Ai-assisted intraoperative soft tissue imaging method for the assessment of success in removing cancerous tissue

Abstract

The present invention relates to an assessment device and tool for medical professionals to increase the success rates of the cancer removal surgeries of breast cancer patients. The removed breast tissue specimen is brought to an X-ray imaging device (200). A first image is taken from the whole specimen (203), and interesting areas (204a) are found either by machine-assisted tools or by human assessment on the actual image. Then a geometric magnification may be performed on the interesting area(s) (204a). The device may provide pixel-specific graphs on the intensities, and this gives help in finding microcalcifications in the specimen (203), comprising calcium oxalate (301) and calcium phosphate (302). The microcalcifications act as markers for various types of tumours. The closest distance from the found carcinoma (304) to the surface of the specimen (203) is also assessed as the healthy margin (303). The result is a human- & machine-made combined assessment, whether the surgery needs to be continued or not.

Description

[0001] AI-ASSISTED INTRAOPERATIVE SOFT TISSUE IMAGING METHOD FOR THE ASSESSMENT OF SUCCESS IN REMOVING CANCEROUS TISSUE

[0002] Technical field

[0003] The present invention relates to breast cancer tissue assessment during a tissue removal surgery of a breast cancer patient, and more specifically, to a method and system performing such as assessment.

[0004] Background

[0005] When a breast cancer is diagnosed within a patient and when a removal surgery i.e. a lumpectomy is scheduled for the patient for the treatment of this serious disease, this represents already a very stressing time for the patient. Furthermore, breast cancer is one of the most common form of cancers among female patients, especially in the Western world, with the rate of more than 5000 new cases annually in Finland. According to some UK statistical sources, 85 % of all breast cancer cases are treated by removal via a surgical operation, and the five- year survival rate is approximately 92 % of all diagnosed breast cancer patients. It can hence be said that the market need around the surgical operations of the breast cancer patients is a notable one, also concerning the quality of life aspects for a large group of people.

[0006] Diagnoses are usually made via screenings through mammography, which is a low-energy X-ray based screening method for the breast tissue. For instance, in Finland there is a regular screening program for females of various ages. Tissue biopsies may also reveal the presence of breast cancer.

[0007] The treatment methods involve surgery, radiation therapy, chemotherapy, hormonal therapy and targeted therapy. Among the surgeries, there is an option of a breast-conserving surgery (i.e. a lumpectomy) or a mastectomy based on the seventy and extent of the cancerous cells within the patient’s body.

[0008] When the patient is given a scheduled time for a surgical removal of at least a part of the breast, it is highly recommended and humanely also a best practice, that all the malignant tissues can be removed in a safe and sound process in a single surgical procedure, because any need for waiting for the results and a possible requirement to undergo a second surgery much later is highly stressful for the patient, and also wasteful for the hospital’s resources in the long-term.

[0009] Therefore, there is a need to perform a reliable scanning method for a removed piece of malignant tissue during the removal surgery of a breast of a patient so that there can be obtained reliable information whether all the malignant tissue has indeed been removed or not. If not, the same surgery operation can then continue for finishing the safe removal process of the whole malignant tissue.

[0010] US 11 ,257,213 (“Hendriks”) discloses tumour boundary reconstruction using hy- perspectral imaging. In other words, Hendriks takes a hyper- or multi-spectral image of a tissue sample comprising tumour or respective diseased tissue. They make a morphological analysis and a spectral analysis. The boundary of the tumour is thereby found out and also the reliability of the boundary is determined.

[0011] A problem in prior art is that the amount of tissue to be removed needs to be determined in a very precise and also in a safe manner. This means that there should be assurance that all cancerous tissues are indeed removed, but there is also an opposite need for letting the patient to keep the breast as much as possible. Obviously, the healthy tissues are desired to be maintained as much as possible but safety matters have to be strictly taken into account.

[0012] Hence, there emerges a need for an efficient, rapid and reliable screening and analysis method for the removed tissue while performing a lumpectomy surgery for the patient, for solving the problems discussed above.

[0013] Summary

[0014] The present invention introduces a method and a respective system for assessing a breast tissue specimen removed from a breast cancer patient during a breastconserving surgery.

[0015] There are five sub-concepts which are part of the same inventive concept and presented in the following as separate embodiments. However, these sub-concepts may be freely combined so that new combinatory sub-concepts are created as new possible embodiments of the present invention. The sub-concepts of the present invention are as follows.

[0016] The magnification concept: At first, a sample of cancerous tissue is X-ray imaged on a lower level of a measurement chamber, on top of a detector. After the first image is taken, the tissue sample is raised higher in the measurement chamber so that the interesting part of the tissue is closer to the X-ray source, thus the focal spot of the beam and the interesting area are centred along the primary axis of the X-ray beam. This requires usually both X- and Y-directional location adjustment for the sample, and also, setting higher in the Z-direction within the measurement chamber. The second image is able to provide a magnified illustration of the interesting part of the tissue sample. The geometric magnification works so that the imaged object is moved closer to the source of the X-ray beam and in that way its projection on the surface of the detector is magnified.

