Examination of Subjects Using Photon Migration With High Directionality Techniques

This invention relates to examination and imaging of biological tissue using visible or infra-red radiation.

Traditionally, potentially harmful ionizing radiation (for example, X-ray or .gamma.-ray) has been used to image biological tissue. This radiation propagates in the tissue on straight, ballistic tracks, i.e., scattering of the radiation is negligible. Thus, imaging is based on evaluation of the absorption levels of different tissue types. For example, in roentgenography the X-ray film contains darker and lighter spots. In more complicated systems, such as computerized tomography (CT), a cross-sectional picture of human organs is created by transmitting X-ray radiation through a section of the human body at different angles and by electronically detecting the variation in X-ray transmission. The detected intensity information is digitally stored in a computer which reconstructs the X-ray absorption of the tissue at a multiplicity of points located in one cross-sectional plane.

Near infra-red radiation (NIR) has been used to study non-invasively the oxygen metabolism in tissue (for example, the brain, finger, or ear lobe). Using visible, NIR and infra-red (IR) radiation for medical imaging could bring several advantages. In the NIR or IR range the contrast factor between a tumor and a tissue is much larger than in the X-ray range. In addition, the visible to IR radiation is preferred over the X-ray radiation since it is non-ionizing; thus, it potentially causes fewer side effects. However, with lower energy radiation, such as visible or infra-red radiation, the radiation is strongly scattered and absorbed in biological tissue, and the migration path cannot be approximated by a straight line, making inapplicable certain aspects of cross-sectional imaging techniques.

Recently, certain approaches to NIR imaging have been suggested. One approach undertaken by Oda et al. in "Non-Invasive Hemoglobin Oxygenation Monitor and Computerized Tomography of NIR Spectrometry," SPIE Vol. 1431, p. 284, 1991, utilizes NIR radiation in an analogous way to the use of X-ray radiation in an X-ray CT. In this device, the X-ray source is replaced by three laser diodes emitting light in the NIR range. The NIR-CT uses a set of photomultipliers to detect the light of the three laser diodes transmitted through the imaged tissue. The detected data are manipulated by a computer of the original X-ray CT scanner system in the same way as the detected X-ray data would be.

Different approaches were suggested by S. R. Arriadge et al. in "Reconstruction Methods for Infra-red Absorption Imaging," SPIE Vol. 1431, p. 204, 1991; F. A. Grunbaum et al. in "Diffuse Tomography," SPIE Vol. 1431, p. 232, 1991; B. Chance et al., SPIE Vol. 1431 (1991), p. 84, p. 180, and p. 264; and others who recognized the scattering aspect of the non-ionizing radiation and its importance in imaging. None of those techniques have fully satisfied all situations.

In summary, there continues to be a need for an improved imaging system which utilizes visible or IR radiation of wavelengths sensitive to endogenous or exogenous pigments.

The invention relates to systems and methods for spectroscopic examination of a subject positioned between input and detection ports of the spectroscopic system applied to the subject.

According to one aspect of the invention, a spectroscopic system includes at least one light source adapted to introduce, at multiple input ports, electromagnetic non-ionizing radiation of a known time-varying pattern of photon density and of a wavelength selected to be scattered and absorbed while migrating in the subject, the input ports being placed at selected locations on the subject to probe a selected quality of the subject; radiation pattern control means adapted to achieve a selected time relationship of the introduced patterns to form resulting radiation that possesses a substantial gradient in photon density as a result of the interaction of the introduced patterns emanating from the input ports, the radiation being scattered and absorbed in migration paths in the subject. The system also includes a detector adapted to detect over time, at a detection port placed at a selected location on the subject, the radiation that has migrated in the subject; processing means adapted to process signals of the detected radiation in relation to the introduced radiation to create processed data indicative of the influence of the subject upon the gradient of photon density; and evaluation means adapted to examine the subject by correlating the processed data with the locations of the input and output ports.

Preferred embodiments of this aspect of the invention include displacement means adapted to move synchronously the optical ports and the detection ports to another location on a predetermined geometric pattern; this other location is used to perform the examination of the subject.

According to another aspect of the invention, a spectroscopic system includes at least one light source adapted to introduce, at multiple input ports, electromagnetic non-ionizing radiation of a known time-varying pattern of photon density and of a wavelength selected to be scattered and absorbed while migrating in the subject, the input ports being placed at selected locations on the subject to probe a selected quality of the subject; radiation pattern control means adapted to achieve a selected time relationship of the introduced patterns to form resulting radiation that possesses a substantial gradient in photon density as a result of the interaction of the introduced patterns emanating from the input ports, the radiation being scattered and absorbed in migration paths in the subject. The system also includes a detector adapted to detect over time, at a detection port placed at a selected location on the subject, the radiation that has migrated in the subject; displacement means adapted to move the detection port to various locations on a predetermined geometric pattern, the various locations being used to detect over time radiation that has migrated in the subject; processing means adapted to process signals of the detected radiation in relation to the introduced radiation to create processed data indicative of the influence of the subject upon the gradient of photon density; and evaluation means adapted to examine the subject by correlating the processed data with the locations of the input and output ports.

