Accurate quantification of near-infrared (NIRS) signal changes during monitoring of the adult head is required for regular clinical use. By incorporating NIRS into a multimodal monitoring system, which documents cerebral haemodynamic changes during carotid endarterectomy, we have provided the opportunity for calibration against parameters known to warn of severe cerebral ischaemia (SCI) under controlled anaesthesia. Application of a particular protocol in which clamping of the External Carotid Artery was performed two minutes before the Internal Carotid Artery enabled isolation of the NIRS measurements from the intracranial tissue. This way the most profound problem in adult NIRS, extracranial contamination, has been successfully addressed. In consequence, NIRS derived thresholds for intraoperative cerebral ischaemia showed in a series of 76 patients high agreement with criteria defined using transcranial Doppler and the cerebral function monitor. In contrast, when non-corrected for extracranial components NIRS measurements were used no thresholds for SCI were apparent.

We used near infrared spectroscopy (NIRS) to study non- invasively changes in cerebral hemoglobin oxygenation during performance of a verbal fluency task in healthy volunteers (n equals 19). We used a multichannel system combining six commercially available NIRS devices (NIRO-500, Hamamatsu Photonics) that allows measurements over the left and right frontal, parietal and occipital cortex simultaneously. We described the NIRS variables both in the time- and frequency-domain using digital signal processing. In the time-domain we measured the blood flow response by calculation of concentration changes for oxygenated and deoxygenated hemoglobin (HbO₂, HbR) and their latency to stimulus variations. We applied frequency selective filters and found an increase of the global signal variability during cognitive stimulation. These effects can be visualized by spectral analysis. The powerspectras of the optical signal varied clearly with rest and activity states and showed a different spectral distribution over the frontal as compared to the parietal cortex.

In this study we assess changes in the hemoglobin oxygenation (oxy-Hb, deoxy-Hb) and the Cytochrome-C-Oxidase redox state (Cyt-ox) in the occipital cortex during visual stimulation by near infrared spectroscopy. For the calculation of changes in oxy-Hb, deoxy-Hb and Cyt-ox from attenuation data via a modified Beer-Lambert equation, the wavelength dependence of the differential pathlength factor (DPF), i.e. the ratio of the mean optical pathlength and the physical light-source-detector separation, has to be taken into account. The wavelength dependence of the DPF determines the crosstalk between the different concentrations and is therefore essential for a high sensitivity. Here a simple method is suggested to estimate the wavelength dependence of the DPF((lambda) ) from pulse induced attenuation changes measured on the head of adult humans. The essence is that the DPF is the ratio of the attenuation changes over absorption coefficient changes and the spectral form of the pulse correlated absorption coefficient change is proportional to the extinction coefficient of blood. Indicators for the validity of the DPF((lambda) ) derived for wavelengths between 700 and 970 nm are the stability of the calculated oxy-Hb, deoxy-Hb and Cyt-ox signals with variations of the wavelength range included for their calculation and its overall agreement with the data available from the literature. The DPF derived from pulse measurements was used for the analysis of attenuation data from cortical stimulations. We show that Cyt-ox in the occipital cortex of human subjects is transiently oxidized during visual stimulation.

Appropriate modeling of light propagation in the adult head is important to deduce the partial optical pathlength for quantitative measurement of brain oxygenation using near infrared spectroscopy. It is known that the clear CSF layer significantly affects the light propagation in the brain and it is likely that the optical properties of the skull adjoining the CSF will also affect the light propagation, especially in the case of a layered (i.e. cortical/trabecular/cortical) bone. In this study a Monte Carlo simulation has been used to investigate the influence of the optical properties of the skull on the partial optical pathlength of light in the adult head. Results show that where the absorption coefficient of the trabecular bone is higher than that of outer cortical bone, the partial optical pathlength in the brain decreases and the detected light has not propagated through the trabecular bone. In this case the outer border of the intensity PMDF tends to be confined to the inner skull boundary with the CSF layer.

The effects of hypercapnia on the intensity and phase of near infrared light back-scattered by the head were studied on eight healthy humans and two patients affected by monohemispheric lesions in the Middle Cerebral Artery (MCA) territory. A decrease in the light intensity and a variation of the phase were detected in all healthy subjects during apnea and hypercapnia. Only negligible changes were observed in the affected hemisphere of the patients. A concomitant study by Transcranial Doppler Sonography (TCD) showed an increase of the blood flow during hypercapnia both in normal hemispheres and, to a less extent, in the affected hemisphere of patients. This suggests that NIRS (Near Infrared Spectroscopy) is more sensitive to alterations of more cortical brain vascular system than TCD which is mainly testing MCA in the depth.

Intracerebral Hemorrhage (IH) is an important cause of secondary brain injury in neurosurgical patients. Early identification and treatment improve neurologic outcome. We have tested Near Infrared Spectroscopy (NIRS) as an alternative noninvasive diagnostic tool compared to CT-Scans to detect IH. We prospectively studied 212 patients with neurologic symptoms associated with intracranial pathology before performing a CT-scan. NIRS signals indicated pathologies in 181 cases (sensitivity 0.96; specificity 0.29). In a subgroup of subdural hematomas NIRS detected 45 of 46 hematomas (sensitivity 0.96; specificity 0.79). Identification of intracerebral hemorrhage using NIRS has the potential to allow early treatment, thus possibly avoiding further injury.