[0017] A search of crystal shapes or forms, i.e. a search process for obtaining data of areas where microcalcification occurs. There are two types of microcalcifications: Calcification oxalate and Calcium phosphate, where the latter locates right in the vicinity of the malignant cancer cells. To be more precise, calcium oxalate is associated with benign breast lesions, while calcium phosphate is associated with both benign and malignant breast lesions. The microcalcifications also form certain shapes. The present invention is capable to zoom into these areas and search similar crystal structures and their typical shapes.

[0018] Applying a multispectral imaging method, and as a result, obtaining an intensity graph for multiple images, where calcium phosphate and calcium oxalate will have their own intrinsic graph shapes. From these results it can be obtained, whether the finding (i.e. tissue sample) comprises calcium phosphate (indicating malignant cancer cells) or calcium oxalate (not indicating malignant cancer cells I indicating merely benign cell changes). Specifying the intensity graph creation in more detail, the present invention applies multispectral imaging methods and it is possible to get multiple spectral components for the image (in practice, multiple images of the same object with different energy settings are obtained) and then, it is possible to isolate (micro)calcifications in those images and it is possible to get an average pixel intensity for the (micro)calcifications. Average pixel intensities for every spectral component are represented as dots in the obtained graph. Defining / assessing a safe healthy margin; in other words, the relevant question is: Is all the malignant tissue already removed with the current specimen? Furthermore, this sub-concept is about making a decision for directing the patient to a second, continuation surgery based on this result. This is an undesired but yet a mandatory continuation of an unsuccessful operation in a form of a second surgery because of a non-sufficient first removal surgery of the cancerous tissues. The present invention indeed desires to minimize the probability for forcing the patient back into the surgery room. However, the healthy margin itself does not need to be minimized in the present invention. Artificial intelligence (Al) algorithms may make this analysis even better and more efficiently. Some examples, which may be performed more efficiently with the help of Al algorithm(s), comprise the tissue segmentation into healthy tissue, malignant tissue and artefacts such as (micro)calcifications. This process may also be called as classification of the tissue as a function of types of the found cells and lesions, and other found interesting parts and details. The Al algorithm is not focussed here in any greater detail.

[0019] Efficient post-processing algorithms are applied, such as CLAHE, HVS and JND and the end result is an enhanced image that allows more efficient classification for image segments both by human observers and by Al segmentation algorithms. CLAHE means “Dynamic Contrast Enhancement”; HVS means “Human Visual System”; and JND means “Just Noticeable Distortion”.

[0020] In other words, the present invention introduces a method for assessing a breast tissue specimen removed from a breast cancer patient, according to a first aspect of the present invention. The method is characterized in that the method comprises the steps of: imaging the specimen at a first time instant for obtaining a first image, wherein an imaging target at the first time instant comprises the whole specimen; inspecting the first image in order to find interesting areas; imaging the specimen at a second time instant for obtaining a second image, wherein an imaging target at the second time instant comprises a focussed geometrically magnified sub-section of the whole specimen comprising at least one of the found interesting areas; assessing locations of cancerous tissue in relation to healthy tissue within the specimen based on the first and / or second images; and based on the assessed locations, determining, whether there is a risk of any cancerous tissue remaining in the operated breast of the breast cancer patient.

[0021] In an embodiment of the present invention, the method comprises the step of: performing X-ray imaging operations in a closed X-ray measurement chamber comprising a detector, capable to apply spectral measurements over a wide spectrum range.

[0022] In an embodiment of the present invention, the method comprises the steps of: after the first image has been taken, adjusting the location of the specimen in the X-ray measurement chamber so that the X- and / or Y-coordinates of the location of the specimen are changed to focus on the interesting area for the second image, and raising the specimen higher in the measurement chamber in order to magnify the interesting area within the specimen on the second image.

[0023] In an embodiment of the present invention, the method comprises the steps of: performing segmentation on the first image; recognizing microcalcifications within the specimen, where the microcalcifications comprise calcium oxalate and calcium phosphate with certain intrinsic shapes within the breast tissue; and taking a second image on the focussed geometrically magnified sub-section of the whole specimen so that the focus is on the microcalcification areas.

[0024] In an embodiment of the present invention, the method comprises the step of: performing imaging at a single frequency and at a single energy level where different types of microcalcifications show most distinguishable characteristics between one another in the first and / or second image.

[0025] In an embodiment of the present invention, the method comprises the steps of: recognizing an outermost sub-section comprising microcalcifications within the specimen, and if it is fully under a surface of the specimen, assessing shortest distance of the outermost sub-section comprising microcalcifications from the surface of the specimen, and if it exceeds a predetermined safe healthy margin, determining the risk of any remaining cancerous tissue in the remaining part of the operated breast of the patient to be negligible. In an embodiment of the present invention, the method comprises the steps of: if an intensity or attenuation as a function of energy of any sub-section of the first or second image resembles respective properties of calcium oxalate or of calcium phosphate, determining the presence of calcium oxalate and / or calcium phosphate, respectively, in the inspected sub-section based on the resemblance of the respective graphs, and noting that the found calcium oxalate and / or calcium phosphate crystal shapes mark a close vicinity of the cancerous cells within the specimen so that calcium oxalate is associated with benign breast lesions, and calcium phosphate is associated with both benign and malignant breast lesions.

