Motion Detection for Internal Breast Tissue in Tomosynthesis

Abstract

Methods and systems for identifying internal motion of a patient's breast during an imaging procedure. The method can include compressing the patient's breast in a mediolateral oblique (MLO) position. During the compression of the breast, a first tomosynthesis MLO projection frame is acquired for a first angle with respect to the breast, and a second tomosynthesis MLO projection frame is acquired for a second angle with respect to the breast. The boundaries of the pectoral muscle are identified in the projection frames and a boundary representation is generated. A difference between the first representation and the second representation is determined. Then a motion score is generated based at least on the difference between the first representation and the second representation.

Description

[0001] Cross - reference to related applications

[0002] This application was filed on August 31, 2020, as a PCT international patent application, and claims the priority and benefit of U.S. Provisional Patent Application Serial No. 62 / 907,079, filed on September 27, 2019, the disclosure of which is incorporated herein by reference in its entirety. Background Art

[0003] X - ray screening examinations are used to detect breast cancer and other diseases. Efforts to improve the sensitivity and specificity of breast x - ray systems have led to the development of tomosynthesis systems. Breast tomosynthesis is a three - dimensional imaging technique that involves acquiring images of a stationary, compressed breast from multiple angles during a short scan. The individual images are reconstructed into a series of high - resolution thin slices that can be displayed individually or in a dynamic cine mode. The reconstructed tomosynthesis slices reduce or eliminate problems caused by tissue overlap and structural noise in single - slice two - dimensional mammographic imaging. Digital breast tomosynthesis also offers the potential for reduced breast compression, improved diagnostic and screening accuracy, reduced recalls, and 3D lesion localization.

[0004] Regarding these and other general considerations, the aspects disclosed herein have been made. Moreover, while relatively specific problems may be discussed, it should be understood that the examples should not be limited to solving the specific problems identified in the background art or elsewhere in this disclosure. Summary of the Invention

[0005] The present technology relates to detecting internal breast tissue movement during an imaging procedure. In one aspect, the technology relates to a method for identifying internal movement of a patient's breast during an imaging procedure. The method includes compressing the patient's breast in the mediolateral oblique (MLO) position; during the compression of the breast, acquiring a first tomosynthesis MLO projection frame for a first angle with respect to the breast; during the compression of the breast, acquiring a second tomosynthesis MLO projection frame for a second angle with respect to the breast; identifying a first boundary of the pectoralis muscle in the first projection frame; generating a first representation of the first boundary of the pectoralis muscle; identifying a second boundary of the pectoralis muscle in the second projection frame; generating a second representation of the second boundary of the pectoralis muscle; determining a difference between the first representation and the second representation; and generating a motion score based at least on the difference between the first representation and the second representation.

[0006] In an example, the first generated representation is a two-dimensional representation. In another example, the difference is based on the area between the first representation and the second representation. In yet another example, the difference is based on the minimum distance between the first representation and the second representation. In yet another example, the method further includes comparing the difference with an expected value, where the expected value is based on at least one of the following: the x-ray angle of the x-ray source of the first projection frame and the x-ray angle of the x-ray source of the second projection frame, or at least based on the fitted curves of the first tomosynthesis MLO projection frame and the second tomosynthesis MLO projection frame; and generating a motion warning based on the comparison of the difference with the expected value. In yet another example, the method further includes concurrently displaying at least a portion of the first projection frame and the second projection frame with a plurality of parallel motion guides in a cine view.

[0007] In another example, the method further includes receiving a selection of one of the plurality of parallel motion guides; receiving an input to move the selected parallel motion guide to a new position; and displaying the selected parallel motion guide at the new position based on the received input to move the selected parallel motion guide. In yet another example, the plurality of parallel motion guides are evenly spaced relative to each other.