Near infrared spectroscopy (NIRS) determined changes in the cerebral blood volume (CBV) were compared to those obtained by positron emission tomography (PET) in five healthy volunteers (2 females). Two NIRS optodes were placed on the left forehead and NIRS-CBV was derived from the sum of oxyhemoglobin and deoxyhemoglobin. CBV changes were induced by hyperventilation and inhalation of 6% CO₂. After 2 min inhalation of labeled carbon monoxide, data were sampled during 8 min for both PET- and NIRS-CBV as well as for the arterial carbon dioxide tension (P_aCO₂). The region of interest for PET-CBV was `banana-shaped' with boundaries corresponding to the position of the NIRS optodes on the transmission scan and to a depth of approximately 2 cm. During hyperventilation, P_aCO₂ decreased from 5.2 (4.6 - 5.8) to 4.6 (4.2 - 4.9) kPa and equally PET-CBV (from 3.9 (2.5 - 5.2) to 3.6 (3.0 - 4.8) ml (DOT) 100 g^-1) and NIRS-CBV were reduced (by -0.14 [-0.38 - 0.50] ml (DOT) 100 g^-1). During hypercapnia P_aCO₂ increased to 6.0 (5.9 - 7.0) kPa accompanied by parallel changes in PET- (to 4.5 (3.9 - 4.9) ml (DOT) 100 g^-1) and NIRS-CBV (by 0.04 [-0.02 - 0.30] ml (DOT) 100 g^-1) and the two variables were correlated (r equals 0.78, p < 0.05). In conclusion, with a moderate change in the arterial carbon dioxide tension, the cerebral blood volumes determined by near infrared spectroscopy and by positron emission tomography change in parallel but the change in NIRS-CBV is small compared to that obtained by PET.

For a long time continuous wave near infrared instruments have been used to detect oxygenation changes in tissue. These instruments have proven to be reliable. The new generation of instruments, such as phase-modulated systems, or time-of-flight instruments, is not yet reliable enough for clinical applications. Most available continuous wave near infrared instruments have low temporal resolution and low signal-to-noise ratio. For functional brain imaging, for example, a sensitive and fast instrument is needed. Therefore we developed the OXYMON, an instrument with a sample frequency up to 50 Hz and for optical densities up to 9 OD. The instrument uses 3 laser diodes, and is equipped with 1 or 2 avalanche photo detectors. Modular building techniques make maintenance easy.

Use of pure diffusion theory produced errors in the measurement of absorbance and scattering using a time-of- flight system. In an attempt to produce quantitative values for absorbance and scattering, we performed both software simulations and empirical collections. For the empiric correction, we first developed a calibration system based upon a homogeneous liquid test phantom. Next, we performed multiwavelength (785, 808, 850, 904 nm) time-resolved spectroscopy upon known standards using this model, and then developed an algorithm to correct for the differences between our system and the known values. We suspect that part of the difference between theory and actual measured values, in the absence of correction, may be due to the response characteristics of our diode-based laser and detection system. If so, then our calibration represents a deconvolution of the response characteristics of the system, and suggests that `less clean' components can be used, reducing cost of time-resolved systems.

Conventional near infrared spectrophotometric monitors have temporal resolution of less than about 1 Hz. However, physiological Hb signals such as pulsation and muscle contraction have higher frequency than 1 Hz. Insufficient sampling rates inevitably lead aliasing of the recorded signals in tissue oximetry for both brain and muscle. Cerebral Hb signals (57 y.o. female artificially ventilated under general anesthesia) and thigh muscle (22 y.o. male with 20 W - 240 W exercise at 1 Hz cycling in semirecumbent ergometer) were measured with NIRS monitor with temporal resolution of 10 Hz (OMRON Co. Ltd., Japan). The detail of physiological fluctuations such as pulsation, ventilation, and muscle pumping was clearly recognized with a 10 Hz sampling. The comparison with recalculated waveforms at slower sampling rate (0.5 Hz, 1 Hz, 2 Hz) revealed that with slower sampling than 1 Hz cerebral respiratory waves were deformed by pulsation, and that magnitudes of muscle pumping could not be properly evaluated in dynamic exercise. In both pulsatile and muscle contractile cycle a phase delay between oxygenated component and deoxygenated one was also detected, which has been overlooked by conventional NIRS monitoring.

Using continuous near infrared spectroscopy, we studied postnatal changes in both cerebral hemoglobin oxygen saturation (SbO2) and cerebral total hemoglobin content (THb) in normal and asphyxiated neonates. With normal neonates, THb showed a tendency to decrease within 12 hours after birth. It then remained constant until day 5 but was found to be at a lower level at 1 month after birth. The SbO2 levels remained constant for the initial 5 days but then decreased to lower levels at 1 month. THb and SbO2 levels were higher in moderately and severely asphyxiated neonates than in normal neonates. In one severely asphyxiated neonate, SbO2 was still higher even at 1 month. In another severely asphyxiated neonate, both THb and SbO2 fell after intracranial hemorrhage occurred.

The present study demonstrated that optical variables HbT and SmcO₂ can be used to monitor changes in cerebral hemodynamics and oxygenation during asphyxia. Unfortunately none of the individual optical variables alone could be used to monitor changes in cerebral hemodynamics and oxygenation under a variety of possible clinical circumstances. However, all variables together, forming patterns unique to the commonly occurring physiological conditions, might potentially serve as a `silver standard' to aid interpretations of optical signals in clinical settings where `gold standard' techniques are not available, i.g. in the human fetus and neonate.

We have used near-infrared spectroscopy to measure the hemoglobin saturation at rest and during exercise on patients affected by peripheral vascular disease (PVD). The instrument used in our study is a frequency-domain tissue oximeter which employs intensity modulated (110 MHz) laser diodes. We examined 9 subjects, 3 of which were controls and 6 were patients affected by stage II PVD. The optical probe was located on the calf muscle of the subjects. The measurement protocol consisted of: (1) baseline (approximately 5 min); (2) stationary bicycle exercise (approximately 5 min); (3) recovery (approximately 15 min). The change in hemoglobin saturation during exercise ((Delta) Y) and the recovery time after exercise (t_rec) were significantly greater in the PVD patients ((Delta) Y equals -21 +/- 3%, t_rec equals 5.9 +/- 3.8 min) than in the control subjects ((Delta) Y equals 2 +/- 3%, t_rec equals 0.6 +/- 0.1 min).