[0026] In an embodiment of the present invention, the method for assessing the breast tissue specimen is performed in the course of a removal surgery of cancerous breast tissue of the patient, in order to allow the current surgery to be continued after the determination result has been obtained.

[0027] In an embodiment of the present invention, the method comprises the step of: enhancing the contrast of a sub-section of the specimen dynamically before the assessment is made.

[0028] In an embodiment of the present invention, the method comprises the step of: allowing a human or a machine vision algorithm to inspect the first and / or second images themselves or their intensity or attenuation vs. energy or spectral component graphs per each sub-section, for revealing possible carcinoma or malignant calcium phosphate occurrences within the specimen, and / or the closest distance between the malignant findings in relation to the outer surface of the specimen.

[0029] In an embodiment of the present invention, the method comprises the step of: in case a carcinoma or malignant calcium phosphate occurrence location extends precisely to a surface of the specimen, instructing a surgeon to continue with the surgery focusing on this interesting area.

[0030] In an embodiment of the present invention, the method comprises the step of: classifying each sub-section of the specimen as either healthy, comprising benign, tissue, or as cancerous tissue, or as an undetermined sub-section where the classification cannot yet be made.

[0031] In an embodiment of the present invention, the method comprises the step of: marking differently classified sub-sections of the specimen with different colours on the image to be shown on a display of a system performing the method.

[0032] In an embodiment of the present invention, the method comprises the step of: if the second image assessment results in only calcium oxalate crystals being found, or with no specific findings, repeating the second imaging with a geometric magnification to a non-focussed volumetric sub-section of the specimen.

[0033] Furthermore, the present invention introduces a system for assessing a breast tissue specimen removed from a breast cancer patient, according to a second aspect of the present invention. The system is characterized in that the system comprises: a measurement chamber; an imaging device comprising a detector, placed inside the measurement chamber; a processor, memory, a display and a user interface for controlling the system and assessing obtained results, wherein the system is configured to: image the specimen at a first time instant for obtaining a first image, wherein an imaging target at the first time instant comprises the whole specimen; inspect the first image in order to find interesting areas on the display; image the specimen at a second time instant for obtaining a second image, wherein an imaging target at the second time instant comprises a focussed geometrically magnified sub-section of the whole specimen comprising at least one of the found interesting areas; assess locations of cancerous tissue in relation to healthy tissue within the specimen based on the first and / or second images; and based on the assessed locations, the system is configured to determine, whether there is a risk of any cancerous tissue remaining in the operated breast of the breast cancer patient. In an embodiment of the present invention, the imaging device is an X-ray imaging device.

[0034] In an embodiment of the present invention, the system further comprises: horizontal specimen movement means configured to move the specimen along X- and Y-directions on a horizontal plane via instructions of a user via the user interface, so that a center of an X-ray beam is set to a found interesting area for the second image; and vertical specimen movement means configured to raise the specimen higher within the measurement chamber, bringing it closer to an X-ray focal spot, in order to obtain a geometric magnification of the found interesting area on the detector.

[0035] In an embodiment of the present invention, the system for assessing the breast tissue specimen is configured to be performed in the course of a removal surgery of cancerous breast tissue of the patient, in order to allow the current surgery to be continued after the determination result has been obtained.

[0036] In an embodiment of the present invention, the system is further configured to: allow a human or a machine vision algorithm to inspect the first and / or second images themselves or their intensity or attenuation vs. energy or spectral component graphs per each sub-section, for revealing possible carcinoma or malignant calcium phosphate occurrences within the specimen, and / or the closest distance between the malignant findings in relation to the outer surface of the specimen.

[0037] In an embodiment of the present invention, the system is further configured to: classify each sub-section of the specimen as either healthy, comprising benign, tissue, or as cancerous tissue, or as an undetermined sub-section where the classification cannot yet be made.

[0038] In an embodiment of the present invention, the system is further configured to: mark differently classified sub-sections of the specimen with different colours on the image to be shown on the display of the system. Brief description of the drawings

[0039] FIG. 1 illustrates a main principle of the method according to an embodiment of the present invention,

[0040] FIG. 2 illustrates some of the applied elements and components in an imaging device configured to image the specimen obtained from a removal surgery of cancerous breast tissues,

[0041] FIG. 3 illustrates an example of a magnified specimen comprising an interesting finding in a form of a tumour in the obtained X-ray image,

[0042] FIG. 4 illustrates some basic forms which calcium oxalate crystals and calcium phosphate crystals form, and

[0043] FIG. 5 illustrates an attenuation vs. radiation energy graph for two different types of microcalcifications.

[0044] Detailed description

[0045] The present invention introduces a method and a respective system for assessing a breast tissue specimen removed from a breast cancer patient. The removal is performed in a respective surgery, which is meant to be a breast-conserving surgery. The presented method is meant to be applied right after the actual surgery, and the results of the method are meant to be used for determining, whether the removal surgery should be continued or not. In other words, the assessment of the removed tissue is meant to happen nearby, when the patient is still in the surgery room undergoing the removal surgery.

