WO2009050712A2

WO2009050712A2 - Method, system and computer program product for tissue characterization

Info

Publication number: WO2009050712A2
Application number: PCT/IL2008/001374
Authority: WO
Inventors: Boris Rubinsky; Mohanad Shini; Shlomi Laufer
Original assignee: Yissum Research Development Company Of The Hebrew Univercity Of Jerusalem
Priority date: 2007-10-15
Filing date: 2008-10-22
Publication date: 2009-04-23
Also published as: WO2009050712A3

Abstract

A method for characterizing a tissue, the method includes: inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest; performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; taking a tissue biopsy at one or more locations of the multiple locations; and generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations.

Description

Method, System and Computer Program Product for Tissue Characterization

Related applications 1. This application claims the priority of US provisional patent application serial number 60/979,939 filing date 15 October 2007.

Field of the invention

2. This disclosure relates generally to a method, system and a computer program product for tissue characterization.

Background of the invention

3. In medicine, it is important to have information on the structure and state of internal tissues and organs under conditions in which the physician has neither visual nor tactile access to those tissues and organs. Two medical technologies produce this type of information. The first is medical imaging, which generates an image of the interior of a body by producing a map of physical attributes of the interior. Various physical attributes have been employed for medical imaging. For instance, magnetic resonance imaging produces a map of proton density, X-ray imaging produces a map of the tissue ability to absorb X-rays, and ultrasound imaging produces a map of the acoustic impedance mismatched interfaces inside the body.

4. Particularly relevant to this study is electrical impedance tomography (EIT). EIT produces an image of the spatial distribution of the electrical impedance of an object from electrical measurements made with an electrode array on its periphery (Boone, Barber et al. 1997), (Polydorides and Lionheart 2002), (Bayford 2006). 5. Image reconstruction in EIT is an inverse problem in which the electrical impedance of a domain is determined from the solution of the electrical field equation with the boundary conditions specified by electrode measurements. In a typical procedure, electric current is injected and removed through a pair of electrodes, while the resulting potentials are measured at other electrodes.

6. A variety of combinations of injecting and measuring electrodes from among the electrodes surrounding the targeted tissue or organ is used. An impedance distribution in the targeted tissue is sought that provides the solution of the electrical field equation which best satisfies all the boundary conditions obtained from the electrode measurements. Increasing the number of electrodes and the number of current injection pair combinations as well as decreasing the distance between electrodes can improve the quality of the EIT image (Tang, Wang et al. 2002). Further details on the technique can be found in a recent book on the topic (Holder 2004).

7. Medical imaging has the advantage that it produces a continuous map of the interior of the body and that it is non-invasive. On the other hand, most medical imaging technologies are expensive and often cumbersome. The resolution that medical imaging can produce is limited by constrains of physical placement of the sensors relative to the tissue or organs of interest as well as the natural limitations of the imaging technique. 8. Histological analysis of biopsies is another technique for producing information on tissues inside the body. Biopsy needles are inserted into the body, usually under imaging guidance, to remove tissue samples, which are then analyzed with a variety of hysto-chemical tests, e.g. (Bear 1998). The advantage of biopsies is that they generate precise information on the sampled tissue inside the body. 9. However, the drawback is that the tissues are sampled only in the discrete locations from which the biopsies are taken. Biopsies do not produce any information on adjacent tissues, except by extrapolation. Because of the limited information available from single biopsy needles sampling, it is becoming increasingly popular to sample larger numbers of locations, e.g. (Lee, Shun et al. 2005), (Rosenblatt, Fineberg et al.), (Jhavar, Corbishley et al. 2005), (Barzell and Whitmore 2003), (Crawford 2005), (Onik and Barzell 2007).

10. The value of multiple needle biopsies is best illustrate through its' use in the detection of prostate cancer (Onik and Barzell 2007). Taking as many as even up to 100 cores as compared to 6-12 in a conventional TRUS (transrectal ultrasound) technique has caught tumors in over 50 percent of the patients that had been inaccurately labeled negative by the more traditional and low number biopsy method (Onik and Barzell 2007). The technique employs cores removed at 5 mm intervals (Crawford 2005), under ultrasound guidance and is done through a grid placed over the peritoneum. This produces precise information also on the place from which the biopsy core has been taken, and has found use in the precise treatment of the prostate known as "male lumpectomy (Onik 2004). It is evident that biopsies produce precise information from the sampled tissue and that increasing the number of biopsies taken will increase the probability that all the malignant sites will be detected. 11. Nevertheless, the use of biopsies remains a discrete site sampling technique and regardless of the number of sites, malignancies can remain hidden between the sites. Furthermore, increasing the number of biopsies increases cost and the complexity of the procedure.

Summary

12. The present invention provides methods and systems as described in the accompanying claims. Specific embodiments of the invention are set forth in the dependent claims. These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. 13. A method for characterizing a tissue, the method includes: (a) inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest; (b) performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; (c) taking a tissue biopsy at one or more locations of the multiple locations; and (d) generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations.

14. A system for characterizing a tissue, the system includes: (a) a memory unit that is configured to store tissue information indicative of a^" tissue region of interest, electrical measurement information and location information; wherein the location information and the electrical measurement information are obtained by inserting, in a sequential manner, a biopsy probe to the tissue region of interest at multiple locations while monitoring the insertion to provide location information; wherein the locations are determined in view of a location of the tissue region of interest; and performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; wherein a tissue biopsy is taken at one or more locations; and (b) an impedance information generator that is configured to generate impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations. 15. A method for characterizing a tissue, the method includes: (a) obtaining electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and (b) classifying the tissue by applying a classifier on the location information and the electrical measurement information.

16. A system for characterizing a tissue, the system includes: (a) an electrical measurement unit configured to measure electrical information of a tissue region of interest, wherein the electrical information is obtained from multiple locations; wherein

• t the tissue region of interest is determined in response to tissue information; (b) an imager configured to generate location information indicative of the multiple locations; (c) a memory unit that is configured to store the tissue information that is indicative of a tissue region of interest, the electrical measurement information and location information; and (d) a classifier that processes the location information and the electrical measurement information to provide tissue class information.

17. A computer program product that includes a computer program medium that stores instructions for: (a) obtaining electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and (b) classifying the tissue by applying a classifier on the location information and the electrical measurement information.

18. A computer program product that includes a computer program medium that stores instructions for: (a) receiving electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and (b) generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations; wherein the electrical information is obtained by inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest and performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; wherein a tissue biopsy is taken at one or more locations of the multiple locations.

Brief Description of the Drawings

19. Further details, aspects, and embodiments of the invention will be described, by way of example only, with reference to the drawings. 20. Figure 1 a illustrates a COMSOL generated mesh based on electrode location;

21. Figure 1a illustrates a COMSOL generated mesh based on electrode location and tumor location;

22. Figures 2a, 2c and 2e illustrate simulations of meshes based on electrode locations and tumor of different sizes and figures 2b, 2d and 2f illustrate EIT images reconstructed in response to these different meshes;

23. Figures 3a, 3c and 3e illustrate simulations of meshes based on electrode locations and a pair of tumors of different size and figures 3b, 3d and 3f illustrate EIT images reconstructed in response to these different meshes; 24. Figures 4a, 4c and 4e illustrate simulations of meshes that differ from each other by the number of electrode locations while the tumor is of the same size and figures 4b, 4d and 4f illustrate EIT images reconstructed in response to these different meshes;

25. Figure 5a illustrates a simulation of a three dimensional mesh based on electrode locations and a tumor and figures 5b and 5c illustrate EIT images reconstructed in response to these different meshes;

26. Figure 6a illustrates a simulation of a three dimensional mesh based on electrode locations and a tumor and figures 6b and 6c illustrate EIT images reconstructed in response to these different meshes;

27. Figure 7a illustrates a simulation of a three dimensional mesh based on electrode locations and a tumor and figures 7b and 7c illustrate EIT images reconstructed in response to these different meshes;

28. Figure 8 illustrates a system according to a first embodiment of the invention;

29. Figure 9 illustrates biopsy probes according to various embodiments of the invention; 30. Figure 10 illustrates a method according to a first embodiment of the invention;

31. Figure 11 illustrates a system according to a second embodiment of the invention; 32. Figure 12 illustrates a method according to a second embodiment of the invention;

33. Figure 13 illustrates a system according to a third embodiment of the invention;

34. Figure 14 illustrates a method according to a third embodiment of the invention; 35. Figure 15 illustrates a three dimensional model used during an evaluation of the second embodiment;

36. Figure 16 illustrates typical data used with the classifier according to an embodiment of the invention; and

37. Figure 17 illustrates results of last classifier in table 1. w - Square cube width, according to an embodiment of the invention.