According to another aspect of the invention, a spectroscopic system includes at least one light source adapted to introduce, at multiple input ports, electromagnetic non-ionizing radiation of a known time-varying pattern of photon density and of a wavelength selected to be scattered and absorbed while migrating in the subject, the input ports being placed at selected locations on the subject to probe a selected quality of the subject; radiation pattern control means adapted to achieve a selected time relationship of the introduced patterns to form resulting radiation that possesses a substantial gradient in photon density as a result of the interaction of the introduced patterns emanating from the input ports, the radiation being scattered and absorbed in migration paths in the subject. The system also includes at least one detector adapted to detect over time, at multiple detection ports placed at selected locations on the subject, the radiation that has migrated in the subject; processing means adapted to process signals of the detected radiation in relation to the introduced radiation to create processed data indicative of the influence of the subject upon the gradient of photon density, and evaluation means adapted to examine the subject by correlating the processed data with the locations of the input and output ports.

Preferred embodiments of this aspect of the invention include displacement means adapted to move at least one of the detection ports to another location on a predetermined geometric pattern, the other location being used to perform the examination of the subject.

Preferred embodiments of this aspect of the invention include rotation means adapted to rotate synchronously the optical input ports while introducing the resulting radiation along a predetermined geometric pattern, the input port rotation being used to perform the examination of a region of the subject.

Preferred embodiments of the above described aspects of the invention are also used to locate a fluorescent constituent of interest in the subject; the wavelength of the introduced radiation is selected to be absorbed in the fluorescent constituent, the detected radiation is emitted from the fluorescent constituent and processed to determine location of the fluorescent constituent.

According to another aspect of the invention, a spectroscopic system includes a light source adapted to introduce, at an input port, electromagnetic non-ionizing radiation of a known time-varying pattern of photon density and of a wavelength selected to be scattered and absorbed while migrating in the subject, the input port being placed at a selected location on the subject to probe a selected quality of the subject; detectors adapted to detect over time, at multiple detection ports placed at selected locations on the subject, the radiation that has migrated in the subject; the time relationship of the detection over time, at the detection ports, being selected to observe a gradient in photon density formed as a result of the interaction of the introduced radiation with the subject. The system also includes processing means adapted to process signals of the detected radiation in relation to the introduced radiation to create processed data indicative of the influence of the subject upon the gradient of photon density, and evaluation means adapted to examine the subject by correlating the processed data with the locations of the input and output ports.

Preferred embodiments of this aspect of the invention of the invention include displacement means adapted to move at least one of the detection ports to another location on a predetermined geometric pattern, the other location being used to perform the examination of the subject.

According to another aspect of the invention, a spectroscopic system includes a light source adapted to introduce, at an input port, electromagnetic non-ionizing radiation of a known time-varying pattern of photon density and of a wavelength selected to be scattered and absorbed by a fluorescent constituent while migrating in the subject, the input port being placed at a selected location on the subject to locate the fluorescent constituent of the subject; detectors adapted to detect over time, at multiple detection ports placed at selected locations on the subject, fluorescent radiation that has migrated in the subject. The system also includes processing means adapted to process signals of the detected radiation in relation to the introduced radiation to create processed data indicative of location of the fluorescent constituent of the subject, and evaluation means adapted to examine the subject by correlating the processed data with the locations of the input and output ports.

Preferred embodiments of this aspect of the invention include displacement means adapted to move at least one of the detection ports to another location on a predetermined geometric pattern, the other location being used to locate the fluorescent constituent of the subject.

Preferred embodiments of the above described aspects of the invention use one or more of the following.

The time-varying pattern of resulting radiation is formed by the intensity modulated radiation introduced from each of the input ports having selected phase relationship that produces in at least one direction a steep phase change and a sharp minimum in the intensity of the radiation.

The phase relationship of the introduced radiation patterns is 180 degrees.

The modulation frequency of the introduced radiation has a value that enables resolution of the phase shift that originates during migration of photons in the subject.

Other advantages and features of the invention will be apparent from the following description of the preferred embodiment and from the claims.

FIGS. 1 and 1A show diagrammatically a phase modulation imaging system including several input ports and one detection port in accordance with the present invention.

FIG. 2 is a block diagram of the phase modulation imaging system including several input ports and several detection ports in accordance with the present invention.

FIG. 2A depicts a phased array antenna that radiates a directional beam.

FIG. 2B depicts sequencing of the phases of an antiphase multi-element array to achieve an electronic scan of the photon density gradient in accordance with the present invention.

FIG. 2C depicts four element antiphased array designed for a conical scan of the photon density gradient in accordance with the present invention.

FIG. 2D depicts the input and output port arrangement of an imaging system in accordance with the present invention.

FIG. 3 depicts a phase modulation imaging system including an input port and several array detection ports in accordance with the present invention.

FIG. 4 is a block diagram of an alternative embodiment of a dual wavelength PMS system.

FIG. 4A is a schematic diagram of an oscillator circuit of FIG. 4.

FIG. 4B is a schematic diagram of a PMT heterodyne modulation and mixing network shown in FIG. 4.

FIG. 4C is a schematic diagram of an AGC circuit shown in FIG. 4.

FIG. 4D is a schematic diagram of a phase detector circuit shown in FIG. 4.

FIG. 5A, 5B, and 5C illustrate changes in optical field propagating in a strongly scattering medium which includes a strongly absorbing component.

FIG. 6 shows an experimental arrangement of a two element phased array used in an interference experiment.

FIGS. 6A, 6B, and 6C show detected interference patterns of two diffusive waves.

FIG. 7 displays the phase shifts measured for a two element array (curve A), and for a single source (curve B).

FIG. 8A depicts an experimental arrangement of sources of a four element phased array and a detector.