[0046] The present invention with its various embodiments relates to a situation, wherein a breast cancer patient has been offered a scheduled surgery time for removal of all tumorous tissues (i.e. cancerous / malignant tissues) in a breast-conserving surgery. The present invention is designed to minimize a risk for the breast cancer patient to undergo a second surgery later in the case of an unsuccessful first removal surgery of cancerous tissues. Such a continuation of the treatment is required, when it is noted, with any appropriate tracking method, that there are still malignant cancer cells within the breast of the patient (or in any other parts of the human body if metastases have been formed). The requirement to undergo also a second removal surgery is very stressful to the patient (also considering a potentially long waiting time before getting access to the actual second surgery) and also a waste of hospital resources, if this double surgery could be safely avoided. Furthermore, during the interval between the first and second surgeries, the cancer may well spread and make the second surgery an even trickier one.

[0047] Therefore, there is a clear need for enabling a possibility to assess the success of the tissue removal surgery during the actual surgery (i.e. during the initial first surgery) and without any excessive waiting time for the patient and the surgeon alike.

[0048] There are two possibilities in general: to assess the patient and the remaining breast tissue during the operation, or to assess a removed body part in another location nearby or even with an imaging device locating in the surgery room. The removed body part may be called as a breast tissue specimen to be assessed and analysed, for clarifying the applied notation here. The analysis of the removed body part for imaging purposes is under main focus in various embodiments of the present invention.

[0049] The present invention applies an X-ray imaging principle for the specimen in a measurement chamber, where predeveloped algorithms are used for analysing the imaged human tissue specimen and classifying the imaged human tissue specimen.

[0050] The method for assessing a breast tissue specimen removed from a breast cancer patient comprises the following features, in an embodiment of the present invention. FIG. 1 is referred to as well, for pointing out the respective features in this embodiment.

[0051] At first, the method comprises imaging 101 the specimen at a first time instant for obtaining a first image, wherein an imaging target at the first time instant comprises the whole specimen.

[0052] Secondly, the method comprises inspecting 102 the first image in order to find interesting areas. Thirdly, the method comprises imaging 103 the specimen at a second time instant for obtaining a second image, wherein an imaging target at the second time instant comprises a focussed geometrically magnified sub-section of the whole specimen comprising at least one of the found interesting areas 204a.

[0053] As a fourth step, the method comprises assessing 104 locations of cancerous tissue in relation to healthy tissue within the specimen based on the first and / or second images.

[0054] Finally, as a fifth step, the method comprises that based on the assessed locations, determining 105, whether there is a risk of any cancerous tissue remaining in the operated breast of the breast cancer patient.

[0055] FIG. 2 illustrates some of the applied elements and components in an imaging device configured to image the specimen obtained from a removal surgery of cancerous breast tissues. Such an imaging device represents an embodiment of the present invention. This embodiment presents a geometric magnification principle of the specimen between two successive image capturing actions. It is also possible to have more than two imaging actions for the same specimen, if required.

[0056] The imaging device is preferably an X-ray imaging device 200. The X-ray imaging device 200 comprises an X-ray source 201 in the top part of the imaging device, and a detector 202 which is an element shaped as a horizontal plate and locating in the bottom part of the imaging device. The X-rays are radiated downwards from the X-ray source 201 in this exemplary setup of the imaging device. The X-ray imaging device 200 is placed inside a measurement chamber (not shown), together with the specimen 203 to be imaged, so that no harmful X-ray radiation will propagate out of such a closed measurement space. The detector 202 may be set at a predetermined height within the measurement chamber, and in this example, the detector 202 locates on the horizontal bottom level marked with black colour. In an embodiment of the present invention, the height of the detector 202 may be set immediately on top of an inner horizontal bottom floor of the measurement chamber. However, the height for the detector 202 may also be different, and it may also be freely adjusted to a desired position e.g. with appropriate supporting rods or the like. As shown in the illustration, the dimensions for the detector 202 may be 300 mm (width) * 240 mm (depth), but this represents merely a single example among many possible options.

[0057] Before the imaging takes place, a part of a human breast has been surgically removed from the patient in the surgery room. The patient remains on the surgical platform while the removed specimen is brought to the imaging process. In other words, the imaging is set to happen in the midst of the cancer removal surgery.

[0058] At the first phase, the specimen 203 is placed directly on top of the detector 202 inside the measurement chamber. A first X-ray image is taken with this measurement setup. The illustration in the left-hand side of FIG. 2 shows the measurement situation in practice at the first phase.

[0059] The detector 202 captures the X-ray image by activating the X-ray source 201 by a user of the imaging device for a predetermined time period. The illustration in FIG. 2 shows the most interesting part of the obtained X-ray image and this part is shown in the form of findings 204a emphasized in the illustration of FIG. 2. The first taken X-ray image on the detector 202 presents the findings 204a in their natural (1x) size. The taken X-ray image is stored in a memory of the X-ray imaging device 200, in a suitable image database as an image file. The image file may also be transferred into a cloud service or to other device / computer / server for further analysis. This could be implemented as a PC connected to the X-ray imaging device which may locate right next to the imaging system.