Detailed Description of the drawings

38. Because the apparatus implementing the present invention is, for the most part, composed of electronic components and circuits known to those skilled in the art, circuit details will not be explained in any greater extent than that considered necessary as illustrated above, for the understanding and appreciation of the underlying concepts of the present invention and in order not to obfuscate or distract from the teachings of the present invention.

39. In the following specification, the invention will be described with reference to specific examples of embodiments of the invention. It will, however, be evident that various modifications and changes may be made therein without departing from the broader spirit and scope of the invention as set forth in the appended claims.

40. It has been shown that biopsy probes, in addition to their function to sample tissues from the interior of the body can be also used as sensors or sensor carriers in medical imaging. This idea is particular relevant to situations which multiple biopsy samples are taken and therefore, sensor measurements from various sampling sites can be combined to produce a refined image of the sampled area. Electrical impedance tomography (EIT) is amenable to this concept, because the technique requires only electrodes to inject currents and measure voltages and it would be technically simple to add electrodes to biopsy probes.

41. Accordingly, biopsy probes can be used, in addition ,to their function to remove tissue samples, also as the EIT sensors (electrodes) that inject currents and measure voltage. Information from multiple samplings can be used to reconstruct an EIT image, in the area sampled by the biopsy probes. This method can be used to detect tumors of various sizes. The EIT image can include information about the impedance of tissue portions that are not in close vicinity of the locations- and can not be estimated using only the measurements obtained from a single biopsy probe inserted in a single location.

42. According to a first embodiment of the invention an EIT image of a region of interest of a tissue can be generated by executing the following stages: (i) Producing or receiving an image of the tissue with one imaging modality such as ultrasound, MRI, CT, optical imaging, mammogram, x-ray or electrical impedance tomography, (ii) Determining, in response to the image of the tissue a tissue region of interest from which to take multiple tissue biopsies, (iii) Placing a biopsy needle (biopsy probe) that includes at least one electrode that can be used to inject current and measure voltage at a location under medical imaging (such as ultrasound, mammogram, x-ray, MRI, CT, and the like- also referred to as imager or monitor) so as to detect the precise location of multiple electrodes that are utilized during an electrical measurement and providing location information; (iv) Making electrical measurements of currents and/or voltage between one biopsy probe electrode and another electrode, wherein the other electrode can be on the biopsy needle, on another tissue biopsy needle with electrode or any other reference electrode, the outcome of the electrical measurements is electrical measurement information. This information can be' indicative of measured currents and, additionally or alternatively, measured voltages, (v) Taking a tissue biopsy from the location, (vi) Repeating at least stages (iii) - (iv) multiple times at a different locations whereas a biopsy can be taken from one or more locations; (vi) combining the location information and the electrical measurement information and providing the information to a system that produces an electrical impedance tomography image (the system can includes a processor that executes a software such as but not limited to such as the EIDORS software) to produce an electrical impedance tomography image of the tissue in the tissue region of interest that was probed by the biopsy needles.

43. Conveniently, a system is provided it can include a medical imaging system (also referred to as imager or monitor) for imaging the tissue, at least one biopsy probe with one or more electrodes, an electrical measurement unit for measuring electromagnetic data from the biopsy probe electrode and another electrode, a system that combines the data from multiple needle biopsies and produces an image of the tissue (or of the tissue region of interest) from location information with electrical measurement information.

44. According to a second embodiment of the invention a classifier can be utilized for analyzing tissue from electromagnetic measurements in an area (tissue region of interest) in which multiple needle tissue biopsies are taken. This method can include: (i) Producing an image of the tissue with an imager such as ultrasound, IVIRI, CT, optical imaging, mammogram, x-ray, electrical impedance tomography, (ii) Determining, from the image of the tissue, locations (locations) from which tissue biopsies should be taken. There can be two locations, three locations and even more, (iii) Providing the tissue sampling biopsy needle with at least one electrode that can be used to inject current and measure voltage (system), (iv) Placing a biopsy needle (biopsy probe) that includes at least one electrode that can be used to inject current and measure voltage at a location under medical imaging (such as ultrasound, MRI, CT, and the like - also referred to as imager or monitor) so as to detect the precise location of multiple electrodes that are utilized during an electrical measurement and providing location information; (v) Making electrical measurements of currents/voltage between the biopsy needle electrode and another electrode in contact with the body to provide electrical measurement information and note from the imaging the exact location of the two electrodes - the precise location is represented by location information. The other electrode can be on the biopsy needle, on another tissue biopsy needle with electrode or any other reference electrode, (vi) Taking a tissue biopsy from the probed site, (vii) repeating at least stages (iii) to (vi) at least one (two, or more) more times at a different locations; (viii) combining the location information and the electrical measurement information and apply a classifier such as but not limited to a Support Vector Machine (SVM) classifier and a data base for these type of measurements to determine the nature (the type) of the tissue in the tissue region of interest.

45. A system can be provided. It can include a combination of a medical imaging system (also referred to as imager or monitor) for imaging the tissue, a biopsy needle (biopsy probe) with one or more electrodes, electrical measurement unit that can obtain data from the biopsy probe and another electrode, and a system that can execute a computer software that is responsive to the location information and electrical measurement information and employs a classifier and a data base for the configurations tested to determine the nature of the tissue in the area probed by the biopsy needles. 46. According to a third embodiment of the invention the classifier can be applied on electrical measurement information that is obtained in a non-invasive manner or at least a portion thereof is obtained in a non-invasive manner. The electrical measurement information can reflect measurements taken in multiple discrete frequencies. The discrete frequencies can be included in the β dispersion range. The classifier can be an SVM classifier and especially a two-class SVM classifier. Accordingly, conventional imaging technique is used to identify a tissue region of interest, and then electrical measurement information is obtained from that tissue region of interest, this information (as well as location information associated with the measurements) is fed to a classifier that determines the nature (the type) of the tissue region of interest.

47. This method can include: (i) Producing an image of the tissue with an imager such as ultrasound, MRI, CT, optical imaging, electrical impedance tomography, (ii) Determining, from the image of the tissue, measurement locations from which measurements should be made. There should be two sites, three sites and even more. (iii) Making electrical measurements of currents/voltage while monitoring the precise location of the components that are involves in the measurements, wherein the precise location is represented by location information, (iv) repeating at least stage (iii) at least one (two, or more) more times at a different sites; (v) combining the location information and the electrical measurement information and apply a classifier and a data base for these type of measurements to determine the nature (the type) of the tissue in the tissue region of interest.

48. A system can be provided, it can include a medical imaging system (imager or monitor) for imaging the tissue and identifying a tissue region of interest, an electrical measurement unit, a system that can execute a computer software that applies a classifier on the location information and electrical measurement information.

A first embodiment

49. Figure 8 illustrates system 800 for characterizing a tissue, according to an embodiment of the invention. 50. System 800 includes memory unit 810, impedance information generator 820. It can also include imager 830, electrical measurement unit 840, one or more biopsy probes such as biopsy probe 850.

51. Memory unit 810 stores tissue information indicative of a tissue region of interest, electrical measurement information and location information. Tissue information can be an image of the tissue. It can indicate whether the tissue includes a suspected tissue region of interest that can be a tumor.

52. The location information is provided by imager 830 and the electrical measurement information is obtained by electrical measurement unit 840 and biopsy probe 850. The mentioned above information is obtained by inserting, in a sequential manner, biopsy probe 850 to the tissue region of interest at multiple locations while monitoring the insertion to provide location information and performing electrical measurements by electrical measurement unit 840 that is connected to biopsy probe 850, at each of the multiple locations. Biopsy probe 850 also takes a tissue biopsy from at least one location.

53. The measurement can include providing an excitation signal at one electrode and measuring a voltage potential between two electrodes. Both electrodes can be placed on biopsy probe 850, but this is not necessarily so. While one electrode belongs to biopsy probe 850 another electrode can belong to a reference probe or to another component that contacts the tissue or the tissue region of interest. The other electrode can be inserted into the body of a patient but this is not necessarily so

54. A sample of a multiple electrode probe is illustrated in US patent application 2002/0026123 of Pearlman, which is incorporated herein by reference. A further example of a probe that is equipped with an electrode in provided in PCT patent application WO 2007/083310 of Hashimshony et el., which is incorporated herein by reference. Various methods for tissue analysis with electromagnetic waves based on a single probe can be applied, some of which are illustrated in the following patents all being incorporated herein by reference: US Patent 5746210, US Patent 5752519, US Patent 598736, and US Patent 6594518.

