We developed a technique for constructing light diffusing devices comprised of a flexible shape memory polymer (SMP) cylindrical diffuser attached to the tip of an optical fiber. The devices are fabricated by casting an SMP rod over the cleaved tip of an optical fiber and media blasting the SMP rod to create a light diffusing surface. The axial and polar emission profiles and circumferential (azimuthal) uniformity are characterized for various blasting pressures, nozzle-to-sample distances, and nozzle translation speeds. The diffusers are generally strongly forward-directed and consistently withstand over 8 W of incident IR laser light without suffering damage when immersed in water. These devices are suitable for various endoluminal and interstitial biomedical applications.
Conventional embolization of cerebral aneurysms using detachable coils is time-consuming and often requires retreatment. These drawbacks have prompted the development of new methods of aneurysm occlusion. We present the fabrication and laser deployment of a shape memory (SMP) polymer expanding foam device. Data acquired in an in vitro basilar aneurysm model with and without flow showed successful treatment, with the flow rate affecting foam expansion and the temperature at the aneurysm wall.
Shape memory polymers (SMPs) are attracting a great deal of interest in the scientific community for their use in applications ranging from light weight structures in space to micro-actuators in MEMS devices. These relatively new materials can be formed into a primary shape, reformed into a stable secondary shape, and then controllably actuated to recover their primary shape. The first part of this presentation will be a brief review of the types of polymeric structures which give rise to shape memory behavior in the context of new shape memory polymers with highly regular network structures recently developed at LLNL for biomedical devices. These new urethane SMPs have improved optical and physical properties relative to commercial SMPs, including improved clarity, high actuation force, and sharper actuation transition. In the second part of the presentation we discuss the development of SMP based devices for mechanically removing neurovascular occlusions which result in ischemic stroke. These devices are delivered to the site of the occlusion in compressed form, are pushed through the occlusion, actuated (usually optically) to take on an expanded conformation, and then used to dislodge and grip the thrombus while it is withdrawn through the catheter.
In this paper the photothermal engineering issues of novel shape memory polymer (SMP) microactuators for treating stroke are presented. The engineering issues for using lasers to heat and subsequently actuate these SMP devices are presented in order to provide design criteria and guidelines for intravascular, laser activated SMP devices. The optical properties of SMP, methods for coupling laser light into SMP, heating distributions in the SMP devices and the impact of operating the thermally activated material in a blood vessel are presented. A total of three devices will be presented: two interventional ischemic stroke devices and one device for releasing embolic coils. The optical properties of SMP, methods for coupling laser light into SMP, heating distributions in the SMP devices and the impact of operating the thermally activated material in a blood vessel are presented. Actuating the devices requires device temperatures in the range of 65 degrees C - 85 degrees C. Attaining these temperatures under flow conditions requires critical engineering of the SMP optical properties, optical coupling into the SMP, and device geometries. Laser- activated SMP devices are a unique combination of laser- tissue and biomaterial technologies. Successful deployment of the microactuator requires well-engineered coupling of the light form the diffusing fiber through the blood into the SMP.
We show that standard tissue phantoms can be sued to mimic the intensity and polarization properties of tissue. Polarized light propagation through biologic tissue is typically studied using tissue phantoms consisting of dilute aqueous suspensions of microspheres. The dilute phantoms can empirically match tissue polarization and intensity properties. One discrepancy between the dilute phantoms and tissue exist: common tissue phantoms, such as dilute Intralipid and dilute 1-micrometers -diameter polystyrene microsphere suspension, depolarize linearly polarized light more quickly than circular polarized light. In dense tissue, however, where scatterers are often locate din close proximity to one another, circularly polarized light is depolarized similar to more quickly than linearly polarized light. We also demonstrate that polarized light is depolarized similar to or more quickly than linearly polarized light. We also demonstrate that polarized light propagates differently in dilute versus densely packed microsphere suspensions, which may account for the differences seen between polarized light propagation in common dilute tissue phantoms versus dense biologic tissue.