[0008] In another aspect, the technique relates to a method for identifying internal motion of a patient's breast during an imaging procedure. The method includes compressing the patient's breast in a mediolateral oblique (MLO) position; acquiring a plurality of tomosynthesis MLO projection frames during the compression of the breast, where the plurality of tomosynthesis MLO projection frames include images of a portion of the patient's breast and a portion of the pectoral muscle; for at least two of the plurality of tomosynthesis MLO projection frames, identifying the boundaries of the pectoral muscle; for at least two of the plurality of tomosynthesis MLO projection frames, generating a representation of the boundaries of the pectoral muscle; determining a first difference between the generated representations for at least two of the plurality of tomosynthesis MLO projection frames; determining a second difference between the first difference and an expected value of the first difference; comparing the second difference with a predetermined threshold; and generating a motion warning based on the comparison of the second difference with the predetermined threshold.

[0009] In an example, the generated representation is a two-dimensional representation. In another example, the first difference is based on the area between the generated representations. In yet another example, the first difference is based on the minimum distance between the generated representations. In yet another example, the second difference is a shift variance value. In yet another example, the method further includes continuously displaying at least a portion of the projection frames concurrently with a plurality of parallel motion guides in a cine view.

[0010] In another example, the method includes receiving a selection of one of a plurality of parallel motion guides; receiving an input to move the selected parallel motion guide to a new position; and based on the received input to move the selected parallel motion guide, displaying the selected parallel motion guide at the new position. In yet another example, the plurality of parallel motion guides are evenly spaced relative to each other.

[0011] In another aspect, the technology relates to a system for identifying internal motion of a patient's breast during an imaging procedure. The system includes an x-ray source configured to rotate around the breast; a compression paddle configured to compress the breast in a mediolateral oblique (MLO) position; and an x-ray detector deployed opposite the compression paddle from the x-ray source. The system further includes at least one processor; and a memory storing instructions that, when executed by the at least one processor, cause the system to perform a set of operations. The operations include, during compression of the breast in the MLO position, emitting a first x-ray emission from the x-ray source at a first angle relative to the breast; after the first x-ray emission passes through the breast, detecting the first x-ray emission from the x-ray source by the x-ray detector; emitting a second x-ray emission from the x-ray source at a second angle relative to the breast; and after the second x-ray emission passes through the breast, detecting the second x-ray emission by the x-ray detector. The method further includes generating a first tomosynthesis MLO projection frame for the first angle based on the detected first x-ray emission; generating a second tomosynthesis MLO projection frame for the second angle based on the detected second x-ray emission; identifying a first boundary of the pectoral muscle in the first projection frame; generating a first representation of the first boundary of the pectoral muscle; identifying a second boundary of the pectoral muscle in the second projection frame; generating a second representation of the second boundary of the pectoral muscle; determining a difference between the first representation and the second representation; and generating a motion score based at least on the difference between the first representation and the second representation.

[0012] In an example, the first generated representation is a two-dimensional representation. In another example, the difference is based on the area between the first representation and the second representation. In yet another example, the difference is based on the minimum distance between the first representation and the second representation.

[0013] In another aspect, the technology relates to a method for identifying internal movement of a patient's breast during an imaging procedure. The method includes compressing the patient's breast; acquiring a plurality of tomosynthesis projection frames during the compression of the breast, wherein the plurality of tomosynthesis projection frames include images of a portion of the patient's breast and a portion of the pectoral muscle; for at least a subset of the plurality of tomosynthesis projection frames, identifying the boundary of the pectoral muscle; for the identified boundary of the pectoral muscle, generating a boundary representation of the identified boundary of the pectoral muscle; measuring the distance between the generated boundary representations for at least a subset of all possible pairs of boundary representations; determining an expected distance value for each boundary representation for which the distance is measured; based on the measured distance and the expected distance value, determining a shift variance for each boundary pair for which the distance is measured; comparing the offset variance with a predetermined threshold; and generating a motion warning based on the comparison of the shift variance with the predetermined threshold.