55. One or more measurements can be made at one or more locations. For example, at one or more site different measurements can be taken at different depths. For example- the measurements can be taken at different penetration depths of biopsy probe 850. These measurements can be taken at depths that are evenly spaced from each other - but this is not necessarily so. Accordingly, each site can include multiple spaced apart measurement locations. According to another embodiment of the invention biopsy probe 850 includes a plurality of biopsy electrodes that are positioned in different heights. By using different electrodes measurements from different measurement locations can be taken without altering the penetration depth of biopsy probe. It is further noted that when multiple probes are used - biopsy probe 850 can be maintained at the same position while a reference probe is lowered or raised. A site can be a biopsy site in which a biopsy needle is inserted but this is not necessarily so. A site can be located entirely outside of the body of a patient but can be located inside the body of a patient. Conveniently, a site is associated with a single inclusion of a measurement element (such as a biopsy probe electrode) into the body of the patient.

56. Conveniently, different excitation signals can be provided per each location. For example, different excitation signals of different frequencies can be provided (in a sequential manner) at each location. For example - twelve excitation signals of twelve different frequencies (within the β excitation range) can be provided.

57. It is noted that a combination of both mentioned above measurement techniques can be provided. Thus, measurements are taken at different depths and involve injecting excitation signals that differ from each other. 58. It is noted that the measurement locations can define a two-dimensional array or measurement locations or even a three dimensional array of measurement locations. The array can be an evenly spaced array (such as an evenly spaced grid) but this is not necessarily so.

59. The locations are determined in view of a location of the tissue region of interest. The determination can be made, in view of the tissue information, by a person.

Additionally or alternatively, the determination can be made by location determination unit 860. The locations are positioned so as to enable sampling the tissue region of interest.

60. Impedance information generator 820 is configured to generate impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations. It can generate a TIM image of the tissue region of interest.

61. Imager 830 is configured to monitor the insertion of the biopsy probe at the multiple locations to provide location information. It can also provide the tissue information, although this is not necessarily so and a different system (not shown) can provide the tissue information. Imager 830 can obtain an image of the tissue in a noninvasive manner, before inserting, in the sequential manner, the biopsy probe. It can include one or more sensors such as an ultrasound sensor, a magnetic sensor, an X- ray sensor, an optical sensor, an electrical sensor, a radio frequency sensor, and a thermal sensor or a combination thereof.

62. Conveniently, biopsy probe 850 and especially biopsy probe electrode 852 has an enlarged surface area. The surface area can be enlarged in various manner such as but not limited to sanding (or otherwise roughening) the surface, coating the surface with Teflon, and the like. The enlarged surface area is expected to improve the interface between the biopsy probe and the tissue and reduce noise associated with the electrical measurements.

63. Biopsy probe 850 can be inserted by a person but this is not necessarily so. For example, the insertion process can be performed by a location insertion controller 880 that is configured to insert the biopsy probe, in a sequential* manner, at multiple evenly spaced probe sites. These locations can be evenly spaced fro each other. Conveniently, location insertion controller 880 inserts the biopsy needle at three or more locations. Location insertion controller 880 can include a robotic arm that can insert the biopsy probe in a precise manner.

64. Figure 9 illustrates biopsy probes 850, 850' and 850" according to various embodiments of the invention. Biopsy probe 850 includes tip 858, biopsy taking element 856, and biopsy probe electrode 852. Biopsy probe electrode 852 can be used to inject an excitation signal and the voltage potential between biopsy probe electrode 852 and another electrode (not shown) can be measured to provide electrical measurement information.

65. Biopsy probe 850' includes tip 858, biopsy taking element 856, and multiple biopsy probe electrodes 852 and 854. Figure 9 illustrates these biopsy probe electrodes as being located in different locations along an imaginary longitudinal axis (not shown) of biopsy probe 850' so that when biopsy probe 850' is inserted into a location these are positioned at different depths.

Biopsy probe electrode 852 can be used to inject an excitation signal and the voltage potential between biopsy probe electrodes 852 and 854 can be measured to provide electrical measurement information. According to an embodiment of the invention the excitation signal can be provided via biopsy probe electrode 854.

66. Biopsy probe 850' includes tip 858, biopsy taking element 856, and an array 853 of biopsy probe electrodes that includes biopsy probe electrode 852 and other biopsy probe electrodes 852(1) - 852(K). Figure 9 illustrates a rectangular array but other arrays can be provided. Biopsy probe electrode 852 can be used to inject an excitation signal and the voltage potential between biopsy probe electrodes 852 and at least one other electrode out of biopsy probe electrodes 852(1) - 852(K) can be measured to provide electrical measurement information. According to an embodiment of the invention the excitation signal can be provided via another biopsy probe electrode of array 853.

67. It is noted that a biopsy probe can include a row of three or more biopsy probe electrodes that are isolated from each other.

68. Figure 10 illustrates method 1000 for characterizing a tissue according to an embodiment of the invention.

69. Method 1000 starts by stage 1010 of receiving or obtaining tissue information. The tissue information can be an image of the tissue.

70. Stage 1010 can include stage 1012 of obtaining the image of the tissue before inserting, in the sequential manner, the biopsy probe. Stage 1012 can include obtaining the image of the tissue in a non-invasive manner, before inserting, in the sequential manner, the biopsy probe. The image can be obtained by a sensor such as an ultrasound sensor, a magnetic sensor, an X-ray sensor, an optical sensor, an electrical sensor, a radio frequency sensor, a mammogram, and a thermal sensor.

71. Stage 1010 is followed by stage 1020 of determining, in view of the tissue information a tissue region of interest and locations (at least two) from which a tissue biopsy should be taken.

72. Stage 1020 is followed by stage 1040 that includes: (a) inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information, (b) performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information and (c) taking a tissue biopsy at one or more of the multiple locations. 73. Stage 1040 can include either one of: (i) inserting the biopsy probe, in a sequential manner, in multiple evenly spaced probe sites, (ii) inserting the biopsy probe, in a sequential manner, in multiple evenly spaced probe sites that form a rectangular grid, or (iii) inserting the biopsy probe, in a sequential manner, in at least three locations. • 74. Stage 1040 can include at least one of the following stages or a combination thereof: (i) stage 1041 of performing at a plurality of locations, multiple electrical measurements taken at different depths, (ii) stage 1042 of performing at a plurality of locations, multiple electrical measurements taken at multiple discrete frequencies, (iii) stage 1043 of performing at a plurality of locations, multiple electrical measurements taken at different depths and performing at a plurality of locations, multiple electrical measurements taken at multiple discrete frequencies, (iv) stage 1044 of performing an electrical measurement utilizing the biopsy probe and a reference probe, (v) stage 1045 of performing an electrical measurement utilizing multiple biopsy probe electrodes, (vi) stage 1046 of performing electrical measurements by at least one biopsy probe that has an enlarged surface area.

75. Stage 1040 is followed by stage 1050 of generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations. 76. Stage 1050 can include stage 1052 of generating an electrical impedance tomography image in response to electrical measurement information and location information obtained from the multiple locations. A second embodiment

77. Figure 11 illustrates system 1 100 for characterizing a tissue, according to an embodiment of the invention.

78. System 1100 includes memory unit 810, SVM classifier 1120. It can also include imager 830, electrical measurement unit 840, and one or more biopsy probes such as biopsy probe 850.

79. System 1100 differs from system 800 of figure 8 by including SVM classifier 1120 instead of impedance information generator 820. According to an embodiment of the invention a system can include both SVM classifier 1120 and impedance information generator 820.

80. SVM classifier 1120 can operate as a two-class (or few class) classifier and not operate in a regressive manner. It processes the location information and the electrical measurement information to provide tissue class information.

81. During a learning phase of SVM classifier 1120 it is fed with electrical measurements and an indication (based upon an analysis of tissue biopsies) that indicate whether the sampled tissue was a tumor and even more especially whether the sampled tissue was malignant or not. At the end of the learning phase the SVM classifier 1120 builds SVM data structure 1122 that will be utilized during its implementation phase. System 1100 performs the implementation phase but can also perform the learning phase.

82. Conveniently, during the learning phase, SVM classifier is fed with measurements taken at different frequencies. It can be fed by measurements taken when the excitation signal had a discrete frequency in the β dispersion range.