FIGS. 8B and 8C display the intensities and the phase shifts measured for the four element array of FIG. 8A, respectively.

FIG. 9A depicts an experimental arrangement of sources of a four element phased array, a detector, and a strongly absorbing object.

FIG. 9B, 9C display respectively the intensities and the phase shifts measured for the four element array of FIG. 9A scanning absorbing objects of different sizes.

FIG. 9D displays the phase shifts measured for the four element array of FIG. 9A scanning absorbing objects of different absorption coefficients.

FIG. 10A an experimental arrangement of sources of a four element phased array, a detector, and two strongly absorbing objects.

FIG. 10B displays the phase shifts measured for the four element array of FIG. 10A scanning two absorbing objects of different sizes.

Imaging system embodiments of the present invention based upon interferance effects of radiation migrating in a subject having scattering and absorptive properties are shown in FIGS. 1, 2, and 3. The systems effectively utilize, in this scattering medium, a directional beam of visible or IR radiation generated and/or detected by an array of sources and/or detectors, respectively. For instance, in the case of an array of sources, each source is placed at a selected location in the array and emits intensity modulated radiation, preferably coherent radiation from a laser diode, of a selected intensity and phase. The criteria for selecting the source locations, the intensities, and the phases of the respective sources is the shape of the desired beam that at any time point possesses a substantial photon density gradient produced by interference effects of radiation from the various sources. This gradient of photon density is localized and has directional properties. Overall, the resulting radiation formed by interference of the radiation of the individual sources migrates in a selected direction in the subject. In an antiphase system, the wavefront of the beam has sections of equal photon density separated by a sharp localized change in photon density. Selected different locations of the photon density gradient are shown in FIG. 2B.

In general, the wavefront propagates in the selected direction in the subject and the gradient of photon density is localized in one or more planes extending from the source array in a selected direction. If the subject includes a localized object having different scattering and absorptive properties from those of the surrounding environment, the propagating radiated field is perturbed. This perturbation is detected and from the source detector geometry the perturbing object can be located.

Referring to the embodiment of FIGS. 1 and 1A, the imaging system utilizes an array of laser diodes 12, 14, 16, and 18 for introducing light into the tissue at selected locations. The geometry of optical input ports 11, 13, 15, 17 and of an optical output port 19 is selected to examine a specific part of the tissue. From the known geometry of the optical input ports and the detection port and from the shape of the introduced and detected radiation, a computer can locate a hidden object 9 of examined tissue 8 (For example, a head or breast). A master oscillator 22, which operates at 200 MHz, excites laser diode 12 through 18, that emit light of a selected wavelength (e.g., 760 nm). The light from each laser diode is conducted to the respective input port placed on a subject via a set optical fibers. A detector 24 detects the light that has migrated through the examined tissue. Preferably, detector 24 includes a photomultiplier tube (e.g., Hamamatsu R928) powered by a high voltage supply which outputs about 900 V in order to ensure a high gain. A local oscillator 26 operating at a convenient offset frequency (e.g., 25 KHz) sends a signal to a mixer 28 and a reference signal to detector 24. Accordingly, an output waveform 25 from detector 24 is at a carrier frequency equal to the difference of the detected and reference frequency, i.e., 25 KHz.

Detector 24 (for example, PMT Hamamatsu R928 or Hamamatsu R1645u) detects the scattered and absorbed light that has migrated through the subject. Detection port 19 is located several centimeters from the location of the input ports. The PMT detector is connected to the subject by the fiber optic guide, or, alternatively, may be directly placed on the subject. It has been found that the most cost-effective detector for measuring signals of frequencies on the order of 10.sup.8 Hz is Hamamatsu R928. However, the Hamamatsu R1645u detector is preferred due to its high precision. The second dynode of the PMT of detector 24 is modulated by 200.025 MHz signal 27 so that the 25 KHz hetrodyned signal 25 is received by a phase detector 30. Phase detector 30 also receives reference signal 29 from mixer 28. If phase detector 30 is a lock-in amplifier then the output signals are the phase shift and the intensity of the detected signal. Both the phase shift and the intensity of the detected light characterize the migration path of photons in the subject (e.g., the brain tissue).

Alternatively, a tunable dye laser or other laser source connected to a wide band acousto-optical modulator operating at the carrier frequency, e.g., 200 MHz can be used instead of the laser diode. The acousto-optical modulator modulates the intensity of the light emitted by the laser at the selected carrier frequency.

The invention also envisions using only one source of coherent light that irradiates one end of several optical fibers at the same time. The other end of each fiber is placed on the subject at a selected input port location. This source radiates light of a selected time varying pattern. The phase relationship and the intensity of the light carried by each fiber is varied by creating a time delay (e.g., different fiber length) and by coupling different amounts of light into each fiber.

The imaging systems of FIGS. 1, 2, and 3 are shown to have a light source of a single wavelength; however, a dual wavelength imaging system is also envisioned according to this invention. In the dual wavelength imaging system two laser diodes or a tunable wavelength laser generate light of two wavelengths that is coupled to an optical fiber. Such a system will now be described.