Aim of this study was to investigate the oxygenation and the total hemoglobin concentration pattern in vastus lateralis and medial gastrocnemius muscle groups during a standardized treadmill exercise (n equals 6) by a new near infrared frequency-domain oximeter (ISS Oximeter model 96208). Vastus lateralis saturation and total hemoglobin concentration were 74 +/- 3% and 71 +/- 15 (mu) M at 0 mph and 72 +/- 5% and 79 +/- 16 (mu) M at 6 mph, respectively. Gastrocnemius saturation and total hemoglobin concentration were 74 +/- 2% and 107 +/- 18 (mu) M at 0 mph and 60 +/- 6% and 113 +/- 23 (mu) M at 6 mph, respectively. The saturation recovered gradually up to the baseline value when the speed was decreased.

Noninvasive measurement of changes in oxygenation of human skeletal muscle can be done with a dual-wavelength near infrared (NIR) spectrophotometer. This allows a noninvasive investigation of muscle mitochondria. An exercise protocol was developed to study the load dependent changes in oxygenation of m. vastus lateralis of myopathic patients. On a bicycle ergometer exercise was done periodically. One period consisted of 1.5 min exercise followed by 3 min rest. Work load in the first period was 20 W, and was increased by 10 W for each subsequent period until maximal work load was reached. In 12 healthy volunteers we observed oxygenation of muscle during periods of low work load (warm-up effect). During periods of high work load the muscle deoxygenated. The work load at transition from oxygenation to deoxygenation (deoxygenation threshold) in controls was 75 W. In 3 patients with myopathies, in addition to NIR- spectroscopy, function of mitochondria of specimen of m. vastus lateralis was investigated biochemically. Muscle fibers were skinned with saponin and investigated with high resolution respirometry and multiple substrate-inhibitor- titration. Mitochondrial function was impaired in patients who had abnormal findings in NIR spectroscopy.

Near infrared spectroscopy (NIRS) has been used to monitor oxygenation changes in muscle. Quantitative values for O₂ consumption, blood flow and venous saturation have been reported by several investigators. The amount of these measurements is, however, still limited and complete validation has not yet been established. The aim of this study was to investigate the different NIRS methods to calculate O₂ consumption (VO₂) and forearm blood flow (FBF) and to validate the data with the accepted method of strain-gauge plethysmography and blood sampling. Thirteen subjects were tested in rest and during static isometric handgrip exercise at 10% MVC. The NIRS optodes were positioned on the flexor region of the arm. A significant correlation was found between plethysmograph data and NIRS [tHb] during venous occlusion in rest (r EQ 0.925 - 0.994, P < 0.05) as well as during exercise (r equals 0.895 - 0.990, P < 0.05). No correlation was found, however, for the calculated FBF and VO₂ values between NIRS and the combination of plethysmography and blood sampling. In rest nor during exercise. It seems that although NIRS is a good qualitative monitoring technique, quantification is difficult due to the great variability that is found between the subjects.

We present a method for determination of muscle blood flow (MBF) using near infrared spectroscopy (NIRS) with indocyanine green (ICG) as the tracer. MBF was quantified using the integrated arterial [ICG] and the accumulation of ICG in muscle. MBF was determined together with ICG-assessed cardiac output (CO) at rest and during incremental cycling. To further modify CO, the same work loads were performed after cardio-selective beta blockade by metoprolol. In one subject both MBF (9 to 110 ml (DOT) 100 g^-1 (DOT) min^-1) and CO increased linearly with work rate (8 to 19 l (DOT) min^-1). Under beta blockade, both the increase in MBF and CO were lower: 5 to 70 ml (DOT) 100 g^-1 (DOT) min^-1 and 5 to 161 DOT min^-1, respectively. During exercise with and without beta blockade, MBF increased with work load to represent a larger proportion of CO. Also, NIRS could detect an attenuated increase in MBF manifest by the restrained CO during leg exercise after cardio-selective beta blockade. Both observations indicate that NIRS detection of indocyanine green provides an estimate of muscle blood flow over the range from rest to intense exercise.

We have used three different oximeters to study finger capillary SatO₂, transcutaneous oxygen tension and vastus lateralis hemoglobin saturation and hemoglobin concentration in 6 volunteers under normobaric normoxia and hypobaric hypoxia. Simultaneously, the ^99mTc-HMPAO uptake in the thigh muscles was assessed by planar scintigraphy. We found a highly selective uptake of ^99mTc-HMPAO in the muscles of all subjects thighs. ^99mTc-HMPAO uptake was significantly higher in hypoxia as compared to normoxia (p < 0.001). By comparing scintigraphy and tissue spectrophotometry (OMNIA), we observed a strong correlation between ^99mTc-HMPAO uptake and hemoglobin saturation (R equals 0.96, p < 0.001) and a good correlation between ^99mTc-HMPAO uptake and hemoglobin concentration (R equals 0.77, p < 0.05). These correlations indicate the occurrence of a protective vasodilatory response during hypobaric hypoxia. During hypoxia, none of the three methods used to measure oxygen saturation correlated with the actual arterial SatO₂. This results suggest that different body districts react to hypoxia in a non-uniform manner. The role of the used NIRS instrumentation in assessing quantitative values needs further investigations.