83. SVM classifier 1100 can be trained to differentiate between tissues that are a benign tumor and a malignant tumor. According to an embodiment of the invention different SVM classifiers can be trained to classify tissues of different types. For example- one SVM classifier (1120) can be trained to locate micro-calcifications and another support vector machine classifier (not shown) can be trained to locate suspicious masses. Yet for another example, the patient body can be virtually segmented to segments and each segment can be associated with a dedicated SVM classifier.

84. According to various embodiments of the invention SVM classifier 1120 can receive multiple electrical measurements taken at multiple depths, can receive (and be responsive to) a size estimate of a tumor from which multiple biopsies were taken, can receive (and be responsive to) electrical measurements obtained in a non-invasive manner.

85. Figure 12 illustrates method 1200 for characterizing a tissue according to an embodiment of the invention.

86. Method 1200 includes stages 1010, 1020, 1040 and 1260. It differs from method 1000 of figure 10 by including stage 1260 instead stage 1050. Stage 1260 is preceded by stage 1040.

87. Stage 1260 includes classifying the tissue by applying a support vector machine classifier on the location information and the electrical measurement information.

88. Stage 1260 can include at least one of the following stages or a combination thereof: (i) stage 1261 of classifying the tissue as being of a tissue class out of few (less than four) tissue classes; (ii) stage 1262 of classifying the tissue as being a benign tumor and a malignant tumor; (iii) stage 1263 of classifying in response to a size estimate of a tumor from which multiple biopsies were taken; (iv) stage 1264 of classifying in response to electrical measurements obtained in a non-invasive manner; (v) stage 1265 of applying multiple support vector machine classifiers; (vi) stage 1266 of applying a first support vector machine classifier configured to classify micro- calcifications and applying another support vector machine classifier configured to classify suspicious masses.

A third embodiment

89. Figure 13 illustrates system 1300 for characterizing a tissue, according to an embodiment of the invention. 90. System 1300 includes memory unit 810, SVM classifier 1120. It can also include imager 830, electrical measurement unit 840, and one or more electrical measurement electrodes such as electrode 1310.

91. During electrical measurements one or more electrical measurement electrode out of electrodes 1310 can be connected (in a non-invasive manner) to the tissue or the body of a patient in a non-invasive manner. Additionally or alternatively, one or more electrode of electrodes 1310 can be inserted into the body of the patient (for example - by using biopsy probes that include an electrode). Electrodes 1310 can be arranged in an array. Various configurations of electrodes are illustrated in US patent application 2002/0026123 of Pearlman, which is incorporated herein by reference.

92. Memory unit 810 stores tissue information indicative of a tissue region of interest, electrical measurement information and location information. Tissue information can be an image of the tissue. It can indicate whether the tissue includes a suspected tissue region of interest that can be a tumor. 93. Electrical measurement unit 840 is configured to measure electrical information of a tissue region of interest, wherein the electrical information is obtained from multiple sites. The tissue region of interest is determined in response to an image of the tissue.

94. Imager 830 is configured to generate location information indicative of the location of the multiple sites. 95. SVM classifier 1120 can process the location information and the electrical measurement information to provide tissue class information.

96. Conveniently, electrical measurement unit 840 obtains at least a portion of the electrical measurement information from a biopsy probe that is inserted, in a sequential manner, to the tissue at multiple locations while the imager monitors the insertion to provide location information; wherein the locations are determined in view of a location of the tissue region of interest.

97. Figure 14 illustrates method 1400 for characterizing a tissue according to an embodiment of the invention. 98. Method 1400 starts by stage 1010 of receiving or obtaining tissue information. The tissue information can be an image of the tissue.

99. Stage 1010 is followed by stage 1420 of determining, in view of the tissue information a tissue region of interest and sites. At least one site can be located within a body of a patient biopsy and can be reached in an intrusive manner. Additionally or alternatively, at least one site can be reached in a non-intrusive manner.

100. Stage 1420 is followed by stage 1430 of obtaining electrical information of a tissue region of interest from multiple sites and generating location information indicative of the location of the multiple sites. Stage 1430 can include stage 1040 but this is not necessarily so.

101. Stage 1430 is followed by stage 1260 of classifying the tissue by applying a support vector machine classifier on the location information and the electrical measurement information.

102. A computer program product can be provided. It includes a computer readable medium such as a disk, diskette, DVD, CD, memory chip, smart card, and the like.

103. The computer program medium can store instructions for: obtaining electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and classifying the tissue by applying a classifier on the location information and the electrical measurement information.

104. The computer program medium can that store instructions for receiving electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations; wherein the electrical information is obtained by inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest and performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; wherein a tissue biopsy is taken at one or more locations of the multiple locations.

Evaluation of the first embodiment

105. Two and three-dimensional models were used to evaluate the feasibility of the first embodiment.

106. The two-dimensional model domain is a square cross-section of tissue (4-cm long sides) with one reference probe placed in an arbitrary chosen point on the boundary. In a typical procedure, biopsy samples are taken one at a time under conventional medical imaging, such as ultrasound. Therefore, in the simulated experiment the inventors assumed that the location of the biopsy probe a.k.a. electrode, is known. The inventors further assumed that in a typical application, currents are injected between the biopsy probe electrode and the reference electrode and the voltages at the two electrodes are measured. For simplicity, the biopsy probe electrodes were modeled as point electrodes. To simulate multiple biopsies the probe was inserted sequentially at various locations, regularly spaced.

107. A typical mesh and the location of the electrodes are shown in Figure 1. During each of the separate "insertions", injected currents and voltages on the probe and on the reference electrode were determined and they served as input data in the electrical impedance tomography image reconstruction process.

108. To provide information to the EIT image reconstruction algorithms, in this simulated experiment the currents and voltages on the electrodes were determined from the solution of the electrical field equation. v(cffφ) = 0 where σ is the complex conductivity and φ is the electric potential.

109. The conductivity of the domain was assumed to be known and reflects the configuration the inventors "image". Insulated boundaries were assumed on the outer surface of the domain and a predetermined current is injected into the biopsy probe electrode to the ground reference electrode. The solution of Equation (1) employs a mesh such as that shown in Figure 1 and the functions of EIDORS (Polydorides and Lionheart 2002). Particular to this analysis is the inventors' use of COMSOL Multyphisics for mesh generation. In COMSOL Multiphysics it is possible to pre-define a set of nodes as a subset of the general mesh and then produce the general mesh. In the simulations, the inventors used the placements of the electrodes and the "tumor" as a defined subset of nodes when the inventors built the mesh in COMSOL. Consequently, the mesh had one node for each electrode. Figure 1a shows a basic mesh that contains only electrodes and Figure 1 b shows a mesh with electrodes and the tumor. To solve the forward problem the inventors inserted a mesh such as in Figure 1 b, in the object of EIDORS to solve the forward problem. The output from each solution was the voltage on the biopsy probe electrode and reference electrode for a certain current and location of the biopsy probe relative to the reference electrode. 110. To simulate various experiments with tumors of different sizes, shapes, and location the inventors have modified the electrical conductivities of the domain to represent the tumors, using a COMSOL generated mesh (Figure 1 b). In each "experiment" the analysis was repeated to simulate the different discrete probe samplings for the different sites in which the probes were inserted. This data was then used as an input in an EIT reconstruction algorithm. Because of the assumed linearity of the problem (Somersalo, Cheney et al. 1992), all the known currents and voltages from the various insertions were combined into one input matrix for the solution of the inverse problem (Bakushinsky 1994). For noise analysis, the inventors added numerical noise to the voltage that results from the forward solution. If VQ is the forward solution, the input to the inverse problem including noise is: V_n = VQ(I + Av) , where v is a vector of normally distributed random numbers, with mean zero and standard deviation one (same dimensions as VQ vector), and A is the noise level (Edd and Rubinsky 2006). In the simulations A = 0.1 %. V_n is the input to the inverse solver. 111. Conventional EIT is the reconstruction of spatial images of the impedivity in a region bounded by electrodes. This is done through the inverse solution of equation (1) using boundary conditions that are data obtained from the application of a set of unique current projections from driving electrodes and collecting of voltage measurements at the non-driving electrodes (Brown 2003). While in conventional EIT the data which is used to find a solution to an inverse problem comes from boundary conditions on the outer surface of the analyzed domain, the data used in the solution of an inverse problem can also come from the interior (Katz and Rubinsky 1984), (Hsu, Rubinsky et al. 1986). Furthermore, solutions to inverse problems can be also found using over specified boundary conditions, such as when the current and voltage specified at the same location (Kececioglu and Rubinsky 1989), (Rubinsky and A. Shitzer 1976). In the simulations the reconstruction was implemented in MATLAB using the program EIDORS v.3.1 (Polydorides and Lionheart 2002). The inventors used the EIDORS by feeding it with information (from which the image was constructed) that includes the currents and voltages calculated on the interior nodes that correspond to the biopsy probe electrodes and on the reference electrode during the insertion experiments. To remove bias the mesh used in the solution of the inverse problem was imported from COMSOL and contained only the information on the location of the electrodes (as illustrated in Figure 1a). 112. The inventors have explored two different 3-D methods for using electrode on multiple biopsy sampling for EIT. The first is similar to the 2D scheme and employs a reference probe with electrodes. In this case the 3D model is a box with dimensions, (I, w, h) = (40, 40, 20) mm. To simulate one of the many possible ways to produce a 3D reconstruction the inventors have assumed that in this case the probe has four (node) electrodes spaced with a distance of 5 mm between them (Figure. 5). In figure 5 each "x" corresponds to a site. It was assumed that all the probes are inserted perpendicular to the 40mm by 40mm bottom surface in such a way that the first electrode is on the bottom surface (0mm) and the fourth electrode is on the tops surface (20 mm) . The inventors assumed that the probes are inserted at discrete and regularly spaced locations. The mesh which was also build with COMSOL on the basis of the location of the electrodes is shown in Figure 5.

