Institutionen för medicinsk teknik (IMT)

A spectroscopic probe with multiple detecting fibers was used for quantifying absorption and scattering in liquid optical phantoms. The phantoms were mixtures of Intralipid and red and blue food dyes. Intensity calibration for the detecting fibers was undertaken using either a microsphere suspension (absolute calibration) or a uniform detector illumination (relative calibration between detectors). Two different scattering phase functions were used in an inverse Monte Carlo algorithm. Data were evaluated for residual spectra (systematic deviations and magnitude) and accuracy in estimation of scattering and absorption. Spectral fitting was improved by allowing for a 10% intensity relaxation in the optimization algorithm. For a multi-detector setup, non-systematic residual spectrum was only found using the more complex Gegenbauer-kernel phase function. However, the choice of phase function did not influence the accuracy in the estimation of absorption and scattering. Similar estimation accuracy as in the multi-detector setup was also obtained using either two relative calibrated detectors or one absolute calibrated detector at a fiber separation of 0.46 mm.

OBJECTIVE-To compare the microcirculatory velocity distribution in type 2 diabetic patients and nondiabetic control subjects at baseline and after local heating. RESEARCH DESIGN AND METHODS-The skin blood flow response to local heating (44 degrees C for 20 mm) was assessed in 28 diabetic patients and 29 control subjects using a new velocity-resolved quantitative laser Doppler flowmetry technique (qLDF). The qLDF estimates erythrocyte (RBC) perfusion (velocity X concentration), in a physiologically relevant unit (grams RBC per 100 g tissue X millimeters per second) in a fixed output volume, separated into three velocity regions: v less than1 mm/s, v 1-10 mm/s, and v greater than10 mm/s. RESULTS-The increased blood flow occurs in vessels with a velocity greater than1 mm/s. A significantly lower response in qLDF total perfusion was found in diabetic patients than in control subjects after heat provocation because of less high-velocity blood flow (v greater than10 mm/s). The RBC concentration in diabetic patients increased sevenfold for v between 1 and 10 mm/s, and 15-fold for v greater than10 mm/s, whereas no significant increase was found for v less than1 mm/s. The mean velocity increased from 0.94 to 7.3 mm/s in diabetic patients and from 0.83 to 9.7 mm/s in control subjects. CONCLUSIONS-The perfusion increase occurs in larger shunting vessels and not as an increase in capillary flow. Baseline diabetic patient data indicated a redistribution of flow to higher velocity regions, associated with longer duration of diabetes. A lower perfusion was associated with a higher BMI and a lower toe-to-brachial systolic blood pressure ratio.

A new method for estimating the measurement depth and volume in laser Doppler flowmetry (LDF) is presented. The method is based on Monte Carlo simulations of light propagation in tissue. The contribution from each individual Doppler shift is calculated and thereby multiple Doppler shifts are handled correctly. Different LDF setups for both probe based (0.0, 0.25, 0.5, and 1.2 mm source-detector separation) and imaging systems (0.5 and 2.0 mm beam diameter) are considered, at the wavelengths 543 nm, 633 nm, and 780 nm. Non-linear speckle pattern effects are accounted for in the imaging system setups. The effects of tissue optical properties, blood concentration, and blood oxygen saturation are evaluated using both homogeneous tissue models and a layered skin model. The results show that the effect on the measurement depth of changing tissue properties is comparable to the effect of changing the system setup, e.g. source-detector separation and wavelength. Skin pigmentation was found to have a negligible effect on the measurement depth. Examples of measurement depths are (values are given for a probe based system with 0.25 mm source-detector separation and an imaging system with a 0.5 mm beam diameter, respectively, both operating at 780 nm): muscle - 0.55/0.79 mm; liver - 0.40/0.53 mm; gray matter - 0.48/0.68 mm; white matter - 0.20/0.20 mm; index finger pulp - 0.41/0.53 mm; forearm skin - 0.53/0.56 mm; heat provoked forearm skin - 0.66/0.67 mm.

Two forced detection (FD) variance reduction Monte Carlo algorithms for image simulations of tissue-embedded objects with matched refractive index are presented. The principle of the algorithms is to force a fraction of the photon weight to the detector at each and every scattering event. The fractional weight is given by the probability for the photon to reach the detector without further interactions. Two imaging setups are applied to a tissue model including blood vessels, where the ID algorithms produce identical results as traditional brute force simulations, while being accelerated with two orders of magnitude. Extending the methods to include refraction mismatches is discussed.

The principle of forced detection; a part of the photon weight. based on the probability of reaching the detector without further interactions, is forced to the detector at each and every scattering event.

An optical microvascular skin model, valid at 780 nm, was developed. The model consisted of six layers with individual optical properties, and variable thicknesses and blood concentrations at three different blood flow velocities. Monte Carlo simulations were used to evaluate the impact of various model parameters on the traditional Laser Doppler flowmetry (LDF) measures. A set of reference Doppler power spectra was generated by simulating 7,000 configurations, varying the thickness and blood concentrations. Simulated spectra, at two different source detector separations, were compared with in vivo recorded spectra, using a non-linear search algorithm for minimizing the deviation between simulated and measured spectra. The model was validated by inspecting the thickness and blood concentrations which generated the best fit. These four parameters followed a priori expectations for the measurement situations, and the simulated spectra agreed well with the measured spectra for both detector separations. Average estimated dermal blood concentration was 0.08% at rest and 0.63% during heat provocation (44°C) on the volar side of the forearm, and 1.2% at rest on the finger pulp. The model is crucial for developing a technique for velocity-resolved absolute LDF measurements with known sampling volume, and can also be useful for other bio-optical modalities.