The influence of the subcutaneous adipose tissue on near infra-red light propagation in leg muscle and on the sensitivity of NIR O₂ monitor is investigated. The simulation model is half infinite medium consisting of skin, adipose and muscle layer. The optical properties of each layer are chosen from the reported data and the absorption coefficient of muscle is rectified by using experimental results. The thickness of adipose layer is varied from 0 - 15 mm and the partial optical pathlength in the muscle layer and the spatial sensitivity profile are predicted by Monte Carlo simulation. The spatial sensitivity profile tends to distribute in the adipose layer and the detected light only grazes the muscle layer. These results suggest that the sensitivity in oxygenation of muscle tissue is impaired by influence of the adipose tissue and NIR O₂ monitor mainly detects the oxygenation change in shallow region of the muscle layer.

Although the inhomogeneity of tissue structure affects the sensitivity of tissue oxygenation measurement by reflectance near-infrared spectroscopy, few analyses of this effect have been reported. In this study, the influence of a subcutaneous fat layer on muscle oxygenation measurement was investigated by Monte Carlo simulation and experimental studies. In the experiments, measurement sensitivity was examined by measuring the falling rate of oxygenation in occlusion tests on the forearm using a tissue oxygen monitor. The fat layer thickness was measured by ultrasonography. Results of the simulation and occlusion tests clearly showed that the presence of a fat layer greatly decreases the measurement sensitivity and increases the light intensity at the detector. The correction factors of sensitivity were obtained from this relationship and were successfully validated by experiments on 12 subjects whose fat layer thickness ranged from 3.5 to 8 mm.

We measured the changes in oxygenation of the brain in animals and humans during changes in oxygen delivery (hypotension, shock) or during brain activation (such as finger movement). We found that such changes were measurable. In addition, we found that we were able to detect a change in the oxygenation signal in an animal undergoing heart surgery before the problem was clinically noted, thus suggesting that such methods may allow for earlier intervention in the intensive care setting.

Each tissue has a unique spectral signature (e.g. liver looks distinct from bowel due to differences in both absorbance and in the way the tissue scatters light). Therefore, we suspect that automated discrimination among tissue types (e.g. blood, nerve, artery, vein, muscle) or tissue state (frozen, unfrozen, viable, dead) is feasible. In this study, we investigated our ability to detect hidden structures (such as blood vessels) or events (such as tissue ablation via freezing) using optical systems. For blood vessel localization, a key step in vascular access, we resolved the component concentration of hemoglobin measured within the tissue, and found that blood vessel depth and direction could be determined. For freezing detection, we found that changes in effective absorbance during freezing allowed the freezing process to be monitored spectroscopically. Such optical techniques may usher in use of light-assisted medical diagnosis, leading to automated and portable diagnostic devices which enable real-time diagnostics and monitoring during medical interventions, such as cryoablation or vascular access.

Optical tomography is a new modality of noninvasive diagnosis for imaging the distribution of optical properties in human bodies. In the process of developing optical tomography systems which can be applicable to human heads for diagnosis of disease and study of brain functions, we need realistic optical phantoms which anatomically and optically simulate human heads with complicated and multi- layered structures. Previously we have reported design and fabrication methods of optical head phantoms based on an MRI human head image. The phantoms have simulated the multi- layered structure with different optical properties specified to each layer. They had five layers; i.e., skin, skull, cerebrospinal fluid layer, gray matter and white matter. Also some inclusions simulating hematoma were embedded in some part of the head phantoms. However, the phantoms were made all of solid plastic resin and the temporal variation of physiological functions in brain could not be simulated by the static phantoms. We have improved the fabrication method and succeeded in making dynamic optical phantoms which are able to include some parts simulating the temporal variation of blood flow rate and oxygenation status.

The Philips Optical Mammoscope Prototype is a versatile CW optical mammography apparatus with which fully 3D images of breast tissue can be obtained at several wavelengths. We present the properties of the machine, and show the first images obtained with it. We elaborate on the tissue parameter that can be quantitatively reconstructed from the measurements, and show some preliminary research into the natural variation of this parameter.

We have measured transillumination spectra collected at various sites on the breasts of normal subjects. The absolute concentration of hemoglobin, lipids and water are derived using ratiometric methods. The lipids and deoxyhemoglobin signals overlap at 760 nm; the accurate separation of these thus requires explicit account to be taken of the diffusion of photons through the breast. The concentrations of Hb, lipids and water vary substantially from site to site, which will influence wavelength selection in imaging systems. Finite element calculations and solid phantom measurements demonstrate that this variability is not due to boundary effects. The spatial variation of (mu) _a at wavelengths 720 nm, 820 nm, 930 nm and 970 nm is illustrated schematically. It is noted that (mu) _a at 930 nm shows the least heterogeneity amongst the 8 normal subjects studied.

The effects of the background optical properties on the performance of imaging of realistic absorption and scattering inhomogeneities in diffusive media were studied experimentally by varying the optical coefficients of both background and inhomogeneity, while keeping constant the optical contrast between the two media. Images were constructed from a matrix of time-resolved transmittance measurements performed with time-correlated single photon counting. Data were collected every 0.2 cm over a 6 cm X 6 cm area from realistic tissue phantoms containing cylindrical inhomogeneities (1 cm height and 1 cm diameter) embedded in a 5 cm thick turbid slab. The optical coefficients of the background were varied in the ranges of 5 - 15 cm^-1 for the transport scattering and 0.02 - 0.08 cm^-1 for the absorption. The optical contrast for the inclusions was kept at values between -50% and +50% for the scattering and between -75% and +300% for the absorption. The data show that a strongly scattering and/or absorbing background can improve the detection capability of time-resolved imaging. From these results, advantage can be taken for a proper choice of the illumination wavelength to be used in transmittance imaging for diagnostic purposes.