[0060] The operation of the X-ray imaging device 200 is controlled by a processor. The processor executes and initiates the start of the X-ray source radiation exposure and ends the same after the predetermined time period has passed. Furthermore, the processor is configured to also set the kV (the voltage value in kilovolts) value that the X-ray tube will use for the exposure. The radiation exposure time (i.e. the length of the predetermined time period) may be readjusted based on the obtained image quality and contrast in the particular application situation so that the findings 204 could be tracked in the easiest possible way. The safety standards concerning the X-ray radiation levels are looser in this situation compared to the direct human body exposure applications, because the imaged tissue part has already been removed from the human breast and transported elsewhere. When the first X-ray image has been taken (and possibly stored in a memory) according to the configuration shown in the left-hand side of FIG. 2, the image may also be shown on the screen of the imaging device for direct presentation to the user of the imaging device. The user may visually track at least one of the interesting findings on the taken image and use this information for inspecting such an interesting finding in more detail in the second imaging taking place thereafter. For the second imaging action, the user may place the interesting finding i.e. the interesting part of the tissue specimen higher and in the centre i.e. in the center of the primary axis of the X-ray beam. This requires usually both X- and Y-directional location adjustment for the specimen along the platform. Either the platform or the specimen on top of it may be moved along the horizontal plane for achieving such X / Y coordinates adjustment for the second image. Otherwise, the X-ray imaging device 200 is prepared for the taking of a second, and magnified, image on the specimen 203. The geometric magnification is obtained by raising the specimen 203 to a higher horizontal level within the measurement chamber. The specimen 203 may locate on an X-ray transparent platform, when raising and setting a new vertical level for the specimen 203 within the measurement chamber (together with setting the correct X and Y coordinates for the specimen 203 for focusing onto the interesting finding seen in the first image). An exemplary height for the specimen 203 is shown in the right-hand side illustration of FIG. 2. The detector 202 remains in the same location as during the taking of the first image, i.e. the detector 202 locates along a bottom horizontal level in the context of the shown view in the illustration of FIG. 2. Summarizing the above movements for the specimen 203, the specimen 203 is raised higher in the measurement chamber so that the selected, interesting finding is high, closer to the focal spot and in the center of the X-ray beam.

[0061] The second image is then taken by exposing the raised specimen 202 for the X- rays emerging from the X-ray source 201 . The illumination beam is set to spread to a wider angle from the X-ray source 201 and, thus, create a magnified projection of the object onto the detector 202 so that the obtained image will look like the example shown along the detector 202 on the right-hand side illustration. The second image now shows magnified findings 204b, and in this example the magnified findings 204b appear visually five times larger on the detector 202 plate than the respective findings 204a in the first image. The magnification factor may be set to some other desired value, too, and this may be accomplished by setting the height of the specimen’s platform to a new height level within the measurement chamber. The illustration in the 5-times magnified image shows three interesting findings 204b on the image, but there may be only a single interesting finding which was noted by the user from the first image. The magnification then reveals this finding in greater detail.

[0062] From the magnified second image and the magnified findings 204b, the assessment of the specimen 203 is easier to make, and also the characteristics of the findings are in more detail available and visually clearer for further analysis.

[0063] The second imaging action may well be repeated for all interesting findings found in the first image. This means that the X and Y coordinates may be set to new values focusing on an unimaged part of the specimen 203, and the magnified image can be taken from this fresh part of the specimen 203. The magnified image taking phase can hence be repeated until all the interesting findings of the first image have been gone through.

[0064] Next, some typical finding types related commonly to human breast tissue and breast cancer cells and agglomeration of such cells and some “markers” common in the vicinity of breast cancer (both benign and malignant) cells, are discussed.

[0065] FIG. 3 is presented in this regard, and it illustrates an example of a magnified specimen 203 comprising an interesting finding in a form of a tumour 304 within the specimen 203, shown in the taken X-ray image (either in the first or second image). It is next assumed that the second image is taken into analysis as comprising the more detailed imaging result of the specimen 203.

[0066] The obtained image may comprise several mutually similar or varying findings, and in this example, it is focussed on a single finding which is shown as starshaped or squid-shaped finding in the obtained X-ray image of the specimen 203. In general, the findings (which differ from the healthy breast tissue) may represent benign cancer cells, malignant cancer cells, or some other kind of tissue which may be even hard to classify. Furthermore, the specimen 203 may comprise microcalcification structures in a form of crystal shapes which act as markers for certain types of cancerous cells. In more detail, the specimen 203 may comprise calcium oxalate (as e.g. calcium oxalate dihydrate) 301 and this matter can be associated with benign breast tissue cell changes. On the other hand, the specimen 203 may comprise calcium phosphate (such as e.g. hydroxyapatite) 302 and this matter can be associated with both benign artefacts and malignant breast tumours. Calcium phosphate may be formed as a result of cell necrosis, in other words, when cancer cells die because the lack of oxygen, they leave microcalcifications formed of calcium phosphate behind. Therefore, the areas comprising calcium phosphate 302 are of particular interest.