A dual wavelength operation is shown in FIG. 4. The system includes a master oscillator 60 operating at 200 MHz and an oscillator 62 operating at 200.025 MHz which is offset 25 KHz from the master oscillator frequency. The offset frequency of 25 KHz is a convenient frequency for phase detection in this system; however, other offset frequencies as high as a few megahertz can be used. Oscillator 60 alternatively drives two laser diodes 64 and 66 using switches 61, 61a, . . . These switches are driven electronically to couple a selected wavelength into the optical fiber and also to achieve a selected radiation pattern resulting from the radiation emanating from the individual fibers. An output 8 mm fiber coupler 72 collects photons for an R928 PMT detector 74. The second dynode (shown in FIG. 3B) of PMT 74 is modulated with a 200,025 MHz reference signal generated by oscillator 62 and amplified by an amplifier 63. Thus, the output signal of the PMT detector has a frequency of 25 KHz. PMT detector 74 alternately detects light of the two laser diodes that has migrated in the tissue and produces corresponding output signals, which are filtered by a filter 78 and leveled by an automatic gain control (AGC) circuit 79. A reference signal of 25 KHz is produced in a mixer 65 by mixing the 200 and 200,025 MHz oscillator signals. The reference 25 kHz signal is also leveled using the second AGC 77 and fed into a phase detector 80. Phase detector 80 generates a signal indicative of the phase of each output signal relative to the phase of the reference signal. The outputs of phase detector 80 are alternately selected by an electronic switch 82, filtered, and then input to an adder 84 and a subtractor 86 to produce sum and difference signals proportional to <L>.sub..lambda.1 +<L>.sub..lambda.2 and <L>.sub..lambda..sub.1 -<L>.sub..lambda.2. The difference and sum signals are then used to calculate changes in the probed pigment and in the blood volume, respectively.

A schematic diagram of preferred oscillator 60 or 62 is shown in FIG. 4A. This circuit has a drift of only 0.03 degrees/hr. (Weng, et al., "Measurement of Biological Tissue Metabolism Using Phase Modulation Spectroscopic Measurement," SPIE, Vol. 143, p. 161, 1991, which is incorporated herein by reference). The crystal is neutralized, which enables operation at resonance, and thus achieves long-term stability. The respective crystals of oscillators 60 and 62 are offset from each other by 25 kHz. This circuit provides a sufficient output to directly drive a 5 mW laser diode.

A modulation circuit 75 for the second dynode of the PMT is shown in FIG. 4B. This circuit uses a resonant circuit 75a with an impedance of 20,000 ohms instead of the usual 50 .OMEGA. load with very high power dissipation, providing a 50 V drive of the photomultiplier dynode while dissipating only a few watts of power.

To obtain stable operation of the phase detector, a stable input signal is required. The 25 KHz AGC circuit 77 illustrated in FIG. 4C includes an MC 1350 integrated circuit U1, featuring wide range AGC for use as an amplifier. The signal amplitude is controlled by a feedback network, as shown. A major reason for the accurate detection of phase changes by the PMT system is that the phase detector input signal level is kept nearly constant by the AGC circuit. Since the input voltage change of between 2 and 6 volts causes variation in the phase shift of only 0.2%, the AGC circuit eliminates the need for a very stable high voltage power supply.

A preferred phase detector circuit is shown in FIG. 4D. Two sinusoidal signals (the measurement signal and the reference signal) are transformed to a square wave signal by a Schmitt trigger circuit 80a. The phase of the square wave signal is shifted by an RC change (composed of R11, R12, C8), which makes it possible to change the measuring range. The detector further includes a 74HC221 integrated circuit. The lock-in amplifier technique obtained to derive the difference of the phase and amplitude of the two signals has the highest signal to noise ratio possible for this type of equipment.

The above-described systems utilize the carrier frequency on the order of 10.sup.8 Hz which is sufficiently fast to resolve the phase shift of the detected light. The characteristic time, the time it takes for a photon to migrate between an input port and an output port, is several nanoseconds. The sensitivity of the system is high, approximately 70.degree. per nanosecond or 3.degree. per centimeter change of pathlength, as observed in experimental models. Selection of the modulation frequency also depends on the desired penetration depth and resolution of the imaging system that will be described below. If deep penetration is desired, low modulation frequency (e.g., 40 MHz) is selected, and if shallow penetration is needed, modulation frequencies as high as 10.sup.9 Hz can be used.

Referring to FIGS. 1 and 1A, a master oscillator 22 operates at a modulation frequency in the range of 40 to 400 MHz selected according to the desired penetration depth of the optical field. The array of laser diodes 12, 14, 16, and 18 generates a highly directional radiation pattern, which is employed in the tissue examination.

In one preferred mode of operation, laser diodes 12 to 18 operate in a phased array pattern which is introduced into the tissue and detected by a single PMT detector 30. Master oscillator 22 operating at 200 MHz drives a multi-channel phased splitter which gives outputs at predetermined phases. Input ports 11 through 17 are located at selected distances and an appropriate phasing of the array creates a directional beam and enables scanning of the optical field in two dimensions across the tissue, as shown in FIGS. 2A, 2B, and 2D. After migrating through the tissue, the optical field is collected in a large area fiber on selected locations 19. The detected signals are heterodyned in the PMT detector 24 by utilizing the output of local oscillator 26, operating at a 25 kHz offset frequency, to detector 24. The resulting 25 kHz signal is phase detected with respect to the output signal 29 of mixer 28 and detector 24. Phase detector 30 outputs the phase and the intensity of signal 25. The detected phase shifts and intensities are stored and used for construction of an image of the subject. This is performed by computer control 34, which governs the operation of the system.