We describe the design of a versatile electronic system performing a lock-in detection in parallel on every pixel of a 2D CCD camera. The system is based on a multiplexed lock- in detection method that requires accurate synchronization of the camera, the excitation signal and the processing computer. This device has been incorporated in an imaging setup based on the optical coherence tomography principle, enabling to acquire a full 2D head-on image without scanning. The imaging experiment is implemented on a modified commercial microscope. Lateral resolution is on the order of 2 micrometers , and the coherence length of the light source defines an axial resolution of approximately 8 micrometers . Images of onion cells a few hundred microns deep into the sample are obtained with 100 dB sensitivity.

The aim of this research is the investigation and improvement of the possibilities of optical coherence tomography (OCT) for evaluation and imaging of dermal lesions, such as melanoma. Single point detection, as done with fiber arrangements using photomultiplier for measurements on human eyes was not successful in reaching high resolutions in strongly scattering media as human skin. The maximum scanning depth was about 500 micrometers . For this reason, a new system with a wide area detection was developed. The experimental set-up consists of a highly sensitive slow-scan CCD-camera and a suitably adapted interferometric arrangement. Measurements were performed on a model made of scattering material with simple objects embedded. They show that images appear to be blurred as a consequence of the wide area detection which allows an increased measuring depth. These investigations lead to a new method for the improvement of the OCT images which is explained and demonstrated on measurements performed on the model. Measurements on preserved specimen of human skin up to a scanning depth of 2 mm are shown in comparison to corresponding histological slides to demonstrate the spatial resolution and the ability of optical coherence methods to distinguish structures with different optical parameters.

In this paper we discuss the application of higher-order transport approximations using finite element-spherical harmonics methods (FE-P_N) to multidimensional photon propagation problems. The combined methodology offers fast and accurate modeling of photon propagation in multidimensional diffusive and non-diffusive media. This is of great importance to the practical solution of the inverse scattering problems which characterize optical tomography.

We present both experimental data and theoretical calculations of near-IR light backscattered from tissue like phantoms with hidden objects. The experimental setup constitutes an innovative detection system with a set of optical fibers sticked into a 2 ps resolution streak-camera scanning the light reflected by the illuminated sample. The data show that small, mm-size objects can be detected up to 12 mm in depth from the illuminated surface. The theory, that solves exactly the diffusion equation and introduces the sample boundary conditions, is in excellent agreement with the data.

It has been suggested that phased array detection is a highly sensitive method of detecting and locating an inhomogeneity embedded within a scattering medium e.g. a tumor embedded in breast tissue. No consideration has so far been made to the effects of noise within the system leading to detection errors. This paper uses a probabilistic approach to find the optimum detector threshold and detection criteria. The technique is demonstrated using experimental data obtained from a novel phased array system designed to eliminate noise due to amplitude fluctuations between the two sources. A brief comparison with single beam frequency detection using the probabilistic method is also made. The technique has applications in optimizing the detection of signals due to inhomogeneities in the presence of a noisy background.

Propagation of photons in multiple scattering media depends on absorbing and scattering properties as well as the boundary conditions of the semi-infinite medium. A new method is shown that makes use of differences in boundary conditions to determine the optical properties. Induced are these different conditions by varying the reflectivity of a sensor head. We describe the influence of the change in reflectivity with the common diffusion theory. By building a ratio between the spatially-resolved diffuse reflectance under different boundary conditions it is possible to calculate the optical properties of homogeneous phantoms. Due to optical heterogeneities in living tissue, limitations of the method was observed, which restricts the application to in vivo measurements.

Local and superficial optical characterization of biological tissues can be achieved by measuring the spatially resolved diffuse reflectance at small source-detector separations. The sensitivity of the signal to the phase function, absorption and scattering coefficients were studied using Monte Carlo simulations. Measurements of spatially resolved reflectance were performed in vivo on human brain with source-detector separations from 0.3 to 1.5 mm. Distinct optical properties were found between normal cortex, astrocytoma of optic nerve and normal optic nerve.

Light propagation in two-layered turbid media having an infinitely thick second layer is investigated in the steady- state, frequency and time domains. A solution of the diffusion approximation to the transport equation is derived employing the extrapolated boundary condition and the Fourier transform technique. We compare the reflectance calculated from this solution to that computed with Monte Carlo simulations and show good agreement. The derived equations are used to calculate mean optical path lengths in a two-layered model representing a fat layer lying above a muscle layer. The results are applied to interpret near infrared spectroscopy measurements on skeletal muscle. It is shown that the fat layer influences strongly the measurements even if the separation of the source and detector is large.

The measurement of absorption and scattering optical parameters (mu) _a and (mu) '_s is difficult in vivo. We developed a new method for non invasive measurement of these coefficients. It is based on existing steady state reflectance methods but it reduces the sensitivity to noise with an integration method. It also decreases the dependence on the sample roughness and on the angular position of the source regarding the sample. The image is circularly integrated versus the radius from the maximum to the nearest side. The integral reflectance is then fitted with a theoretical formula using 2 point sources model. We validate the method on simulation images and on calibrated Latex beads suspensions used as scattering phantoms. The integral reflectance method has then been applied to the measurements of (mu) _a and (mu) '_s coefficients of the human skin in vivo. We have obtained a good reproducibility of the results.

The development of time-resolved optical diagnostic techniques for biomedical applications requires an accurate description of the time-dependent photon propagation in tissues. In many applications the diffusion approximation is used for this purpose. However, in case of a highly anisotropic scattering and in the vicinity of light sources the diffusion equation becomes inadequate. To overcome this limitation, we introduce another approximation of the time- dependent radiative transfer equation. The approximation is based on the assumption that the scattering phase function of the medium is strongly forward-peaked, which has been established for a variety of tissues. We show that in this case, the integro-differential time-dependent transfer equation can be reduced to a partial differential equation. Furthermore, we demonstrate that this approximate equation is valid at much shorter distances from the source than the diffusion equation. At the same time, this approach is amenable for a combination with an inverse technique in order to determine the optical properties of the medium from a time- or frequency-resolved experiment.