[0014] In another aspect, the technology relates to a method for identifying internal movement of a patient's breast during an imaging procedure. The method includes compressing the patient's breast; acquiring a plurality of tomosynthesis projection frames during the compression of the breast, wherein the plurality of tomosynthesis projection frames include images of a portion of the patient's breast and a portion of the pectoral muscle; for at least a subset of the plurality of tomosynthesis projection frames, identifying the boundary of the pectoral muscle; for the identified boundary of the pectoral muscle, generating a boundary representation of the identified boundary of the pectoral muscle; generating a reference line that intersects the generated boundary representation; identifying reference points along the reference line; for at least a subset of the generated boundary representations, calculating the intersection distance from the reference point to the intersection of the corresponding boundary and the reference line; determining an expected intersection distance value based on the calculated intersection distance; determining an intersection shift variance for each boundary representation for which the intersection distance is calculated; comparing the intersection shift variance with a predetermined threshold; and generating a motion warning based on the comparison of the intersection shift variance with the predetermined threshold.

[0015] The present invention content is provided to introduce some concepts in a simplified form, which will be further described in the following detailed description. The present invention content is neither intended to identify the key features or essential features of the claimed subject matter, nor intended to be used to limit the scope of the claimed subject matter. Additional aspects, features, and / or advantages of the examples will be partly set forth in the following description, and partly will become apparent from the description, or can be learned through the practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The following non-limiting and non-exhaustive examples are described with reference to the accompanying drawings.

[0017] Figure 1 An example portion of a tomosynthesis system is depicted, where the breast is compressed in the mediolateral oblique (MLO) position.

[0018] Figure 2ADepicts a plurality of projection frames for a set of tomosynthesis projection frames acquired during a MLO tomosynthesis imaging protocol.

[0019] Figure 2B Depicts Figure 2A a plurality of projection frames having a representation of the boundaries for the pectoral muscles.

[0020] Figure 3A Depicts an example drawing of the pectoral muscle boundary representation in the case where no patient movement occurred during the tomosynthesis imaging protocol.

[0021] Figure 3B Depicts an example drawing of the pectoral muscle boundary representation in the case where patient movement occurred during the tomosynthesis imaging protocol.

[0022] Figure 3C Depicts a drawing of the distances between the pectoral muscle boundary representations identified in the projection frames.

[0023] Figure 3D Depicts another example drawing of the pectoral muscle boundary representation.

[0024] Figure 3E Depicts an example intersection distance (I) measurement.

[0025] Figure 3F Depicts an example drawing of the intersection distance (I).

[0026] Figure 4A Depicts an example method for approximating or identifying the movement of internal breast tissue during a tomosynthesis protocol.

[0027] Figure 4B Depicts another example method for approximating or identifying the movement of internal breast tissue during a tomosynthesis protocol.

[0028] Figure 4C Depicts another example method for approximating or identifying the movement of internal breast tissue during a tomosynthesis protocol.

[0029] Figure 4D Depicts another example method for approximating or identifying the movement of internal breast tissue during a tomosynthesis protocol.

[0030] Figure 5A Depicts an example medical image of a breast with multiple movement guides.

[0031] Figure 5B Depicts an example series of projection frames for a tomosynthesis imaging protocol for a stationary breast.

[0032] Figure 5C Depicts an example series of projection frames in which breast movement occurred during the tomosynthesis imaging protocol.

[0033] Figure 6 depicts an example method for displaying a motion guide for medical images.

[0034] Figure 7 depicts an example of a suitable tomosynthesis system in which one or more of the present embodiments may be implemented.