FIG. 2 depicts a phase modulation imaging system comprising an input port array for introducing radiation and a detection port array for detecting radiation that has migrated in the subject. The operation of the system is controlled by computer control 34, which coordinates a transmitter unit 32 with a receiver unit 42. Transmitter unit 32 comprises several sources of visible or IR radiation adapted to introduce a selected time-varying pattern of photon density into subject 8 by array of input ports 31, 33, 35, and 37. Receiver unit 42 detects radiation that has migrated in the subject from the input port array to an array of detectors 39, 41, 42, and 47.

The radiation sources of transmitter unit 32 are intensity modulated at a frequency in the range of 40 MHz to 200 MHz, as described for the imaging system of FIG. 1. Receiver unit 42 detects and processes the radiation using the same principles of the phase and amplitude detection as described above. The signal detected at individual ports can be phased using appropriate delays.

Several modes of operation of the transmitter array and receiver array are described in FIGS. 2A, 2B, 2C, and 2D. Referring to FIG. 2A, it has been known, that for a simple horizontal linear array of N identical elements radiating amplitude modulated light spaced a distance, d, apart. The radiating wavefront is created by the interference effect. If all elements radiate in phase the wavefront propagates in a direction perpendicular to the array. However, by appropriately phasing the radiating elements, the resulting beam can scan space in two dimensions. We consider the phases of the signal along the plane A--A whose normal makes an angle .theta..sub.0 with respect to the array normal. The phase of the signal from the first radiator lags the phase of the second radiator by a phase angle (2.pi./.lambda.)d sin .theta..sub.0 because the signal from the second radiator has to travel a distance d sin .theta..sub.0 longer than the signal from the first radiator to reach plane A--A. Similarly, the phase of the signal from the n.sup.th radiator leads that from the first radiator by an angle n(2.pi./.lambda.))d sin .theta..sub.0. Thus, the signals from the various radiators can be adjusted to be in-phase along the A--A plane, if the phase of each radiator is increased by (2.pi./.lambda.)d sin .theta..sub.0. Consequently, at a point on the wavefront in the far field of the antenna array the signals from the N radiators will add up in phase, i.e., the intensity of the total normalized signal is a sum of the signals from the individual sources. The constructed pattern has a well defined directional characteristic and a well pronounced angular dependence, i.e., the antenna pattern has a well defined transfer characteristic of the antenna with respect to the angle .theta..sub.0.

FIG. 2B depicts an arrangement of phases for the sources the system of FIG. 2 operating in one preferred mode of operation. The array of five sources is divided into two or more portions that are phased 180.degree. apart. Each portion has at least one source. The sources of each portion radiate amplitude modulated light of equal intensity and are spaced so that the resulting beam of two or more equally phased sources has a substantially flat wavefront, i.e., no gradient of photon density. On the other hand, there is a sharp 180.degree. phase transition, a large gradient in photon density between two antiphased portions of the array. Thus, the radiated field possesses an amplitude null and a phase transition of 180.degree., which is due to the large gradient of photon density.

Electronic scanning is performed by appropriately varying the apportionment of 0.degree. and 180.degree. phases on the sources. The five element array of FIG. 2B can have the 180.degree. phase transition along four different parallel planes extending from the array. Scanning is achieved by electronically switching the sources by 180.degree., so that the photon density gradient moves in the direction parallel to the location of the sources.

Using the principles described in FIGS. 2A and 2B, a conical scan of a directional beam possessing at least one substantial photon density gradient can be accomplished using a four element antiphased array, as shown in FIG. 2C. The laser diodes are antiphased using a push pull transformer. The phasing and amplitude of four laser diodes S.sub.1, S.sub.2, S.sub.3, and S.sub.4 arranged into a two dimensional array is modified sequentially using the switches Sw.sub.1, Sw.sub.2, Sw.sub.3, and Sw.sub.6 and inductances L.sub.1, L.sub.2, L.sub.3, and L.sub.4.

FIG. 2D shows a possible arrangement of the transmitter array and the receiver array. The above described directional beam enters subject 8 at the transmitter array location and is pointed to hidden absorber 9 which perturbs the migrating beam. The field perturbation is measured by the receiver array. Scanning of the transmitter array or the receiver array is envisioned by the present invention.

A hidden absorber that includes a fluorescent constituent is detected using a selected excitation wavelength of the laser sources of the transmitter array. Then, the radiation is absorbed, and almost instantly a fluorescent radiation of a different wavelength is re-emitted. The re-emitted radiation propagating in all directions is detected by the receiver array.

FIG. 3 depicts a phase modulation imaging system comprising one input port and several arrays of detection ports. This system operates comparably to the systems of FIGS. 1 and 2. The 754 nm light of a laser diode 48 is amplitude modulated using master oscillator 22. The light is coupled to subject 8 using an input port 49. The amplitude modulated light migrates in the subject and is scattered from hidden object 9. It is also expected that hidden object 9 has a different effective index of refraction than subject 8. The migrating radiation is governed by the laws of diffusional wave optics that are described below. The scattered radiation migrates in several directions and is detected by detection systems 50, 52, and 54.

Ports 51, 53, and 55 of the detection systems can include either large area fibers or arrays of detection ports. If large area fibers are used then detector systems 50, 52, and 54 correspond to detector 24 of FIG. 1. If arrays detection ports are used, then each of detector systems 50, 52, and 54 includes several individual PMT detectors. The PMT detectors of each detector system are phased utilizing a selected phase mode, as described above. The phasing is controlled by the computer control. The detected signals are heterodyned at the PMT's and sent to a phase detector 58. Phase detector 58 detects alternatively the heterodyned signals using a switch 56. Operation of phase detector 58 is similar to the operation of phase detector 30 of FIG. 1. The detected phase and amplitude are alternatively sent to the computer control using a switch 56a. Even thought only one phase detector is shown in FIG. 3, the invention envisions use of several phase detectors.