The optical parameters of a red blood cell suspended in the blood plasma, namely, the scattering and absorption cross sections and the phase function describing the small-angle distribution of the scattered light are calculated. Dependence of the optical parameters on all possible values of sizes, shapes, orientations, hemoglobin concentration, and oxygenation degree is considered. The data are calculated with the so-called straight-ray approximation. The accuracy of the approximation is estimated by comparison with the Mie theory.

A semi-analytic diffusion model to describe frequency domain imaging through heavily scattering media has been validated by a quantitative comparison with Monte Carlo simulations and experimental results. Although the model is not an exact solution it can deal with situations where exact analysis fails. The method lends itself to modeling responses from restricted objects and deals with the fact that the object is embedded in a finite medium. The presence of inhomogeneities is accounted for by a two stage propagation of Green's functions from source to object to detector. Using a combined photon flux and photon density term at the object plane to represent the radiance provides the best match to Monte Carlo simulations.

OCT with wide area detection has proven to be a good instrument to obtain tomographic images of epithelial structures. A depth of 2 mm was reached with a CCD-camera setup. Due to the lack of an effective image preprocessing the data acquisition time was too long to carry out measurements in vivo. A new detector head using line detector arrays was developed with the advantage of interference detection by complementary detectors and data reduction by analog computing.

Comparison of lucigenin- and luminol-dependent chemiluminescence (LC-CL and LM-CL, respectively) in nondiluted healthy donors' blood revealed significant differences in their patterns. LM-CL was low in fresh blood and disappeared after it storage for 3 hours. LC-CL was already high in fresh blood and was steadily increasing with blood storage. Serial dilution of blood with saline after addition of chemiluminescence indicators resulted in elevation of LM-CL, but decrease in LC-CL. LM-CL elevation after the initiation of respiratory burst (RB) in blood with zymosan was observed only in aerated samples and immediately dropped down when air supply to a blood sample was ceased. On the contrary, LM-CL did not depend on air supply to a blood sample for about 30 min. after RB initiation. The results suggest that there are at least two mechanisms for reactive oxygen species production in nondiluted blood. The first one is reflected predominantly by LM-CL. It is activated during RB and uses prevalently oxygen dissolved in cell medium. Another one is reflected predominantly by LC- LM. It does not depend upon initiation of RB in neutrophils, operates in blood constantly, and uses oxygen supplied by erythrocytes. It needs blood integrity for its manifestation.

An integral equation for the probability density of the photon migration is constructed. The medium is assumed to be infinite and scatter photons isotropically. The solution is found as an expansion over the number of scatterings. The series occurs to converge rapidly suppose the travel time is fixed. The diffusion approximation is obtained from the integral equation in the limit of large travel time t and distance r such that r << t. It is shown that the probability density tends to infinity near the front. The density calculated is compared with the diffusion limit and the results of a method using path integrals. The latter results appear to be generally poorer than the diffusion theory ones, especially near the front.

A modification of the Monte Carlo simulation of light propagation in turbid isotropically scattering media is proposed. The modification consists in that, besides simulating free paths, multi-scattering photon migrations are statistically modeled by using the exact values of the transition densities obtained for infinite media. An estimate of the efficiency of the method is given. The method is shown to essentially reduce the computing time when modeling the photon propagation far from the boundaries. The angle distribution for the photon after multi-scattering migration is derived. The advantages and drawbacks of the method suggested are discussed.

This work reports a spectroscopic investigation on controlled size latex particle solutions. Theoretical estimates of scattering coefficients and g-factor have been obtained by Mie theory. The measurements have been carried on using time-resolved transmittance technique. Numerical evaluation of optical properties have been obtained from a comparison between experimental data and analytical solutions to diffusion equation. A Ti:Sa laser system (pulse duration: 100 fs; repetition rate: 76 MHz; average power: 1.5 W at 800 nm; investigated range: 790 - 920 nm) and a streak camera with a temporal resolution of few picoseconds have been used. The experimental results allowed us to analyze the applicability of Mie theory when different wavelengths, particle sizes, and scattering conditions are under investigation.

In last years time-resolved transillumination seems to be particularly interesting for human tissue imaging. In spite of the huge quantity of work already done in this field, it seems still necessary to investigate in a systematic way the interdependence among the basic image parameters: spatial resolution, contrast and noise. This work reports the preliminary results of such a study using a Ti:Sa laser and a streak camera. The investigated samples have been Intralipid solutions with an object test introduced in them. Measurements on spatial resolution have been performed by measuring edge response function and the line spread and modulation transfer functions have been calculated. Moreover noise pattern and contrast have been evaluated.

We have used, in our paper, the Monte Carlo method to study the multiple light scattering effect on the medium polarization characteristics. We have estimated the influence of a scattering system parameters and experimental conditions on the LSM measurement results. In our model a monodisperse system of spherical particle was used. The scattering substance was considered to be placed in a cylinder cell. Calculations were conducted for various parameters of incident beam, cell diameter and scattering particles dimensions. The calculations were performed for three values of particle diameters and different concentrations.

Photoacoustic spectroscopy is powerful method of investigation of absorption spectra of scattering media. This method was successfully applied for investigation of different biotissues. However the multiple scattering is essential in these media. It disturbs the output photoacoustic signals. So it is necessary take this phenomenon into account. The purpose of our work is to analyze the variations of photoacoustic signal versus the multiplicity of light scattering. Calculations are held for the model monodisperse spherical particles. Mie theory was applied to describe scattering by single spherical particle. The case of Gaussian profile beam was considered. Monte- Carlo method was used to calculate the photons trajectories in three dimension space. Such events as photon absorption, scattering and reflection from the sell walls was evaluated by random numbers with appropriate probability density. Angular dependence of probability density of the photon scattering is determined in terms of Mie theory for depolarized light. So we define coordinates of photon's collision and absorption points. It enables us to determine the spatial distribution of the points of photon absorption, which depends on the shapes of scattering volume and beam profile. These calculations were made for the model of former elements of blood. It was shown that the multiple scattering varies the absorption profile of the beam. Since multiplicity of light scattering depends on the wavelength of incident light, then absorption spectrum is disturbed. The depth of light penetration into the medium appears to be essentially smaller in the case of short wavelengths.