[0035] Figure 8 depicts an example of a suitable operating environment in which one or more of the present embodiments may be implemented. DETAILED DESCRIPTION

[0036] As discussed above, breast tomosynthesis is a three-dimensional imaging technique that involves acquiring images of a stationary, compressed breast at multiple angles during a short scan. The individual images are reconstructed into a series of thin, high-resolution slices. Because multiple images are captured over a period of time and used for reconstruction, there is a possibility that the patient may move during the tomosynthesis imaging procedure. Motion during the procedure can have a negative impact on the resulting reconstruction and the quality of the tomosynthesis slices. Specifically, patient motion can cause blurring, anatomical distortion, and / or artifacts, which are exaggerated during longer exposure times. If the patient's motion is substantial, additional imaging procedures may be required to obtain better quality tomosynthesis images of the patient. The ability to automatically detect motion at or near the end of the tomosynthesis imaging procedure allows the patient to be re-imaged while the patient is still at the imaging facility. For example, without an automated motion detection technique, patient motion during the imaging procedure will not be recognized, if at all, until the physician reviews the set of medical images and notices blurring or other signs of patient motion. This review often occurs days or even weeks after the imaging procedure. Consequently, the patient will then have to return to the imaging facility for additional imaging at a later date. With automated motion detection, if there is substantial motion during the first imaging procedure, the patient can be re-imaged almost immediately after the first imaging procedure. Additionally, the automated motion detection techniques discussed herein can also provide a fraction or measure of the detected motion. This measure or fraction can further be used in motion correction or inference techniques to improve the final image quality.

[0037] Some motion detection concepts are discussed in U.S. Patent No. 9,498,180 (the '180 patent), which is incorporated herein by reference in its entirety. The '180 patent discloses techniques for identifying the motion of the skin line of the breast. Identifying the motion of the skin line has many benefits, including the fact that the skin line appears in the most commonly acquired image views (i.e., cranio-caudal (CC) and mediolateral oblique (MLO) views). It has been found that in some cases, the motion of the skin line does not accurately reflect the motion of the internal breast tissue. That is, in some cases, the skin line can move during the imaging protocol, but the internal breast tissue can remain substantially stationary. Conversely, examples also occur where the internal breast tissue moves during the imaging process but the skin line remains substantially stationary. This distinction is important where a lesion or region of interest occurs away from the skin line, and the examiner may need to know whether breast motion is occurring near the lesion.

[0038] To help address this issue, a new technique for approximating the motion of the internal breast tissue has been developed. More specifically, the technique examines the position of the pectoral muscle in a plurality of tomosynthesis projection frames. Based on the position of the pectoral muscle, the presence of motion and the magnitude of such motion can be identified. Due to the position of the pectoral muscle in the projection, the motion of the pectoral muscle provides a more accurate approximation of the motion of the internal tissue of the breast compared to the motion of the skin line. The technique can also utilize other internal structures of the breast or the patient, such as implants in the patient's breast or chest wall muscles.

[0039] However, a corresponding disadvantage of some embodiments of the technique is that some embodiments may only be used for a subset of the medical image views of the breast. For example, in images where the pectoral muscle is generally absent, such as in the CC view, the technique may not be able to approximate the internal motion of the breast based on the pectoral muscle. For images where the pectoral muscle is present, the technique provides an improved approximation of the motion of the internal breast tissue. The most common view where the pectoral muscle is present is the MLO view. To obtain the MLO view, the tomosynthesis gantry is rotated approximately 45 degrees and the patient's breast is compressed at a 45-degree angle. Due to the 45-degree compression, the MLO compression is often more uncomfortable for the patient than other views (e.g., the CC view). Due to the increased discomfort, the patient is more likely to move during the protocol and the motion is more likely to be substantial. In an internal study, it has been identified that approximately 66% of patient motion occurs during MLO compression. Thus, while some embodiments of the technique may not be used for all views, the technique is useful for the view where substantial motion is most likely to occur.