[0067] As the illustration points out, malignant tumour 304 may locate in close proximity to calcium phosphate 302 crystal shapes. Furthermore, such crystals can be seen and recognized in the X-ray image, and the shapes of the crystals can also be assessed to some extent; the greater the magnification factor, the better. For the shapes and forms usually found in connection with the described matters, see also the description in connection with FIG. 4 later. As a result, the areas comprising calcium phosphate crystal shapes should be inspected in detail, in order to obtain sufficient information on the tumour 304 area and its expanded branches in all possible directions within the 3-dimensionally observed tissue specimen 203. Indeed, the shape of the tumour 304 may be notably non-symmetric, and the question now arises on the location of the whole tumour 304 regarding the removed specimen 203 and its outer surfaces.

[0068] In this context, the term “healthy margin” 303 comes into play. The surgeon and the whole medical team operating the patient is interested in whether the removed specimen 203 comprises a tumour 304 (or all found tumours / carcinomas) as a whole, and whether there can be made an assessment that the remaining breast tissue of the patient is completely tumour-free. The assessment of a healthy margin 303 helps this decision-making. When the visible carcinoma has been identified, and all the calcium oxalate comprising sub-sections of the specimen 203 have been determined and also all the calcium phosphate comprising sub-sections of the specimen 203 have been determined, there can be made a calculation for a distance from each malignant form towards the outer surface of the specimen 203 (= the closest distance). If the malignant findings (meaning at least one of such agglomerations) locate and extend right to the tissue surface of the specimen 203, then the area is thoroughly examined and if calcifications are indeed associated with malignant lesions, then a decision to continue the surgery in that respective location of the breast can be made and this information can be provided to the surgeon. Furthermore, the purpose of the assessment of the calcifications is that the carcinoma is sometimes difficult to be distinguished from the healthy tissue and the calcifications help the radiologist or the image analysis algorithm to focus the attention to the area, which is the most relevant and most interesting in this sense. Otherwise, the margins from each of the crystal shapes agglomerations towards the surface of the specimen 203 is calculated, and their minimum value is picked, and set as the healthy margin 303. If this minimum value i.e. the healthy margin 303 exceeds 2 millimeters, and if there can be made an assessment that all cancerous tissues are indeed already removed from the patient’s body, it is then possible to deduce that the removal surgery has already been successful, and all cancerous tissues have been successfully removed by surgical means. After this assessment, the surgeon may be given orders to conclude the surgery in a successful fashion. After all, this assessment has been performed while the surgery was still ongoing in the surgery room.

[0069] Usually, there is no absolute need to minimize the healthy margin 303, but rather, to have the healthy margin 303 to be on a safe side, but not in an over-exaggerated fashion. The surgeons familiar with the breast cancer surgery know, what is the correct amount of tissues to be removed in such a breast-conserving surgery. Therefore, if the specimen 203 is confirmed to include the carcinoma 304, and the healthy margin 303 exceeds the above-mentioned threshold value, there can be made a conclusion that the already performed surgery is enough, and the surgery can be concluded by the surgeon.

[0070] Regarding the location of the X-ray imaging device, it may be performed in different room of the hospital, where the X-ray imaging device with all the processing and data storing capabilities is already available. In an embodiment, the X-ray imaging device may locate even in the surgery room, making the transport time and distance very short for the specimen. The surgeon may perform the imaging and the visual assessments personally or with his / her assistant, so that the whole process around the removal surgery happens within the same room. This gives the added benefit that it is easy to continue the surgery from the correct location of the patient’s body, as the display for the imaging device is available in the same surgery room for the surgeon (and his / her team) to use.

[0071] FIG. 4 illustrates some basic forms which calcium oxalate 301 crystals and calcium phosphate 302 crystals form. Tetragonal form of crystals is common for calcium oxalate (dihydrate) 301. For hydroxyapatite (i.e. calcium phosphate 302), needle-like crystals are common. If the X-ray image as magnified shows clear crystal structures in a tetragonal or needle-like form, there can be made a preliminary assessment that such crystals are markers for benign artefacts and / or malignant cancer cells (i.e. carcinoma), as described above in connection with FIG. 3. Further imaging or other assessment on the tissue specimen 203 may follow this initial assessment.

[0072] There is a complementary manner of assessing the obtained X-ray image (either the first one or the second one), which may be applied in an embodiment of the present invention. Depending on the radiated X-ray energy levels, the attenuation due to calcium oxalate crystals within the specimen varies. In other words, the larger the photon energy, the smaller the linear coefficient of attenuation. This means that higher energy radiation reaches the detector to a larger extent, which is a natural result. However, for calcium phosphate crystals, the attenuation graph is different compared to calcium oxalate crystals. In other words, the slopes between these two graphs differ. This phenomenon is shown in FIG. 5 for these two materials. If the user using some post-processing tools can obtain such an attenuation graph from various parts (e.g. pixels) of the X-ray image, he / she is capable to determine whether such a tissue section includes crystal shapes formed of calcium oxalate or calcium phosphate. The analysis work could be automized too, in an embodiment of the present invention.