113. As with the 2D case, here the inventors also used a reference probe that was placed arbitrarily on the outer surface of the box and had four electrodes with an arrangement similarly to that of the probe electrodes. While many combinations of current sources and sinks are possible, in this simulated experiments the inventors assumed that during each insertion of the biopsy probe the current source and sink are the electrodes on the biopsy probe and the reference probe which are at the same distance from the bottom surface. Therefore, during each insertion of the biopsy probe four different pairs of current sources and sinks are possible. The inventors further assumed that the walls of the box are insulated. The analysis was performed in a similar way to the 2D case. The solution of Equationi produced the voltage values at the four node electrodes on the biopsy probe and at the four electrode nodes on the reference probe for each source-sink current pairs. The reconstruction is also done with EIDORS as described in the 2D case.

114. The second 3D configuration was inspired by earlier studies on liner electrode arrays (Powell, barber et al. 1987). Particular to this configuration is that both electrodes - the sink and source for the electrical currents - are on the same probe, which removes the need for a reference probe. In this case, the inventors have employed a cubical box of 40 mm on each side.

115. It was assumed that the biopsy probes had eight electrodes separated by 5 mm. The electrodes are arranged in such a way that when inserted perpendicular to a surface the first and last probes correspond to the bottom and top surfaces of the box. It was assumed that all the probes were inserted at regularly spaced distances. The source and sink electrodes were on the same probe and the inventors studied two modes of the many possible (Powell, barber et al. 1987).

116. In the first mode, marked S1 the source and sink electrodes were adjacent to each other and in the second mode, marked S3, they were separated by two electrodes. In this model the current electrodes supplied information only of current and the remainder of the electrodes supplied information on voltage. For each insertion the inventors tested all the possible permutations of the defined spacing to obtain the data for the reconstruction. The reconstruction is also done with EIDORS, using a superposition of all the data from all the insertion experiments and all the current injection experiments, as in the previous case.

117. A variety of geometrical tumor configurations was tested to simulate possible experimental scenarios. The simulated experiments assumed that the conductivity ratio between the normal tissue and the tumor is one to five. This ratio is typical (Smith, Foster et al. 1986), (Miklavcic and Hart 2006). The inventors used a noise parameter A, of 0.1% as in (Edd and Rubinsky 2006). The results were chosen from a large number of tests for their ability to illustrate the concept and demonstrate its feasibility. In all the results, the reference electrodes were in the same location for a particular figure, since the complete studies have shown that the location of the reference electrode does not change the quality of the results in a significant way. 118. Figures 2, 3 and 4 are for the 2-D model.

119. Figure 2 examines the effect of tumor size on the ability of the technique to detect tumors. The left hand column of figures shows the simulated experiments and the right hand column of figures shows the reconstructed images produced with the EIDORS reconstruction algorithm. This case simulates a typical prostate multiple sampling biopsy procedure in which the distance between the probe insertion sites is 5 mm (Crawford 2005), (Onik and Barzell 2007). The reconstructed figures were obtained in the presence of noise. The top results were obtained for a simulated tumor of 1 cm, which would be also detected with the biopsies. However, the advantage of imaging with the biopsy needles is that the image also produces information on the actual size of the tumor, which could contribute to the design of the treatment.

120. The middle result is for a 5 mm tumor, which is placed, however, in such a way that it would be missed by the biopsy needles. The bottom result is for a 3 mm tumor, which is missed by the biopsy needles. It is evident from Figure 2 that the proposed technique has the potential to detect tumors between biopsy needles of a size that could be missed by the biopsies.

121. Figure 3 employs the same number of probe insertions and investigates a situation in which there are two tumors in the tissue. It was also assumed that the measurements are with noise. The right hand column of figures is the reconstructed images and the left hand column figures are the simulated experiments. The top panels are for two tumors of 1 cm each side by side. It is interesting to note that the imaging captures the discrete nature of the two tumors and separates between them. The middle panels show two 4 mm tumors which are both, missed by the biopsy needles. It is evident that with the proposed technique the tumors, which would be both missed by biopsies, are identified. This technique is therefore not limited to the detection of one tumor and could be used to single out a number of non-detectable tumors. The bottom panels are for two tumors, one of 1 cm and the other of 4 mm. Using traditional biopsy techniques, only the large sized tumor would be identified. The conclusion from the biopsy may have been that this is the only tumor and tnat it is responsible for the symptoms that have led to the need for performing a biopsy. It is obvious that with the proposed technique the hidden tumor is also detected.

122. Figure 4 illustrates another advantage of the proposed technique. Here the inventors examine the effect of the number of biopsy probe samples on the ability to detect the tumor. The left hand column of figures shows the simulated experiments and the right hand column of figures shows the reconstructed images. In these cases a 5 mm diameter tumor is simulated. The top figures are for a case in which 64 locations are sampled, as in the prostate, (Onik and Barzell 2007) and the tumor is missed by the sampling. It is evident that the technique can detect the tumor. However, the middle and bottom figures show that the presence of this tumor would be detected even with 36 or 16 biopsy sampling sites, respectively. With multiple sampling biopsy assisted EIT, decreasing the number of biopsies taken reduces the accuracy with which the tumor is depicted. Nevertheless, even reduced numbers of samplings provide the information on the presence of a non-detectable tumor. In the prostate, it is considered that 5 mm tumors are clinically significant and therefore the current 3-D mapping sampling biopsies employ 5 mm spacing between the biopsy probe insertion sites (Crawford 2005), (Onik and Barzell 2007). However, a substantial disadvantage of multiple biopsies is the cost and the time required for the procedure. If multiple biopsy sampling assisted EIT is proven to produce with fewer and sparse sampling similar information to that produced with convention more dense sampling biopsies this would have a significant impact on the economics of cancer detection.

123. Figures 5 and 6 are for the first of the two 3-D techniques. The experimental setup is shown by the top figure. The figure shows that the sampling was taken at sixteen equally spaced insertion sites separated by 10mm, as marked in the figure, for a total of 256 measurements. Figure 5b shows the reconstructed image in a case without simulated noise and Figure 5c in a case with simulated noise. The reconstructions are shown in figures 5b and 5c on nine panels in which each panel shows a cross section in planes perpendicular to the page along the wide side, separated by 2.5 mm from the bottom. The sequence of the panels is from left to right and top to bottom. It is evident that the technique can detect a hidden tumor, in 3-D also in the presence of noise.

124. Figure 6 was obtained in a similar way to Figure 5, albeit for a 6 mm tumor. It illustrates the limit of the technique with the current electrode configuration. The figure shows that with this configuration a 6 mm hidden tumor can be detected when there is no noise in the system. However, the noise distorts the images and produces artifacts which could be taken as false positives. False positives are obviously a potential problem with this imaging technique that needs to be addressed through optimization of electrode design and measurements.