The laser-tissue interaction code LATIS is used to analyze photon scattering histories representative of optical coherence tomography (OCT) experiments performed at Lawrence Livermore National Laboratory. Monte Carlo photonics with Henyey-Greenstein anisotropic scattering is implemented and used to simulate signal discrimination of intravascular structure. An analytic model is developed and used to obtain a scaling law for optimization of the OCT signal and to validate Monte Carlo photonics. The appropriateness of the Henyey-Greenstein phase function is studied by direct comparison with more detailed Mie scattering theory using an ensemble of spherical dielectric scatterers. Modest differences are found between the two prescription for describing photon angular scattering in tissue. In particular, the Mie scattering phase functions provide less overall reflectance signal but more signal contrast compared to the Henyey-Greenstein formulation.
Imaging through biologic tissue relies on the discrimination of weakly scattered from multiply scattered photons. The degree of polarization can be used as the discrimination criterion by which to reject multiply scattered photons. Polarized light propagation through biologic tissue is typically studied using tissue phantoms consisting of dilute aqueous suspensions of microspheres. We show that, although such phantoms are designed to match the macroscopic scattering properties of tissue they do not accurately represent biologic tissue for polarization-sensitive studies. In common tissue phantoms, such as dilute Intralipid and dilute 1-micrometers -diameter polystyrene microsphere suspensions, we find that linearly polarized light is depolarized more quickly polarized light. In dense tissue, however, where scatterers are often located in close proximity to one another, circularly polarized light is depolarized similar to or more quickly than linearly polarized light. We also demonstrate that polarized light propagates differently in dilute versus densely packed microsphere suspensions, which may account for the differences seen between polarized light propagation in common dilute tissue phantoms versus dense biologic tissue.
Polarimetry, which is a comparison of the polarization state of light before and after it has interacted with a material, can be used to discriminate unscattered and weakly scattered photons from multiply scattered photons. Weakly scattered photons tend to retain their incident polarization state whereas highly scattered photons become depolarized; thus, polarization-based discrimination techniques can be used to image through tissue with decreased noise and increased contrast. Many previous studies investigating polarization- based discrimination have been conducted on tissue phantoms, with the ultimate goal being noninvasive imaging of breast tumors. We demonstrate here that linearly and circularly polarized light propagate differently in common tissue phantoms than in two independent techniques on tissue phantoms consisting of polystyrene and Intralipid microsphere suspensions, and on porcine adipose tissue and porcine myocardium. We show that contrary to expectations made from studies in the phantoms, linearly polarized light survives through more scattering events than circularly polarized light in both adipose tissue, which contains quasi-spherical scatterers, ad myocardium, which contains quasi-spherical and cylindrical scatterers. Differences between spherical and biological scatterers are discussed, along with the impact of tissue birefringence on degree of polarization measurements.
Formation of vapor bubbles is characteristic of many applications of short-pulse lasers in medicine. An understanding of the dynamics of vapor bubble generation is useful for developing and optimizing laser-based medical therapies. To this end, experiments in vapor bubble generation with laser light deposited in an aqueous dye solution near a fiber-optic tip have been performed. Numerical hydrodynamic simulations have been developed to understand and extrapolate results from these experiments. Comparison of two-dimensional simulations with the experiment shows excellent agreement in tracking the bubble evolution. Another regime of vapor bubble generation is short-pulse laser interactions with melanosomes. Strong shock generation and vapor bubble generation are common physical features of this interaction. A novel effect of discrete absorption by melanin granules within a melanosome is studied as a possible role in previously reported high Mach number shocks [Lin and Kelly, SPIE 2391, 294 (1995)].
We present a study of the short-timescale fluid dynamic response of water to a fiber-delivered laser pulse of variable energy and spatial profile. The laser pulse was deposited on a stress confinement timescale. The spatial profile was determined by the fiber core radius, r, and the water absorption coefficient, (mu) a. Considering 2D cylindrical symmetry, the combination of fiber radius and absorption coefficient parameters can be characterized as near planar, symmetric, and side-directed. The spatial profile study shows how the stress wave varies as a function of geometry. For example, relatively small absorption coefficients can result in side-propagating shear and tensile fields.
In various pulsed-laser medical applications, strong stress transients can be generated in advance of vapor bubble formation. To better understand the evolution of stress transients and subsequent formation of vapor bubbles, 2D simulations are presented in channel or cylindrical geometry with the LATIS computer code. Differences with 1D modeling are explored, and simulated experimental conditions for vapor bubble generation are presented and compared with data.