[0073] In an embodiment, another option for analyzing the obtained X-ray image is to pick a sub-section (represented by a group of pixels) and calculate the detected signal intensity (= level of received X-ray radiation; the intensity is also proportional to the attenuation) as a function of energy (in other words, across varying photon energies). This may also be called as a certain spectrum in relation to the incoming energy levels. This of course requires some scanning of energy levels in the X-ray source’s 201 end. The tissue parts containing only calcium oxalate 301 crystals will have a different attenuation / energy graph than the tissue parts containing only calcium phosphate 302 crystals. From the obtained graphs of different interesting sub-sections (pixels) of the specimen 203, it is possible to deduce what subsections comprise calcium oxalate and which comprise calcium phosphate. From this assessment, there can be made a conclusion on sub-sections which contain only benign artefacts and which contain possibly malignant tumours (i.e. carcinomas). In case the characteristics of the subsection do not follow either of these intrinsic graphs, it can be deduced that such a sub-section comprises a fully healthy tissue. It is possible that the assessment cannot be made, and the result leaves reasonable doubt on the type of the removed tissue in a certain group of sub-sections. In that situation, it is possible to image such a sub-section once again with a different magnification factor. In an embodiment, the undetermined sub-section is raised higher in the measurement chamber so that it is closer to the X-ray source’s focal spot and centered along the primary axis of the X-ray beam, and then the X-ray imaging is performed for all the undetermined sub-sections in a successive fashion.

[0074] In an embodiment of the present invention, the calculation of the healthy margin 303 is performed by the processor via appropriate software, and the result is given to the user of the X-ray imaging device by traffic lights style of indication means: green LED light means a safe healthy margin and the surgery can be safely concluded, a yellow light means reasonable doubt on the assessment and that the analysis process should be continued, and a red light means that the carcinoma extends right to the immediate surface of the specimen, meaning that the continuation order for the surgeon needs to be given to the surgery room. The taken X-ray images may be provided to the surgeon for giving proper feedback, in order to have a good plan for the continuation of the operation. This could be made even directly via the display of the X-ray imaging device, if it locates in the surgery room.

[0075] In an embodiment, the imaging device may be set to classify each group of pixels of the obtained X-ray image with the above-mentioned colour scheme using an image segmentation method for each sub-section (i.e. for each area), so that the whole image is marked with safe sub-sections (green), sub-sections under clear concern (red) and also the “grey spots” which cannot yet be classified (yellow). This resulting colour-coded image could be provided to a medical professional user for further assessment. The user hence may have a final say, whether the already performed surgery is already a successful one, or whether the surgery needs to be continued for removing the rest of the carcinoma(s). The user may also decide to continue the imaging process of the specimen towards different findings under interest than earlier, and / or with different magnification factors than earlier for the currently assessed findings. With such repeated imaging process on the specimen, it is possible to assess the “grey spots” better so that it is easier to classify them as either a green or red sub-section. As a summary for all of the above disclosure, the present invention is a machine- assisted tool for the medical professional to be assisted in making a decision on whether the already performed cancerous tissue removal surgery is sufficient, for obtaining a long-term successful end result where all cancer cells have been eliminated from the patient’s body. The machine-assisted tool gives classification information on the whole specimen and the safe healthy margin, and these results are combined with human assessment which is done by the medical professional. The decision on whether to continue with the surgery, is thus based on a combination of human assessment and the results provided by the presented imaging device. The present invention is a technical assistance tool for medical professionals for improving the success rates of the breast-conserving cancer surgeries.

[0076] In an embodiment, the energy of the exposed X-ray radiation may be selected based on the spectral analysis of pixels (or pixel groups) in the images based on the two types of microcalcifications discussed above. Such an energy is picked, which gives a notable distinction between the attenuation graphs of the two microcalcification types. In an embodiment, such a distinction (i.e. difference) could even be maximized when performing a design process of the imaging device, and the found energy could thereafter be used in the actual X-ray imaging process according to the present invention. With such a selection process, the identification of different microcalcification types becomes easier, and presumably, there are less “grey spots” which require reassessment or reimaging.

[0077] Back to the geometric magnification principle described earlier, there is a potential problem of losing sharpness in the edges of the X-ray image, if the magnification factor is high enough (i.e. the specimen is raised very high, relatively speaking). This problem can be assessed by making combined images by image fusion techniques, in an embodiment of the present invention. In this way, the less magnified image(s) may be used in obtaining images with good sharpness on the interesting findings. In addition, the more magnified image(s) may be used for obtaining spatial information on the imaged findings. These pieces of information may be combined so that both the exact spatial information, and e.g. the shapes of the findings (i.e. forms of the crystal shapes of the microcalcifications) can be obtained. This improves the quality of the assessment, and also decreases the risk for the patient to undergo a second surgery at a later time. The present invention may vary within the scope of the claims.