If hidden absorber 9 includes a fluorescent constituent, laser diode 48 is selected to introduce an excitation wavelength (e.g., 754 nm). The introduced, intensity modulated radiation, excites the fluorescent constituent which re-emits radiation in all directions, as shown in FIG. 3. The re-emitted radiation is detected using detector systems 50, 52, and 54. To increase the system resolution, each detector can be furnished with an interference filter selected to pass only the fluorescent radiation.

The interference of several waves, as described in FIG. 2A, has been long known in a non-scattering medium, wherein the radiation propagates on a straight line, but not in a strongly scattering medium. Referring to FIGS. 6, 6A, 6B, and 6C, in a simple experiment, interference of two different diffusive waves in a strongly scattering medium was demonstrated. Propagation of visible IR radiation in a scattering medium such as tissue can be described by diffusion of photons, and thus we describe it as a diffusive wave.

Referring to FIG. 6, the two laser diodes were separated at a distance of 4 cm and 1.2 cm from the detection port. The intensity modulated light of the two laser diodes at frequency 200 MHz was sent through two optical fibers to a container with an Intralipid.TM. suspension. The source detector distance was varied by moving the optical port of the detection fiber along a line parallel to the position of the sources. FIGS. 6A, 6B, and 6C show measured maxima and minima of the optical field migrating in the medium. This data demonstrates interference between two diffusive waves created by two coherent emitting sources of phase difference 180 degrees. FIG. 7 summarizes the experiment, wherein the displacement of the detector is plotted against the phase shift measured by the detector. The phase shift displays the steepest part of the trace, curve A, (slope of about 360.degree./cm) at the displacement of about 2.25 cm. Curve B is measured with an optical field of source S.sub.2. Here, the measured slope is about 30.degree./cm. When comparing curves A and B we demonstrate much higher sensitivity of the null detection of the two element array contrasted with a diminished sensitivity to the detector displacement when using a single source arrangement. The sensitivity of the two source arrangement is increased by about a factor of 10. The sensitivity is further increased when using four or more element phased array, which sharpens the photon density gradient and thus provides a higher resolution for locating the hidden object.

In a strongly scattering medium, the emitted photons undergo a large number of collisions and their migration can be determined by applying the diffusion equation. The diffusion equation for photons in a uniformly scattering medium was solved by E. Gratton et al., "The possibility of a near infrared optical imaging system using frequency domain methods." in Mind Brian Imaging Program, Japan 1990; and by J. Fishkin et al., "Diffusion of intensity modulated near-infrared light in turbid media", SPIE Vol. 1413 (1991) p. 122. A solution of the diffusion equation was obtained for the light of a point source (at r=0) radiating S(1+M exp[-i(.omega.t+e)] photons, wherein S is the source strength (photons/sec.), M is the modulation of the source at frequency .omega., and e is an arbitrary phase. The photon intensity can be calculated as

wherein .rho.(r,t) is the photon density and c=10.sup.8 m/s is the velocity of light.

When solving the diffusion equation using a spherical-harmonics approximation in a non--absorbing medium for the density of photons .rho.(r,t) than

wherein the diffusion constant D is 1/3 of the mean free path. In the absence of an amplitude modulated signal (.omega.=0) the solution corresponds to a spherical wave propagating without attenuation. For a non-zero frequency, the amplitude of the signal at a frequency .omega. decreases exponentially. The light wave front the emitted advances at the constant velocity V

and has wavelength

The above equations show that higher modulation frequencies yield shorter effective wavelengths, and smaller diffusion constants also give shorter effective wavelengths. In principle, short wavelengths can be obtained using high frequency modulated waves in a very turbid medium. However, the amplitude of the modulated wave decreases exponentially with the modulation frequency. Therefore, the best resolution, i.e., the shortest wavelength, is obtained using the highest frequency which still gives a measurable signal. The diffusion process limits the penetration depth at any given modulation frequency, because of the exponential decrease of the wave's amplitude, and also decreases the velocity of light propagation.

The above described diffusion wave approach treats amplitude modulated light waves in scattering media using the framework of wave optics. The photon intensity, calculated as superposition of different waves, constitutes a scalar field, propagating at a constant velocity. At any given modulation frequency, the wave optics phenomenology of scalar fields is valid. Therefore, in the frequency-domain, the measurement and analysis of light diffusing in tissues from several sources will undergo constructive and destructive interference. Furthermore, wave refraction occurs at a boundary between two different tissues. It causes a deviation of the direction of propagation of the wave front, and thus there is a change in the amplitude and phase shift of the propagation wave. The direction change is a function of the ratio of the effective index of refraction in the two tissues. In diffusional wave optics, on the other hand, the wave's amplitude is exponentially attenuated as the wave propagates in the scattering medium. This attenuation is in addition to the exponential attenuation caused by finite absorption of the medium.

Amplitude modulated waves propagate coherently in the scattering medium; this is crucial for image reconstruction. It is possible to accurately measure in real time, the average intensity, amplitude, and phase of the wave front over a large area using a single detector or an array of detectors applying well-established frequency-domain methods.