Laser injury by sub-nanosecond pulses in the eye and skin is related to strongly absorbing pigment particles such as melanin with dimension of order 10-15 nm. Single melanosomes, with size of approximately 1 micrometers and containing many such melanin particles, were isolated in water and irradiated with 100 psec pulses. Using time resolved imaging techniques, they observed the emission of a strong shock wave followed by rapid bubble expansion on a nanosecond timescale. The shock had a supersonic speed of approximately 2700 m/sec and an initial pressure of nearly 35 kbars. The shock wave can induce further tissue damage in addition to that produced by the bubble expansion and reduce the threshold for laser damage in the retina. In this work we simulate the system by using the hydrodynamic computer code LATIS with a realistic equation of state for water. We simulate both isolated melanin particles and whole melanosomes. Our melanosome model considers a spherical structure of order 1 micrometers in diameter with a uniform energy. This is consistent with the fact that the melanin particles are not stress confined while the melanosome is almost stress confined; thus, the pressure builds up uniformly in the melanosome. The details of the dynamics of the supersonic shock wave emission and rapid bubble evolution on both the melanin and melanosome scales are investigated. Comparison between modeling and experiments is presented. In order to achieve peak pressures and shock speeds comparable to the reported values, it is necessary to model the melanosome as having an absorption coefficient of approximately 6000 cm-1. Another way to achieve agreement with experiment is if the superposition of shock waves from the many melanin particles inside the melanosome produces a stronger shock than calculated by assuming a smooth absorption, as in our melanosome model. A better experimental determination of the values of linear and non- linear absorption coefficients for a single melanosome is needed in order to decide between the two approaches.
An in vivo study of vascular welding with a fiber-delivered argon laser was conducted using a canine model. Longitudinal arteriotomies and venotomies were treated on femoral vein and artery. Laser energy was delivered to the vessel wall via a 400 micrometer optical fiber. The surface temperature at the center of the laser spot was monitored in real time using a hollow glass optical fiber-based two-color infrared thermometer. The surface temperature was limited by either a room-temperature saline drip or direct feedback control of the laser using a mechanical shutter to alternately pass and block the laser. Acute patency was evaluated either visually (leak/no leak) or by in vivo burst pressure measurements. Biochemical assays were performed to investigate the possible laser-induced formation or destruction of enzymatically mediated covalent crosslinks between collagen molecules. Viable welds were created both with and without the use of feedback control. Tissues maintained at 50 degrees Celsius using feedback control had an elevated crosslink count compared to controls, while those irradiated without feedback control experienced a decrease. Differences between the volumetric heating associated with open and closed loop protocols may account for the different effects on collagen crosslinks. Covalent mechanisms may play a role in argon laser vascular fusion.
The strength and stability of laser-welded tissue may be influenced, in part, by the effects of laser exposure on collagen crosslinking. We therefore studied the effects of diode laser exposure (805 nm, 1 - 8 watts, 30 seconds) plus indocyanine green dye (ICG) on calf tail tendon collagen crosslinks. The effect of ICG dye alone on crosslink content prior to laser exposure was investigated; unexpectedly, we found that ICG-treated tissue had significantly increased DHLNL and OHP, but not HLNL. Laser exposure after ICG application reduced elevated DHLNL and OHP crosslink content down to their native levels. The monohydroxylated crosslink HLNL was inversely correlated with laser output (p less than 0.01 by linear regression analysis). DHLNL content was highly correlated with content of its maturational product, OHP, suggesting that precursor-product relationships are maintained. We conclude that: (1) ICG alone induces DHLNL and OHP crosslink formation; (2) subsequent laser exposure reduces the ICG-induced crosslinks down to native levels; (3) excessive diode laser exposure destroys normally occurring HLNL crosslinks.
The exposure of human skin to near-infrared radiation is numerically simulated using coupled laser, thermal transport and mass transport numerical models. The computer model LATIS is applied in both one-dimensional and two-dimensional geometries. Zones within the skin model are comprised of a topical solder, epidermis, dermis, and fatty tissue. Each skin zone is assigned initial optical, thermal and water density properties consistent with values listed in the literature. The optical properties of each zone (i.e. scattering, absorption and anisotropy coefficients) are modeled as a kinetic function of the temperature. Finally, the water content in each zone is computed from water diffusion where water losses are accounted for by evaporative losses at the air-solder interface. The simulation results show that the inclusion of water transport and evaporative losses in the model are necessary to match experimental observations. Dynamic temperature and damage distributions are presented for the skin simulations.
