an image (or shape or wavefront) from measurements of the derivatives of the image in two orthogonal directions is a common problem. We demonstrate how a particular reconstructor, commonly referred to as the Fried algorithm, can be used with megapixel derivative images to recover the original image. Large datasets are handled by breaking the derivative images into smaller tiles, applying the Fried algorithm and stitching the tiles back together. The performance of the algorithm is demonstrated using differential interference contrast microscopy on a known test object.

Optical wire guided lumpectomy: frequency domain measurements

In practice, complete removal of the tumor during a lumpectomy is difficult; the published rates of positive margins range from 10% to 50%. A spherical lumpectomy specimen with tumor directly in the middle may improve the success rate. A light source placed within the tumor may accomplish this goal by creating a sphere surrounding the tumor that can serve as a guide for resection. In an optical phantom and a prophylactic mastectomy specimen, sinusoidally modulated light within the medium was collected by optical fiber(s) at fixed distance(s) from the source and used to measure the optical properties. These optical properties were then used to calculate the distance the light had traveled through the medium. The fiber was coupled to an 830nm diode laser that was modulated at 100, 200 and 300 MHz. A handheld optical probe collected the modulated light and a network analyzer measured the phase lag. This data was used to calculate the distance the light traveled from the emitting fiber tip to the probe. The optical properties were μa = 0.004mm-1 and μ1s = 0.38mm-1 in the phantom. The optical properties for the tissue were μa = 0.005mm-1 and μ1s = 0.20mm-1. The prediction of distance from the source was within 4mm of the actual distance at 30mm in the phantom and within 3mm of the actual distance at 25mm in the tissue. The feasibility of a frequency domain system that makes measurements of local optical properties and then extrapolates those optical properties to make measurements of distance with a separate probe was demonstrated.

Extracting Absolute Phase and Amplitude from DIC Imagery

We discuss the use of a DIC microscope for characterization of the scatter and absorption properties of thin tissue samples. We demonstrate the calibration process, illustrate phase-stepping approaches, and show representative results.

Measuring distance through turbid media: a simple frequency domain approach

In both industry and medicine there is no optical technique to measure distance through light scattering media. Such a technique may be useful for localizing embedded structures, or may be a non-contact method of measuring turbid media. The limits of a frequency domain based technique were explored in three polyurethane optical phantoms. We have demonstrated a simple method to measure the distance between an intensity modulated light source and detector in turbid media based on the proportionality of the phase lag to the distance. The limits of the technique were evident for distances less than 5 mm, particularly when μ1s <0.1mm-1 and distances greater than 55mm for the phantoms studied. This method may prove useful in industry and medicine as a non destructive way measure distance through light scattering media.

Light guided lumpectomy: is continuous wave or frequency domain more accurate

Improving the success of lumpectomies would reduce the number of procedures, cost, and morbidity. A light source could be placed in a lesion to assist in finding and removing the lesion. A quantitive measurement of the distance between such a light source and a detector would further aid in the procedure by providing surgeons with easy to use intra-operative guidance to the lesion. Two methods, continuous wave and frequency domain, of accomplishing this measurement were compared. Within one radio frequency experimental system, the amplitude at 15MHz was taken to represent the continuous wave signal and the phase at 100MHz was taken to represent the frequency domain signal. For the continuous wave method, data at source-detector separation distances of 20, 30 & 50mm were used to predict other distances of 10, 20, 30, 40, & 50 mm. Data at source-detector separation distances of 20 & 40mm was used to predict distances for the frequency domain method. When the difference between the predicted distance and the actual distance was compared to zero the continuous wave method was significantly different (students t-test, p = 0.03) while the frequency domain method was not statistically different from zero (studentst-test, p > 0.05). The frequency domain method was more accurate at predicting the source-detector separation distance between 10 & 50 mm. This frequency domain method of measuring distance may be useful in locating and removing lesions during lumpectomy procedures.