125. Figure 7 was obtained using the linear array method in combination with multiple biopsy sampling to produce the EIT image of the sampled domain. The experimental system is shown in the top figure. Sixteen discrete biopsy samples were taken on sites equally spaced and separated by 10 mm. The reconstructions are shown in Figures 7b and 7c on seventeen panels in which each panel shows a cross section in planes perpendicular to the page, separated by 2.5 mm from the bottom of the cube-like domain. The sequence of the panels is from left to right and top to bottom. A central 4 mm tumor that was not detected by biopsies is examined. The data includes noise.

126. Figure 7b shows results obtained with the so-called S1 data acquisition scheme while Figure 7c is for the S3 data acquisition scheme. The S1 scheme employed 560 measurements for the reconstruction and the S3 scheme, 488.

127. Figure 7c shows that the technique can image with the same number of biopsy probe samplings as in Figure 6, smaller tumors than those which produced false positives in Figure 6. It is noted that the use of linear arrays in single measurements produced results with limited resolution and utility (Powell, barber et al. 1987). However, when used in a multi-biopsy sampling EIT technique, linear arrays seem promising. Nevertheless, the observation that the S3 measurement scheme produces, for the same number of insertions as S1 , false positives suggests the need for substantial research to optimize the technique.

Evaluation of the second embodiment 128. Breast cancer is the most common cancer among women in the Western world, and X-ray mammogram is the standard screening tool for breast cancer. However, while mammogams identify suspicious areas, the character of these areas is frequently inconclusive and a needle biopsy is often performed. There are several types of biopsies, but they are all invasive and expensive, and none are risk-free. Furthermore, in about 60-85% of the cases, the suspicious tissues are found to be benign. Although imaging techniques, such as sonography, have been shown to reduce the number of unnecessary biopsies the problem is yet to be solved. The classifier concept developed in this study could alleviate the need for needle biopsies for tissue characterization

129. Studies have shown that cancerous breast tissue and normal breast tissue have different impedance values , making it reasonable to assume that electrical impedance measurements can be used as an imaging and diagnostic tool. As an imaging system, several electrical impedance tomography (EIT) systems have been developed. The classical approach of EIT is to use the electrical measurements to reconstruct the impedance image of the tissue, from which the tumor location and properties can be determined. However, this method suffers from low resolution

130. The inventors suggested that the mammogram image be used to define the suspicious area and that the usage of spectroscopic electrical measurements be limited to distinguishing between benign and malignant tumors. This makes the problem one of classification, which can be solved using standard classification tools, such as Support Vector Machines (SVMs). Considering the large number of unnecessary biopsies performed, we believe that reducing this number through the use of this non-invasive, not harmful and inexpensive method is of practical value 131. The inventors suggest apply the SVM as a two-class classifier, rather than a regression method, and the employment of discrete multi-frequency electrical spectroscopy in order to distinguish between benign and malignant breast tumors. Accordingly, a first-order feasibility study is provided in which a mathematical simulation model is used to create the database. 132. The inventors assumed that the data which will be available for the construction of the classifier includes a mammography image, electrical measurements of currents, and voltages from electrodes and needle biopsy data.

133. It should be noted that according to different embodiments of the invention, the classifier may be fed with data that is obtained in an invasive manner, or in a non- invasive manner (e.g. from data that is only taken from the exterior, without biopsy data). According to an embodiment of the invention, the classifier is fed with data that is obtained from biopsy of a prostate. According to an embodiment of the invention, the classifier is fed with data that is acquired from the exterior of a breast. It is noted that in different embodiments of the invention, any combination of data that is acquired by biopsy and of data that is acquired from the exterior may be implemented.

134. According to an embodiment of the invention, that data is needed only in the construction of the classifier. In this theoretical study, the electrical measurements are replaced by the solution of the field equation obtained from a computer simulation of the problem. The biopsy data correspond to the tissue properties used in the mathematical simulation

135. The mathematical mode attempts to mimic the information that is available from the X-ray mammography and the biopsy. 136. It is assumed that the electrical measurements are performed in a configuration similar to the one suggested by C. Myoung Hwan, K. Tzu-Jen, D. Isaacson, G. J. A. S. G. J. Saulnier, and J. C. A. N. J. C. Newell, "A Reconstruction Algorithm for Breast Cancer Imaging With Electrical Impedance Tomography in Mammography Geometry," Biomedical Engineering, IEEE Transactions on, vol. 54, pp. 700-710, 2007, meaning that the breast tissue has the same geometry during the electrical measurements as during the mammography and that the measurement electrodes are on the mammography plates (see Rg. 14). This particular configuration was chosen since it is assumed knowledge of the suspicious tissue location from the mammogram image, and this method avoids the need to register the mammogram image. The breast between the mammography plates is modeled as a 3D box. The rounded edges of the breast are ignored assuming that they are "far enough" from the suspicious tissue. The 3D model of figure 15 includes five square cubes. The four pairs of electrodes used are marked in the four smaller images.

137. It is also assumed that the mammogram image provides only rough information of the suspicious tissue's characteristics and dimensions and that it is impossible to extract the exact size and shape of the suspicious tissue. In order to simulate this partial knowledge situation, it is assumed that the suspicious tissue has a cube-like geometry, as shown in the center of Fig 14, with randomly generated variations of the tissue composition. 138. Mathematical model of electrical measurements - Measurements made with electrodes are simulated through the solution of the field equation: f - (<7?u) = 0, xf Ω (1) in the geometry of Fig 14. It is assumed for boundary conditions that the entire outer surface of the cube is insulated except for two electrodes used for injecting the currents, I and -I. The "measurement" is the calculated voltage on the sites of two "measurement" electrodes. The voltage measurement electrodes can either be the same electrodes as the current injection electrodes or different ones. For each electrode couple the voltage is "measured" for "n" different current injection frequencies. 139. Data Preprocessing - Several different types of data preprocessing were

evaluated and found to be useful: ^" " ¹W^; (2)

(3)

141 . ¹WrJ = ^Vnjnπ,} ~ *V.*αr. (4)

142. Where* ^-4 ^{= 1^} -) are a set of measurements (n frequencies) for a specific electrode configuration. %umor are measurements done near the suspicious tumor.

Vreauisr are measurements done far enough from the tumor, or on the other breast.

This "regular" area can be verified using the mammography image. Finally

143. The first two preprocessing steps can be useful in two ways. First, they help in estimating the influence of the tumor on the measurements with respect to regular tissue, as can be seen in Figure 16. The second point is related to the fact that the electrical properties of normal tissue vary among different women.

144. Figure 16 illustrates typical data used with the classifier: raw data (a) normalized data as in eq. (2) (b) homogenized data as in eq. (3)(c). Using the normalized preprocessing, it can be seen that the effect of tumor size of 0.4cm is less than 1%, which may explain the difficulty in classifying small tumors in the presence of noise, w - Square cube width.

145. In using these data processing steps, these differences can be masked and emphasize the changes caused by the tumor. The third preprocessing step will result in shifting all the measurements around a mean value of zero. By using this shift, the classifier's generalization for different tumor sizes is improved. 146. SVM Classifier Training - An SVM classifier was used in order to distinguish between malignant and benign breast tumors. The data are presented as: f(^χi.yi I -_'(Xm-V₀₁ )}.^χ ε s^B .y 6 {-i, ij (5)

147. The vector ^χ: is a vector of length n, and each entry in the vector is a voltage value measured for a different electrical excitation frequency. M = 1 for a malignant tumor and K = -1 for a benign tumor. The index U = I-- -1 is one index for each "simulated woman's electrical spectroscopy study."

148. Measurement Noise - The robustness of the classifier to measurement noise was also checked. Noise was added in the following manner: ^⁰-* ⁼ ° l~ ^{+ v}) (6) 149. Where ^vz is a vector of voltage measurements, v is a vector of normally distributed random numbers, with mean zero and variance one (same dimensions as ^v- ), and A is the noise level.

150. Electrode Combination - The basic configuration used only one pair of electrodes. A more complex classifier can use multiple electrode configurations. These multiple measurements where used in two ways: (i) Majority: Train a separate classifier for each configuration, then give each classifier a binary score (1 for malignant and -1 for benign), and finally sum all the classifiers, (ii) Summation: Sum all the measurements and then train one classifier.