We consider an anisotropically scattering layered cylinder or sphere illuminated by a point source. The scattered field generated by these bodies is calculated in terms of the appropriate mode decomposition of the radiance. This leads to mode-expansions for the radiance which are very similar to the mode expansions derived for the diffraction- scattering problems of classical mathematical physics.

Optical radiation plays an important role in the life of the human and animals. Particularly the ultraviolet radiation (< 0,38 mkm) which fulfill antirachitic action, erythema and immunologic action. Knowledge ways penetration of the optical energy to the organism and the transformation mechanism of its energy will permit correctly to dose the biological action, to raise which the productivity of the organism. Humans and animals are covered with the skin which is covered by the hair and by the hearse-cloth they perceive the energy of the optical irradiation. This manuscript is devoted to the presentation of the experimental investigation results for studying opportunity of the optical irradiation penetration into the biologically active subcutaneous layer through the skin or by the separate hair as by the light channel.

We consider imaging a spherical inhomogeneity (object) imbedded in a scattering and absorbing medium. We use the Bayesian approach to study the error bars on the parameters to be reconstructed (volume, absorption and scattering coefficients), and on combinations of them. We study the characterization threshold, i.e. the minimum size of the object permitting its characterization with a reasonable accuracy. We are interested in (1) absorbing and scattering spheres versus purely absorbing or purely scattering ones; (2) modulated (frequency-domain) versus CW illumination. We show that CW illumination cannot characterize absorbing and scattering-objects except for very large sizes. We demonstrate that a significant reduction of the characterization threshold can be obtained if complete characterization is replaced by a partial one, giving information only on the nature (absorbing versus scattering) of the object.

The trajectory approach to the problem of the optical imaging through the strongly scattering media with given macroinhomogeneities was considered. It was shown, that the shadow inside an object with any shape can be calculated by the representation of the photon mean path integrals. The influence of the object's boundary to the photon path statistical characteristics was investigated. The corresponding graphical dependencies conveniently illustrating trajectory alterations were represented. A comparative study of the three basic boundary geometry such as semi-infinite medium, flat layer and rectangular sector showed that the trajectory R((tau) ) of the photon statistical distribution center can be approximated by the three-segment polygonal line and the photon path root-mean- square deviation (Delta) ((tau) ) and the value d(tau) /dl ((tau) ), which is in inverse proportion to the speed of the distribution center movement can be replaced by the simplified functions in general case.

The changeable scattering properties of fibrous tissue like the human sclera were studied in vitro. The collimated light transmittance, photon-correlation, and polarization measurements were done. The optical characteristics of fibrous tissue were controlled using various osmotically active solutions including glucose. It was shown that refractive indices matching of collagen fibrils and base material of fibrous tissue is responsible for tissue optical enlightening and coherent beam transmittance.

Simple optical model of pulp, surrounded by bone shell, is suggested. Results of Monte-Carlo simulations of Doppler shift ware presented for random medium that containing moving particles. The single-layered and two-layered configurations of the medium are considered. Doppler shift of the frequency of incident laser light is investigated as a function of such parameters as absorption coefficient, scattering coefficient, and thickness of the medium.

A portable CW tissue oximeter of a 10-Hz sampling rate was developed for examination of pulsatile components of the output signals as a mean of checking the signal reliability during long-term monitoring. Feasible studies were performed on a healthy subject. Changes in Hb and HbO₂ signals of cerebral tissue were continuously measured by placing a photoprobe on the forehead during 6-hour sleep. Pulsatile changes in Hb and HbO₂ were steadily observed over a whole period of the recording. The phase relation of pulsation in Hb and HbO₂ was almost inverse. Not only information for reliable monitoring but also physiological parameters with respect to cerebral circulation and metabolism could be obtained by measuring the pulsatile components.

The cutaneous hemoglobin of the forehead can contribute to near infrared spectrophotometric cerebral tissue oximetry in both light scattering and absorption. Wide distance between the light source and detector was theoretically proved to increase substantial signal from cerebral tissue, however increase in signal to noise ratio was practically undesirable. The tight compression of the probe onto the head expels the subcutaneous blood, which is supposed to improve cerebral signal detection. The spectrophotometric effect of probe compression with a head band was examined by NIRS monitor having temporal resolution of 10 Hz which provided detailed information in tissue hemoglobin signals including pulsatile fluctuations. The measurement for healthy volunteers was done in both supine and sitting positions. The oxygenated Hb and total Hb levels significantly decreased with probe compression in supine position. Compared with pre-compression state, the amplitude of Oxy-Hb pulsation was reduced to 61.0 +/- 10.0 (mean +/- SD)% in sitting position, and 53.3 +/- 68% in supine position (p < 0.01). The postischemic hyperemia was also observed in every measurement. The results indicate that cutaneous blood in the scalp significantly contributes pulsatile Hb signals in cerebral NIRS, and probe compression is a good measure to reduce scalpel effect especially in supine position.

We develop a stochastic approach for the description of wave propagation in random media and present a derivation of the time independent bilocal radiative transport equation which generalizes the Boltzmann transport equation by taking into account wave scattering effects by correlated scatterers. We obtain an expression for the extinction coefficient that includes effects of correlation.