The emitters are varied sequentially in phase starting with the first emitter in the line and followed by subsequent emitters. Each emitter emits a spherical wave and propagation of the resultant beam is perpendicular to the wavefront. If all the transmitter delays are equal, the beam travels straight ahead. Delay lines which produce variable transmitter delays can be used to obtain appropriate phasing for steering the beam across the tissue. The same principle can apply during reception.

There are two important aspects of imaging as envisioned by the present invention. The first is a geometrical aspect and the second is phasing of the transmitters and receivers.

It is also possible to construct a two-dimensional array for two-dimensional pointing (e.g., FIG. 2C). The multiplexing switches used with these arrays can be constructed as an integral part of the array and can consist of field effect transistors arranged so that access to any element may be obtained by the application of two adverse signals.

In addition to electronic scanning, the two-dimensional scanning can be achieved by moving the array of sources and detectors in a regular pre-determined pattern in a plane parallel to that being investigated in the subject. For maximum detection, the detector is places in the plane of the photon density gradient of the resulting field created by the array of sources. The plane of the photon density gradient is swept as the array moves. In this sweeping action, as a strongly or weakly absorbing object enters the radiation field, the detector registers a field imbalance due to the above described refraction of the propagating radiation. A two-dimensional image is formed by storing the information while the probe is moved across the subject. Several scans in different imaging planes are envisioned by the invention. If the system is duplicated or time shared in two other faces of a cube, an algorithm would be used to provide a 3-dimensional picture of the object by triangulation, as known in the art. The data storage is accomplished electronically.

The detector detects the intensity and the phase shift of the radiation that has migrated in the subject. The phase shift depends on the tissue properties, i.e., absorption and scattering. For the low frequencies the phase shift is proportional to ((1-g).mu..sub.s /.mu..sub.a).sup.1/2 and for the high frequencies proportional to 1/.mu..sub.a. To obtain desired penetration depth, appropriate frequency for both master oscillator 22 and local oscillator 26 is chosen; however, the phase relationship of the laser diodes is maintained.

Different types of phased arrays are designed for optimal examination and imaging of different human organs (e.g., human head or breast). The amplitude and phase of the signals can be monitored on a precision oscilloscope. In order to scan the phased array past a fixed object of approximately known position, as in needle localization procedures, the location of the input and output ports will be scanned past the object and the position of maximum phase shift will be recorded in one-dimension; however, detection in two and three dimension can be performed in the same way.

In the preferred mode of operation, the array of sources is phased 180.degree. apart, as shown in FIG. 8A. There is a sharp 180.degree. transition of photon density wave, a large gradient in photon density, from S.sub.2, S.sub.2 sources to the S.sub.3, S.sub.4 sources. Thus, the radiated field gives an amplitude null and a phase transition of 180.degree. corresponding to the y-z plane, i.e., perpendicular to the detector. If a larger number of similarly phased sources is used, the transitions are even sharper. The array produces a uniform photon density pattern on each side of the array, as shown in FIGS. 8B and 8C. If an absorbing object is placed in this directional field of diffusing optical waves, imbalance in the photon density is measured. The detection of a hidden object is accomplished by translating the experimental transmitter-receiver system of FIG. 8A.

In addition to the mechanical scanning achieved by moving of the input-output port system, electronic scanning can be performed using the multiple source and multiple detector system of FIG. 2. As shown in FIG. 2B for an array of five sources, there is a 180.degree. phase transition in the resulting migrating field due to the 180.degree. phase difference between the antiphased sources radiating amplitude modulated light. The plane of the 180.degree. phase transition can be shifted in parallel by appropriately varying the apportionment of 0.degree. and 180.degree. phases on the sources. This is performed by sequentially switching the phase of the sources by 180.degree.. In each case, the detection port located on this plane is used for collecting the data. As the sources are electronically switched by 180.degree., the detection array can be also electronically switched from one detection port to another. The signal from the receiving optical fiber is coupled to one shared PMT detector. However, the system can also include several detectors. If the systems of FIGS. 1 or 1A are used, the electronic source scanning can be combined with synchronous mechanical movement of the detection port.

In general, the invention utilizes the photon density gradient created in the migrating field since it increases the resolution of the detection. As known to one skilled in the art, the photon density gradient formed by interference effects of the individual wave can be created not only by appropriate phasing of the sources but also by other methods such as appropriately spacing the sources, creating an imbalance in the radiated intensity of the individual sources, and other.

FIG. 8A shows the arrangement of the input ports 11 to 17 and detection port 19 of FIG. 1. As described above, light of each laser diode 12 through 18 is intensity modulated at the 200 MHz frequency. Wavelength of the intensity modulated radiation is ##EQU1## wherein f is the modulation frequency of 200 MHz, .mu..sub.s is the scattering factor which is approximately 10 cm.sup.-1 in an Intralipid solution with refractive index n, and c is 3.times.10.sup.8 cm/s. Thus, the expected wavelength is about 7 cm. The input ports S.sub.1, S.sub.2, S.sub.3, and S.sub.4 are set 3.5 cm apart and are anti-phased by 180.degree. using a push pull transformer. The antiphased array creates a large gradient in photon density chosen to take advantage of the destructive interference with the null detection. The laser diodes emitting 754 nm light are intensity modulated at 200 MHz using master oscillator 22, and the local oscillator 26 is operating at 200.025 MHz to perform the dynode modulation of PMT detector 24. The detected intensities and phase shifts of an x-direction scan (FIG. 8A) of detection port 19 are plotted in FIGS. 8B and 8C, respectively. As expected, the intensity has a sharp minimum in between sources S.sub.2 and S.sub.3 where the phase is changed 180.degree.. The peak width at half maximum is about 2 cm. In addition to the x-direction scan of the detection port, the detection port was scanned in y-direction wherein, as expected, no variation was observed.