Claims

Claims1 . A system for assessing a breast tissue specimen (203) removed from a breast cancer patient, characterized in that the system comprises: a measurement chamber; an X-ray imaging device (200) comprising a detector (202), placed inside the measurement chamber; a processor, memory, a display and a user interface for controlling the system and assessing obtained results, wherein the system is configured to: image (101 ) the tissue specimen (203) at a first time instant for obtaining a first image, wherein an imaging target at the first time instant comprises the whole tissue specimen (203); inspect (102) the first image in order to find interesting areas (204a) on the display; raise the tissue specimen (203) higher in the measurement chamber so that an interesting area (204a) of the tissue specimen (203) is closer to an X-ray source, thus the focal spot of a beam and the interesting area (204a) are centred along a primary axis of the X-ray beam; image (103) the tissue specimen (203) at a second time instant for obtaining a second image, wherein an imaging target at the second time instant comprises a focussed geometrically magnified sub-section of the whole tissue specimen (203) comprising at least one of the found interesting areas (204a); assess (104) locations of cancerous tissue in relation to healthy tissue within the tissue specimen (203) based on the first and / or second images; and based on the assessed locations, the system is configured to determine (105), whether there is a risk of any cancerous tissue remaining in the operated breast of the breast cancer patient.

2. The system according to claim 1 , characterized in that the system further comprises: horizontal specimen movement means configured to move the tissue specimen (203) along X- and Y-directions on a horizontal plane via instructions of a user via the user interface, so that the center of the X-ray beam is set to the found interesting area (204a) for the second image; and vertical specimen movement means configured to raise the tissue specimen (203) higher within the measurement chamber, bringing it closer to the X-ray focal spot, in order to obtain a geometric magnification of the found interesting area (204a) on the detector (202).

3. The system according to claim 1 , characterized in that the system for assessing the breast tissue specimen (203) is configured to be performed in the course of a removal surgery of cancerous breast tissue of the patient, in order to allow the current surgery to be continued after the determination (105) result has been obtained.

4. The system according to claim 1 , characterized in that the system is further configured to: allow a human or a machine vision algorithm to inspect the first and / or second images themselves or their intensity or attenuation vs. energy or spectral component graphs per each sub-section, for revealing possible carcinoma or malignant calcium phosphate (302) occurrences within the tissue specimen (203), and / or the closest distance (303) between the malignant findings in relation to the outer surface of the tissue specimen (203).

5. The system according to claim 1 , characterized in that the system is further configured to: classify each sub-section of the tissue specimen (203) as either healthy, comprising benign, tissue, or as cancerous tissue, or as an undetermined subsection where the classification cannot yet be made.

6. The system according to claim 5, characterized in that the system is further configured to: mark differently classified sub-sections of the tissue specimen (203) with different colours on the image to be shown on the display of the system.

7. The system according to claim 1 , characterized in that the system is further configured to: perform segmentation on the first image; recognize microcalcifications within the tissue specimen (203), where the microcalcifications comprise calcium oxalate (301 ) and calcium phosphate (302) with certain intrinsic shapes within the breast tissue; andtake a second image on the focussed geometrically magnified sub-section of the whole tissue specimen (203) so that the focus is on the microcalcification areas.

8. The system according to claim 7, characterized in that the system is further configured to: perform imaging at a single frequency and at a single energy level where different types of microcalcifications show most distinguishable characteristics between one another in the first and / or second image.

9. The system according to claim 7, characterized in that the system is further configured to: recognize an outermost sub-section comprising microcalcifications within the tissue specimen (203), and if it is fully under a surface of the tissue specimen (203), assess shortest distance (303) of the outermost sub-section comprising microcalcifications from the surface of the tissue specimen (203), and if it exceeds a predetermined safe healthy margin, determine the risk of any remaining cancerous tissue in the remaining part of the operated breast of the patient to be negligible.

10. The system according to claim 1 , characterized in that the system is further configured to: if an intensity or attenuation as a function of energy of any sub-section of the first or second image resembles respective properties of calcium oxalate (301 ) or of calcium phosphate (302), determine the presence of calcium oxalate (301 ) and / or calcium phosphate (302), respectively, in the inspected sub-section based on the resemblance of the respective graphs, and note that the found calcium oxalate and / or calcium phosphate crystal shapes mark a close vicinity of the cancerous cells within the tissue specimen (203) so that calcium oxalate (301 ) is associated with benign breast lesions, and calcium phosphate (302) is associated with both benign and malignant breast lesions.11 . The system according to claim 1 , characterized in that the system is further configured to:enhance the contrast of a sub-section of the tissue specimen (203) dynamically before the assessment is made.

12. The system according to claim 3, characterized in that the system is fur- ther configured to: in case a carcinoma or malignant calcium phosphate (302) occurrence location extends precisely to a surface of the tissue specimen (203), instruct a surgeon to continue with the surgery focusing on this interesting area.

13. The system according to claim 1 , characterized in that the system is further configured to: if the second image assessment results in only calcium oxalate (301 ) crystals being found, or with no specific findings, repeat the second imaging with a geometric magnification to a non-focussed volumetric sub-section of the tissue specimen (203).

Images