Monte Carlo simulations have been employed to examine several different techniques proposed for the improved detection of scattering inhomogeneities in turbid media. Scaling formulas were used to determine the effect of a scattering inhomogeneity starting from a set of trajectories evaluated for a homogeneous medium. In a relatively short time it was possible to vary both the dimension and the optical properties of the inhomogeneity and to evaluate the temporal point spread functions (TPSFs) for a large number of positions. Using the derived TPSFs, a comparison was performed of the images acquired using techniques based on CW, time domain, and frequency domain measurements.

An approach of using phase modulation spectroscopy (PMS) system has been proposed which has single frequency, dual- wavelength with phase-only output for oximetry to minimize the influences of room light in the clinical environment. The availability of this method has been verified by laboratory experiments both in vitro and in vivo. In this study, a 200 Mhz PMS system has been used for measuring on piglet brain in vivo, to detect blood volume and saturation change during normoxia to hypoxia periods, and to monitor the brain depolarization in the hypoxic stress. Our experimental results show that the brain saturation estimated by PMS is approximately 84% and lies between arterial and venous blood values of 97% and 80% in normoxia, and tends to move close to the venous side in hypoxia. In addition, the excessive hypoxic stress triggers the brain to be bio-energetic deficits and finally introduced the neural depolarization, which can be effectively detected by recording the increment of tissue scattering.

In last years NIR radiation was extensively used to detect targets hidden in optically turbid media. We performed transillumination measurements on tissue-like phantoms (absorbing objects embedded in Intralipid and black ink solutions) to determine the intensity of transmitted radiation. The experimental apparatus include a diode laser ((lambda) _emiss equals 820 nm, CW power equals 250 mW), a detection fiber (diameter equals 600 mm), a PMT and a Digital Signal Analyzer (bandwidth equals 1 GHz). At variance of very complex systems that reject highly scattered photons by time-resolved detection, the developed system performs this selection by collimation. The experimental data were compared with diffusion approximation predictions.

The purpose of this study was to examine the relationship between histochemical characteristics obtained from the gastrocnemius and changes in muscle energetics. Muscle oxygenation was determined by near infrared spectroscopy, and phosphocreatine (PCr) by 31-phosphorus magnetic resonance spectroscopy (³¹P-MRS) during a submaximal plantar flexion exercise and recovery. The relative occurrence (%) of slow-twitch fibers (ST), fast-twitch oxidative fibers (FTa), fast-twitch glycolytic fibers (FTb), and the number of capillaries per fiber (Cap/Fiber) were also determined. The sum of %ST and %FTa (%ST + %FTa) was 85.8 +/- 8.74% (mean +/- SD), while the number of Cap/Fiber was 2.52 +/- 0.63. The initial rate of deoxygenation (Rate_deoxy) at the beginning of the exercise was 2.43 +/- 0.95 %(DOT)sec^-1. The time constants (Tc) for the recovery of PCr and muscle oxygenation after exercise were 22.1 +/- 6.3 sec. and 20.3 +/- 13.6 sec., respectively. The %ST + FTa, and the number of Cap/Fiber were each positively correlated to the Rate_deoxy (P < 0.05). The %ST + %FTa, and the number of Cap/Fiber were negatively correlated to the Tc for PCr recovery (P < 0.05), but not correlated to that for muscle oxygenation recovery. In conclusion, the Rate_deoxy during localized submaximal exercise, and the Tc for PCr recovery are indicators of the muscle's oxidative capacity.

Three ways are explored to improve the accuracy of currently used Near InfraRed Spectroscopy algorithms, based on the Beer-Lambert law, for quantification of chromophore concentration changes in living tissues. Therefore, new two- steps models are developed in which the estimates obtained in the first step are refined in the next step with the use of prior knowledge. In particular, when certain chromophore concentration changes are assumed to be small or in chemical equilibrium it is shown that these models can considerably improve the accuracy of the computed concentration changes of Hb, HbO₂, Cytaa₃, and Icg, if present, compared to the conventional least squares regression based algorithm.

In the past, absolute quantification of tissue chromophore concentrations has not been possible using near infrared spectroscopy (NIRS). On the contrary, the Critikon 2020 Cerebral Redox Monitor (Johnson and Johnson Medical Ltd) provides real time quantified value for oxyhemoglobin (Hb02), deoxyhemoglobin (HbH) and total hemoglobin (THb) concentrations. It achieves quantification by incorporating the following features : • a 2 detectors approach (to delete extracranial contribution) • an algorithm based on quantified absorption spectra • inclusion of an optical pathlenght and a wavelength dependent scattering factors

An effective way to investigate the competence of NIR Imaging is to associate NIR images with ones from other, well established modalities, such as X-ray Mammography, Tomography or Magnetic Resonance Imaging (MRI). MRI is an excellent candidate for such an approach since besides having immense sensitivity and rendering great anatomical information, it also employs non-ionizing radiation. The simultaneous acquisition of MR and NIR data may afford very precise co-registration of images and investigate the potential and limitations of the technique. In return, by establishing confidence on the NIR Tomography capability, the MR specificity may be enhanced due to the additional information content of the simultaneous examination. We have developed a 24 X 8 source-detector multi-channel NIR imager and spectrometer based on the time-correlated single photon counting technique. The instrument is capable of operating as a stand-alone modality or coupled to an MR scanner. We have tested the instrument as a breast imager with volunteers and patients in the MR examination room. We have used specially designed soft compression plates bearing the optical fibers and the MR coils. The acquisition time has been optimized to be within the time limits of a typical MR breast examination protocol. We have obtained approximately 25 dB signal to noise ratio per sec of averaging time for 6 cm breast separation and sensitivity of absorption coefficient changes, following contrast agent administration, of the order of 10^-3 cm^-1. Additionally we demonstrate the use of the instrument as a stand-alone motor cortex activity imager/spectrometer.