Referring to FIG. 9A, cylindrical objects of different diameter, d, were scanned using the previously described phased array. The objects were placed in the middle of the linear array displaced 2.5 cm from the x-axis. The detection port was located on the x-axis and each object was moved parallel to the x-axis at the 2.5 cm y displacement. The intensity and phase shift detected at different locations are plotted in FIGS. 9B and 9C, respectively. The intensity pattern for each moving object has two maximum and one minimum when the scanned object was located at x=0, y=2.5 point during its scan along the x-axis. At this point, a large phase change is detected, as shown in FIG. 9C. The phase detection has inherently larger resolution of a localized absorber; a hidden object of size as small as 0.8 mm can be detected.

The response due to different absorption of the hidden object was studied using a 5 mm cylinder of different absorption coefficient scanned by the 4 element phased array of FIG. 9A. The detected phase change is shown in FIG. 9D. The 5 mm black rod displays the largest phase change due to its high absorption, and the cylinder filled with cardiogreen 3.5 mg/l which has absorption coefficient .mu..sub.a =200 cm.sup.-1 shows the smallest phase change. In scanning of a hidden object, these experiments correspond to mechanically displacing the source detector system, or electronically scanning the subject.

Scanning of two objects of a different diameter is shown in FIG. 10A. Two cylinders of different diameter are scanned across the four element phased array located on the x-axis. The detection port in located at y=5 cm. In FIG. 10B the detected phase change in plotted against the displacement of these objects. Curve A represents the phase change of two cylinders of diameters 5 mm and 10 mm separated 3 cm apart. Curve B was measured using 16 mm cylinder instead the 5 mm cylinder. In this case, wherein the two cylinder separation is smaller, the phase detector can not resolve the two objects.

The imaging resolution is increased by increasing the number of elements of the phased array, since the main lobe of the resultant beam becomes much sharper, the gradient of photon density is larger. Phased arrays of different number of elements and different shapes are used for imaging different organs. For example, in tumor imaging, the four element phased array of FIG. 8A having an approximately linear shape can be used for imaging of the brain. On the other hand, a rectangular or a circular phased array would be used for imaging of a hidden tumor in the breast. The modulation frequency and the element spacing is adjusted to obtain proper focussing in each case.

In addition to the above described directional detection, the present invention envisions imaging systems able to calculate the average migration pathlengths. Referring to FIGS. 1 and 1A, in one mode of operation, the signal from master oscillator 22 is mixed with a set of four local oscillators operating at offset frequencies of 25, 35, 45, and 55 kHz (not shown in FIGS. 1 and 1A); there is one local oscillator operating at an offset frequency associated with each laser diode. Thus, the output of each laser diode is intensity modulated at the master oscillator frequency plus the frequency of its local oscillator. The intensity modulated radiation of each laser diode is simultaneously coupled to the tissue.

Detection of the optical field is performed in the same way as described for the other embodiments. The detected signal is heterodyne mixed directly at the PMT detector. The detector outputs signals at four different offset frequencies associated with each diode. These signals are fed into the phase detector wherein the phase and the intensity of the detected radiation are measured. There are either four phase detectors (only one detector is shown in FIG. 1) operating alternatively at different frequencies or one phased detector is used in a time shared mode of operation. The phase shift and the intensity of a detected heterodyned signal depend on the tissue through which said scattered and absorbed radiation migrated. When using several radiation sources of selected carrier frequency and phase, the resulting radiation has directional properties and the detected intensity and phase shift depend on the pathlength along which said radiation was scattered and absorbed. The tissue properties are determined from the detected phase shift and intensity values and from the known input ports and detection port geometries. The measured average pathlengths, <L>, can also be determined. The detected phase shift is converted to an effective migration pathlength <L> by using the low frequency approximation .theta.=2.pi.f<L>n/c, wherein f is the modulation frequency, c is the speed of light (3.times.10.sup.8 cm/s), and n is the refractive index of the medium.

To illustrate imaging by detecting migration pathlengths, we use an example of photon migration in a tissue with a strongly absorbing object, a perfect absorber (.mu..sub.a .fwdarw..infin.) of radius R. Referring to FIGS. 5A, 5B, and 5C the distribution of pathlengths defines an optical field that exists between a point detector, D, and source, S, separated by distance p and located on the exterior of an examined tissue which is a semi-infinite, strongly scattering medium. As shown in FIG. 5A, infinitely far away from the field, a perfect absorber does not alter the banana-shaped optical field of photons emitted by source S and detected at detector D. As the object enters the optical field (FIG. 5B), the photons which have migrated the farthest distance from D and S are eliminated by the absorption process inside the perfect absorber of radius R. Since photons which travel the longest pathlengths are absorbed, the approach of an object shortens the distribution of pathlengths, or alternatively, shortens the average pathlength <L>. As the object moves closer, and the optical field surrounds the object (FIG. 5C), some of the detected photons have travelled "around" the object, which is detected as lengthening the distribution of pathlengths. Thus, the average pathlength measurement can reveal location of a strongly absorbing component of a tissue (e.g., tumor or localized bleeding).

Even though this pathlength computation approach requires in most cases extensive computational capabilities, it can yield useful information in the localization procedures and can provide an useful supplement to the above described directional approach.