151. Results - Model Description - A 3D mesh (~ 40,000 elements) was created using Comsol Multiphysics Version 3.3. The solution of Eq. 1 was obtained using the forward solver of EIDORS. The size of the box was 10x10x4 (length x width x height). In total, 18 electrodes were used (3x3 on each side), of which four pairs were sufficient for this study. Five different sized square cubes were used to represent the suspicious tissue, with widths of 0.4; 0.8; 1.2; 1.6; and 2.0, respectively. In a study conducted by Jossinet , the electrical impedance values of breast tissue were measured for 12 dispersion range. The mean values from that study were beta frequencies in the used in our simulations. For the normal tissue, the properties of AT (adipose tissue: normal breast tissue) were used. The following pathological tissues were differentiated : CA (carcinoma: malignant tumors) versus MA (mastopathy; a general term covering various benign breast diseases) and FA (fibroadenoma: benign tumors of the breast). In order to simulate the variations in tissue properties and their uncertainty associated with imaging and biopsy (see section A.1), the inventors randomly chose 75%-90% of the cube and set its connectivity to one of the three pathological tissues. The rest of the box was left with AT.

152. Data Collection - All the data for training and testing the classifiers were obtained from the computer simulations. For each different pathological tissue type (3 types) and for each different suspicious tissue cube size (5 sizes), the inventors ran 100 different simulations (a total of 3 tissue type x 5 cube size x 100 random selections). Thus, a total of 1500 simulations were completed. In each simulation, the inventors randomly chose 75%-90% of the cube to be the pathological tissue and the rest to be adipose tissue. This means that each simulation had a unique impedance distribution. Each simulation included 16 different combinations of electrode current and voltage pairs and was simulated for the 12 different frequencies. In relation to real life classifiers, each model represented one woman for whom the mammography and biopsy were known and on whom electrical measurements were done.

153. Classifiers - Several classifiers were trained in order to examine the different parameters discussed in the methods section. In the first group of classifiers, the suspicious area size was ignored. The different classifiers in this group are presented in table 1 , and the SVM scores for one of them in Fig. 17. In the second group, the information on the suspicious area size was used, and five different classifiers were trained, one for each size. The results for several different noise levels are provided in table 2. 154. Since the main aim of this method is to reduce the number of unnecessary biopsies with a minimal number of subsequent malignancy misses, the inventors set the minimal sensitivity (the percent of malignant tumors classified as such) at 95%. The specificity (the number of benign tumors classified as such) can be seen as the percent of unnecessary biopsies avoided. Similar assumptions have been made in other studies . All the classifiers in this study were trained and tested using the program svmLight.

155. This feasibility study introduced the use of a classifier based on multi-frequency electrical spectroscopy measurements for breast cancer tissue characterization. The results of the study show that the usage of more than one electrode can in fact improve the classification of benign and malignant breast tumors. It also appears that knowledge of the estimated tumor size can improve the classifier's capabilities. This size estimation can be provided by the radiologist or by a CAD program. In clinical examinations, more information can be extracted, such as different classifiers for micro- calcifications and for suspicious masses. Further clinical work can also investigate the option of adding electrical data to existing CADx systems in order to improve their present classification capabilities and to facilitate the construction of one multi-modality classifier. In the presence of noise, the electrical measurements do not appear to be sufficient for the classification of small tumors (0.4 cm and 0.8cm for 2% noise and above).

156. Table 1 Results of classifiers with no suspicious area size. 'A' is the noise parameter taken from eq. (6). Electrode combination definition as in eq. (7) & (8). ± standard deviation. The training data subset included 70 of the 100 simulations for each combination of suspicious tissue type and cube size (i^Qsi∞itianoKi ' hum* :y»* ' hub* n:t ,a total of 1050 simulations), and the test data subset included the other 30 (^3βjr:«a!βti*«f ^* 'rt*«i« :*?« ' 5-_aj,, «« _>a total of 450 simulations).

157. Table 2 Results of classifiers with suspect area size. 'A' is the noise level from eq. (6). Sens.: sensitivity. Spec: specificity. All 16 electrode combinations were used as in eq. (8). The training data subset included 70 simulations for each tissue type

- tissue f>-?_* ,a total of 210 simulations), and the test data subset included the remaining 30 for each tissue type

^{• 3}.-:sr._β ry_?* ,a total of 90 simulations). 158. This study was conducted to evaluate the ability of a classifier using noninvasive electrical spectroscopy measurements to distinguish between malignant and benign tumors in a suspicious area identified on mammography. The results demonstrate the feasibility of this tissue characterization technique. Obviously, much more work is needed to optimize the design of the classifier for this use and to build a clinical database of information. However, the technique should have the potential to improve the characterization abilities of conventional imaging with relatively simple electrical measurements. Furthermore, it may provide a convenient alternative to needle biopsies.

159. Furthermore, those skilled in the art will recognize that boundaries between the functionality of the above described operations are merely illustrative. The functionality of multiple operations may be combined into a single operation, and/or the functionality of a single operation may be distributed in additional operations. Moreover, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be altered in various other embodiments. 160. Thus, it is to be understood that the architectures depicted herein are merely exemplary, and that in fact many other architectures can be implemented which achieve the same functionality. In an abstract, but still definite sense, any arrangement of components to achieve the same functionality is effectively "associated" such that the desired functionality is achieved. Hence, any two components herein combined to achieve a particular functionality can be seen as "associated with" each other such that the desired functionality is achieved, irrespective of architectures or intermediate components. Likewise, any two components so associated can also be viewed as being "operably connected," or "operably coupled," to each other to achieve the desired functionality.

161. In addition, the invention is not limited to physical systems or units implemented in non-programmable hardware but can also be applied in programmable systems or units able to perform the desired system functions by operating in accordance with suitable program code. Furthermore, the systems may be physically distributed over a number of apparatuses, while functionally operating as a single system.

162. However, other modifications, variations, and alternatives are also possible. The specifications and drawings are, accordingly, to be regarded in an illustrative rather than in a restrictive sense.

163. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word 'comprising' does not exclude the presence of other elements or steps from those listed in a claim. Moreover, the terms "front," "back," "top," "bottom," "over," "under" and the like in the description and in the claims, if any, are used for descriptive purposes and not necessarily for describing permanent relative positions. It is understood that the terms so used are interchangeable under appropriate circumstances such that the embodiments of the invention described herein are, for example, capable of operation in other orientations than those illustrated or otherwise described herein.

164. Furthermore, the terms "a" or "an," as used herein, are defined as one or more than one. Also, the use of introductory phrases such as "at least one" and "one or more" in the claims should not be construed to imply that the introduction of another claim element by the indefinite articles "a" or "an" limits any particular claim containing such introduced claim element to inventions containing only one such element, even when the same claim includes the introductory phrases "one or more" or "at least one" and indefinite articles such as "a" or "an." The same holds true for the use of definite articles. Unless stated otherwise, terms such as "first" and "second" are used to arbitrarily distinguish between the elements such terms describe. Thus, these terms are not necessarily intended to indicate temporal or other prioritization of such elements. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

WE CLAIM

1. A method for characterizing a tissue, the method comprises: inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest; performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; taking a tissue biopsy at one or more locations of the multiple locations; and generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations.

2. The method according to claim 1 comprising generating an electrical impedance tomography image in response to electrical measurement information and location information obtained from the multiple locations.

3. The method according to claim 1 comprising performing, at a plurality of sites, multiple electrical measurements taken at different depths.

4. The method according to claim 3 comprising performing, at a plurality of locations, multiple electrical measurements taken at multiple discrete frequencies.

5. The method according to claim 1 comprising performing, at one or more of a plurality of locations, multiple electrical measurements taken at multiple discrete frequencies.

6. The method according to claim 1 comprising performing an electrical measurement utilizing the biopsy probe and a reference probe.

7. The method according to claim 1 comprising performing an electrical measurement utilizing multiple biopsy probe electrodes.

8. The method according to claim 1 wherein the tissue information is an image of the tissue.

9. The method according to claim 8 comprising obtaining the image of the tissue before inserting, in the sequential manner, the biopsy probe.

10. The method according to claim 8 comprising obtaining the image of the tissue before inserting, in the sequential manner, the biopsy probe.

11. The method according to claim 8 comprising obtaining the image of the tissue in a non-invasive invasive manner, before inserting, in the sequential manner, the biopsy probe.

12. The method according to claim 8 comprising obtaining the image of the tissue by a sensor selected from a group consisting of an ultrasound sensor, a magnetic sensor, an X-ray sensor, an optical sensor, an electrical sensor, a radio frequency sensor, and a thermal sensor.

13. The method according to claim 1 comprising determining the locations in response to the image of the tissue.

14. The method according to claim 1 comprising monitoring the insertion by a noninvasive imager.

15. The method according to claim 1 comprising performing electrical measurements by at least one biopsy probe that has an enlarged surface area.