A method to separate a Doppler power spectrum into a number of flow velocity components, measured in absolute units (mm/s), is presented. A Monte Carlo software was developed to track each individual Doppler shift, to determine the probability, p(n), for a photon to undergo n Doppler shifts. Given this shift distribution, a mathematical relationship was developed and used to calculate a Doppler power spectrum originating from a certain combination of velocity components. The non linear Levenberg-Marquardt optimization method could thus be used to fit the calculated and measured Doppler power spectra, giving the true set of velocity components in the measured sample. The method was evaluated using a multi tube flow phantom perfused with either polystyrene microspheres or undiluted/diluted human blood (hct = 0.45). It estimated the velocity components in the flow phantom well, during both low and high concentrations of moving scatterers (microspheres or blood). Thus, further development of the method could prove to be a valuable clinical tool to differentiate capillary blood flow.

 Light absorption in tissue is generally decreased when chromophores are spatially concentrated rather than being homogeneously distributed. In tissue, this applies to hemoglobin located in blood vessels (vessel packaging). In this paper, the diffusely reflected light from 41 tissue models with discrete blood vessels with diameters ranging from 6.25 to 100 μm were simulated using the Monte Carlo technique. A reverse engineering approach was then utilized to find the model that had an optimal spectral fit to each of the simulated models. The average vessel diameter was one fitting parameter in the adaptive model. The estimated vessel diameter from the optimal fit model was compared to the known diameter from the simulated models. Two different methods to calculate the vessel packaging effect were used, one existing based on a simple analytic expression and a new method based on path length distributions. Both methods had similar performance. For the new method, the absolute RMS deviation of the estimated vessel diameter was 5.5 μm for vessel diameters ≤ 25 μm, and the relative RMS deviation was 21 % for vessel diameters > 25 μm.

Laser Doppler flowmetry (LDF) is virtually the only non-invasive technique, except for other laser speckle based techniques, that enables estimation of the microcirculatory blood flow. The technique was introduced into the field of biomedical engineering in the 1970s, and a rapid evolvement followed during the 1980s with fiber based systems and improved signal analysis. The first imaging systems were presented in the beginning of the 1990s.

Conventional LDF, although unique in many aspects and elegant as a method, is accompanied by a number of limitations that may have reduced the clinical impact of the technique. The analysis model published by Bonner and Nossal in 1981, which is the basis for conventional LDF, is limited to measurements given in arbitrary and relative units, unknown and non-constant measurement volume, non-linearities at increased blood tissue fractions, and a relative average velocity estimate.

In this thesis a new LDF analysis method, quantitative LDF, is presented. The method is based on recent models for light-tissue interaction, comprising the current knowledge of tissue structure and optical properties, making it fundamentally different from the Bonner and Nossal model. Furthermore and most importantly, the method eliminates or highly reduces the limitations mentioned above.

Central to quantitative LDF is Monte Carlo (MC) simulations of light transport in tissue models, including multiple Doppler shifts by red blood cells (RBC). MC was used in the first proof-of-concept study where the principles of the quantitative LDF were tested using plastic flow phantoms. An optically and physiologically relevant skin model suitable for MC was then developed. MC simulations of that model as well as of homogeneous tissue relevant models were used to evaluate the measurement depth and volume of conventional LDF systems. Moreover, a variance reduction technique enabling the reduction of simulation times in orders of magnitudes for imaging based MC setups was presented.

The principle of the quantitative LDF method is to solve the reverse engineering problem of matching measured and calculated Doppler power spectra at two different source-detector separations. The forward problem of calculating the Doppler power spectra from a model is solved by mixing optical Doppler spectra, based on the scattering phase functions and the velocity distribution of the RBC, from various layers in the model and for various amounts of Doppler shifts. The Doppler shift distribution is calculated based on the scattering coefficient of the RBC:s and the path length distribution of the photons in the model, where the latter is given from a few basal MC simulations.

When a proper spectral matching is found, via iterative model parameters updates, the absolute measurement data are given directly from the model. The concentration is given in g RBC/100 g tissue, velocities in mm/s, and perfusion in g RBC/100 g tissue × mm/s. The RBC perfusion is separated into three velocity regions, below 1 mm/s, between 1 and 10 mm/s, and above 10 mm/s. Furthermore, the measures are given for a constant output volume of a 3 mm³ half sphere, i.e. within 1.13 mm from the light emitting fiber of the measurement probe.

The quantitative LDF method was used in a study on microcirculatory changes in type 2 diabetes. It was concluded that the perfusion response to a local increase in skin temperature, a response that is reduced in diabetes, is a process involving only intermediate and high flow velocities and thus relatively large vessels in the microcirculation. The increased flow in higher velocities was expected, but could not previously be demonstrated with conventional LDF. The lack of increase in low velocity flow indicates a normal metabolic demand during heating. Furthermore, a correlation between the perfusion at low and intermediate flow velocities and diabetes duration was found. Interestingly, these correlations were opposites (negative for the low velocity region and positive for the mediate velocity region). This finding is well in line with the increased shunt flow and reduced nutritive capillary flow that has previously been observed in diabetes.