A study of laser tissue welding mediated with an indocyanine green dye-enhanced protein solder was performed. Freshly obtained sections of porcine artery were used for the experiments. Sample arterial wall thickness ranged from two to three millimeters. Incisions approximately four millimeters in length were treated using an 805 nanometer continuous- wave diode laser coupled to a one millimeter diameter fiber. Controlled parameters included the power delivered by the laser, the duration of the welding process, and the concentration of dye in the solder. A two-color infrared detection system was constructed to monitor the surface temperatures achieved at the weld site. Burst pressure measurements were made to quantify the strengths of the welds immediately following completion of the welding procedure.
Many aspects of the physical processes involved in a pulsed laser interacting with an occlusion in the intra-cranial vascular system, e.g., a blood clot, are included in the simulation codes LATIS and LATIS3D. Laser light propagation and thermo-mechanical effects on the occlusion can be calculated by these codes. The hydrodynamic response uses a realistic equation of state which includes melting and evaporation. Simple material strength and failure models now included in these codes are required to describe clot breakup. The goal is to ascertain the feasibility of laser thrombolysis, and to help optimize the laser parameters for such therapy. In this paper detailed numerical results for laser interaction with water is considered as an initial model for laser thrombolysis of soft blood clots which have high water content. Three regimes of water response to increasing laser energy are considered: (1) the linear stress pulse, (2) the nonlinear evaporation bubble, and (3) the nonlinear inertial bubble. It is shown that later in time the inertial bubble evolves into a slowly growing cavitation bubble. More physical processes will be added in the near future to better model realistic occlusion-vessel wall geometries.
We have previously demonstrated that the thermal denaturation of collagen results in a repeatable loss of linear birefringence (LB), which can be measured with birefringence compensation techniques or transmitted or remitted intensity measurements. Consequently, we have used LB to measure the kinetic changes in collagenous tissues. Analysis of the birefringence data results in the calculation of the kinetic parameters entropy, (Delta) S, enthalpy, (Delta) H and the rate constant, k. The birefringence data show that multiple kinetic processes are occurring during collagen denaturation. The most likely physical interpretation of the observed data is that structural changes are occurring on the molecular (tropocollagen) as well as the collagen fibril structural levels. Each of these structures contributes to the total birefringence. When a parallel kinetic model, which is composed of two first order exponential decay models, is applied to the birefringence data, excellent agreement between the data and model is seen. Moreover, the data also show that the relative contributions of the two constituent reactions to the total birefringence measured are in excellent agreement with reported values of intrinsic and form birefringence contributions in collagen. The knowledge that multiple reactions exist during collagen denaturation impacts the understanding of how denaturation occurs and what information can be utilized for the control of the heating process.
Linear birefringence (LB) is a polarization-specific property of many semi-crystalline structures in tissue. Specifically, collagen, with its triple helix conformation, exhibits LB in its native state. Rat tail tendon (RTT) was chosen for the LB experiments because it is > 90% collagen and the collagen fiber alignment is nearly parallel with the RTT length. This alignment results in RTT exhibiting uniaxial characteristics such that two properly chosen optical axes display differing refractive indices ((Delta) n equals nslow - nfast). RTT, which has an elliptical cross section, has its slow axis parallel to the tendon's length and a fast axes along the tendon's cross section. Native RTT has a refractive index difference of (Delta) n equals 1.5 X 10-3. For a typical tendon thickness of 200 micrometers , the phase shift, (delta) equals n*d (d, diameter), is approximately equal to 300 nm (transmission measurement). Heating of RTT results in a repeatable loss of (delta) . If monochromatic light is used the sample's output intensity is proportional to sin2((delta) (pi) /(lambda) ) where (lambda) is the wavelength of the light. Thus, given the native phase shift, the incident light's wavelength may be chosen such that the sample's loss of LB with heating is intensity- mapped on the sample's image.