16. The method according to claim 1 comprising performing electrical measurements by at least one biopsy probe that has a rough surface.

17. The method according to claim 1 comprising performing electrical measurements by at least one biopsy probe that has a portion that is coated by Teflon.

18. The method according to claim 1 comprising inserting the biopsy probe, in a sequential manner, in multiple evenly spaced probe sites.

19. The method according to claim 1 comprising inserting the biopsy probe, in a sequential manner, in multiple evenly spaced probe sites that form a rectangular grid.

20. The method according to claim 1 comprising inserting the biopsy probe, in a sequential manner, in at least three locations.

21. The method according to claim 1 wherein the tissue information is indicative of suspicious areas of the tissue and wherein the locations correspond to the suspicious areas of the tissue.

22. The method according to claim 1 further comprising classifying the tissue by applying classifier on the location information and the electrical measurement information.

23. The method according to claim 22 comprising applying a support vector machine classifier.

24. A system for characterizing a tissue, the system comprises: a memory unit that is configured to store tissue information indicative of a tissue region of interest, electrical measurement information and location information; wherein the location information and the electrical measurement information are obtained by inserting, in a sequential manner, a biopsy probe to the tissue region of interest at multiple locations while monitoring the insertion to provide location information; wherein the locations are determined in view of a location of the tissue region of interest; and performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; wherein a tissue biopsy is taken at one or more locations; and an impedance information generator that is configured to generate impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations.

25. The system according to claim 24 comprising an imager that is configured to monitor the insertion of the biopsy probe at the multiple locations to provide location information.

26. The system according to claim 24 comprising an electrical measurement unit that is configured to perform, at a plurality of sites, multiple electrical measurements taken at different depths.

27. The system according to claim 24 comprising an imager that is configured to obtain the image of the tissue before inserting, in the sequential manner, the biopsy probe.

28. The system according to claim 24 comprising a location determination unit that is configured to determine the locations in response to a location of the tissue region of interest.

29. The system according to claim 24 comprising a location insertion controller configured to insert the biopsy probe, in a sequential manner, at multiple evenly spaced probe sites that form a rectangular grid.

30. The system according to claim 24 comprising a location insertion controller configured to insert the biopsy probe, in a sequential manner, at three or more locations.

31. A method for characterizing a tissue, the method comprises: obtaining electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and classifying the tissue by applying a classifier on the location information and the electrical measurement information.

32. The method according to claim 31 comprising obtaining at least a portion of the electrical information in an intrusive manner.

33. The method according to claim 31 comprising obtaining at least a portion of the electrical information in a non-intrusive manner.

34. The method according to claim 31 comprising obtaining at least a portion of the electrical information from a biopsy probe that is inserted, in a sequential manner, to the tissue at multiple locations while the imager monitors the insertion to provide location information; wherein the locations are determined in view of a location of the tissue region of interest.

35. The method according to claim 31 comprising applying a support vector machine classifier.

36. The method according to claim 31 comprising: inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest; performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; and taking a tissue biopsy at one or more of the multiple locations.

37. The method according to claim 31 comprising classifying the tissue as being of a tissue class out of few tissue classes.

38. The method according to claim 31 comprising classifying the tissue as being a benign tumor and a malignant tumor cyst, calcification.

39. The method according to claim 31 wherein the electrical measurement information comprises multiple electrical measurements taken at multiple discrete frequencies.

40. The method according to claim 31 wherein the electrical measurement information comprises multiple electrical measurements taken at multiple depths.

41. The method according to claim 31 comprising classifying in response to a size estimate of a tumor from which multiple biopsies were taken.

42. The method according to claim 31 comprising classifying in response to electrical measurements obtained in a non-invasive manner.

43. The method according to claim 31 comprising applying multiple support vector machine classifiers.

44. The method according to claim 31 comprising applying a first classifier configured to classify micro-calcifications and applying another support vector machine classifier configured to classify suspicious masses.

45. The method according to claim 31 comprising monitoring the insertion of the biopsy probe at the multiple locations to provide location information.

46. The method according to claim 31 comprising performing electrical measurements by utilizing a biopsy probe and a reference probe.

47. The method according to claim 31 comprising performing electrical measurements by utilizing at least one biopsy probe that comprises at least one biopsy probe electrode.

48. The method according to claim 31 comprising performing, at a plurality of locations, multiple electrical measurements.

49. The method according to claim 31 comprising performing electrical measurements by utilizing multiple biopsy probe electrodes.

50. The method according to claim 31 comprising obtaining an image of the tissue before inserting, in the sequential manner, the biopsy probe.

51. The method according to claim 50 comprising obtaining the image of the tissue in a non-invasive manner, before inserting, in the sequential manner, the biopsy probe.

52. The method according to claim 50 comprising obtaining the image of the tissue by a sensor selected from a group consisting of an ultrasound sensor, a magnetic sensor, an X-ray sensor, an optical sensor, an electrical sensor, a radio frequency sensor, and a thermal sensor.

53. The method according to claim 31 comprising determining the locations in response to a location of the region of interest.

54. The method according to claim 31 comprising performing electrical measurements by utilizing at least one biopsy probe that has an enlarged surface area.

55. The method according to claim 31 comprising performing electrical measurements by utilizing at least one biopsy probe that has a rough surface.

56. The method according to claim 31 comprising performing electrical measurements by utilizing at least one biopsy probe that has a portion that is coated by Teflon.

57. The method according to claim 31 comprising inserting the biopsy probe, in a sequential manner, at multiple evenly spaced probe sites.

58. The method according to claim 31 further comprising generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations.

59. A system for characterizing a tissue, the system comprises: an electrical measurement unit configured to measure electrical information of a tissue region of interest, wherein the electrical information is obtained from multiple locations; wherein the tissue region of interest is determined in response to tissue information; an imager configured to generate location information indicative of the multiple locations; a memory unit that is configured to store the tissue information that is indicative of a tissue region of interest, the electrical measurement information and location information; and a classifier that processes the location information and the electrical measurement information to provide tissue class information.

60. The system according to claim 59 wherein the electrical measurement unit is coupled to a biopsy probe and wherein at least a portion of the electrical measurement information is obtained in an intrusive manner.

61. The system according to claim 59 wherein the electrical measurement unit is configured to obtain at least a portion of the electrical measurement information in a non-intrusive manner.

62. The system according to claim 59 wherein the electrical measurement unit obtains at least a portion of the electrical measurement information from a biopsy probe that is inserted, in a sequential manner, to the tissue at multiple locations while the imager monitors the insertion to provide location information; wherein the locations are determined in view of a location of the tissue region of interest.

63. The system according to claim 59 wherein the electrical measurement unit is configured to generate electrical measurement information from a biopsy probe that is inserted , in a sequential manner, to the tissue at multiple locations while the imager monitoring the insertion to provide location information; wherein the biopsy probe takes a tissue biopsy from at least one location;

64. The system according to claim 59 wherein the classifier is a support vector machine.

65. The system according to claim 59 wherein the classifier is a support vector machine configured to classify the tissue as being of a tissue class out of few tissue classes.

66. The system according to claim 44 wherein the support vector machine classifier is configured to classify the tissue as being a benign tumor and a malignant tumor, a cyst or a calcified region.

67. The system according to claim 44 comprising multiple support vector machine classifiers.

68. The system according to claim 44 comprising at least one biopsy probe that has an enlarged surface area.

69. The system according to claim 44 comprising a location insertion controller configured to insert the biopsy probe, in a sequential manner, at multiple evenly spaced probe sites.

70. A computer program product that comprises a computer program medium that stores instructions for: obtaining electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and classifying the tissue by applying a classifier on the location information and the electrical measurement information.

71. A computer program product that comprises a computer program medium that stores instructions for: receiving electrical information of a tissue region of interest from multiple locations and generating location information indicative of the multiple locations; wherein the tissue region of interest is determined in response to tissue information; and generating impedance information reflecting a spatial distribution of electrical impedance of the tissue in response to electrical measurement information and location information obtained from the multiple locations; wherein the electrical information is obtained by inserting, in a sequential manner, a biopsy probe to the tissue at multiple locations of a tissue region of interest, while monitoring the insertion to provide location information; wherein the locations are determined in view of tissue information indicative of a tissue region of interest and performing electrical measurements at each of the multiple locations utilizing at least one biopsy probe electrode to provide electrical measurement information; wherein a tissue biopsy is taken at one or more locations of the multiple locations.