[0040] Figure 1Illustrates an example portion of a tomosynthesis system, where the breast is compressed in the MLO position. The example system includes an x-ray source 100 that moves along an arc 101, a compression paddle 104, a breast platform 106, and an x-ray detector or receiver 110. During a tomosynthesis scan, the patient's breast 102 is fixed and compressed between the compression paddle 104 and the breast platform 106. The x-ray receiver 110 is deployed within a housing located below the breast platform 106. After the x-rays have passed through the breast 102, the x-ray receiver 110 receives and / or detects the x-rays emitted from the x-ray source 100. The x-ray source 100 moves along the arc 101, which can be centered on the top surface of the receiver 110. At predetermined discrete positions, the source 100 is energized to emit a collimated x-ray beam, for example but not limited to at every 1.07° of an arc of + / -7.5°. The beam irradiates the breast 102, and the radiation that has passed through the breast is received by the receiver 110. The receiver 110 and associated electronics generate image data in digital form for each pixel of a rectangular grid of pixels at each predetermined discrete angular position of the source 100. In the MLO position, the breast is compressed at an angle (θ) of approximately 45 degrees with respect to the vertical direction. In some examples, the compression can be between approximately 40 - 60 degrees.

[0041] The movement of the source 100 can be continuous or discontinuous. If the movement is continuous, then corresponding sets of image data are accumulated over small increments of the continuous movement, such as an arc of 0.1° to 0.5° of the movement of the source 100, but these non-limiting parameters are merely examples. Different ranges of movement of the source 100 can be used, and the movement of the source 100 can be along an arc centered on a different axis, such as inside the fixed breast 102 or at the breast platform 106 or at the receiver 110. Additionally, the source movement need not be along an arc and can be translational or a combination of different types of movement, such as partial translation and partial rotation. In some examples, the x-rays can be emitted between 7.5° and 7.5° from the midpoint of the arc, and 15 different projection frames can be obtained from a single tomosynthesis imaging protocol.

[0042] Different features 103 of the breast will project onto the detector at different positions for each different image, resulting in projection paths 120, because the x-ray source position is different for each image. Additionally, the projection paths 120 among all viewpoints generally follow a smooth trajectory for tomosynthesis scans, because of, for example, the way the x-ray source movement is defined in a controlled arc, and because the x-rays are exposed in a time and space uniform manner, and this smooth trajectory is without patient movement. However, if the patient moves during the scan, then the projection of the features will not follow a smooth trajectory.

[0043] Figure 2ADepicts a plurality of projection frames that are a set of tomosynthesis projection frames acquired during a MLO tomosynthesis imaging protocol. In the depicted example image, an example first projection frame 202, an example second projection frame 204, and an example third projection frame 206 are depicted. The first MLO projection frame 202 can be the first MLO projection frame acquired during the tomosynthesis imaging protocol, the second MLO projection frame 204 can be the eighth MLO projection acquired during the imaging protocol, and the third MLO projection frame 206 can be the fifteenth MLO projection acquired during the tomosynthesis imaging protocol. Thus, in an imaging protocol where fifteen projection frames are acquired (e.g., the x-ray source emits radiation at fifteen different angular positions), the first MLO projection frame 202 represents the starting image, the second MLO projection frame 204 represents an intermediate image, and the third MLO projection frame 206 represents the ending image. In each projection frame, breast tissue as well as pectoral muscle can be seen.

[0044] In the present technique, the boundaries or edges of the pectoral muscle are identified. The identification of the edges of the pectoral muscle can be performed automatically by using computer-aided detection (CAD). The CAD system can analyze the projection frames to identify anatomical features within the projection frames, such as the pectoral muscle boundaries. Such identification can be based on changes between pixel values within the projection frames. For example, a particular pattern of pixel intensities can indicate the pectoral muscle boundary, and the pattern of pixel intensities allows the CAD system to identify the boundary. Once the pectoral muscle boundary is identified, a representation of the boundary is generated. The representation of the boundary can be a curve indicating the position of the pectoral muscle, and the curve can be based on the detected positions or points of the pectoral muscle.

[0045] Figure 2B Depicts Figure 2A a plurality of projection frames, where a representation of the pectoral muscle boundary is present. The boundary representation is indicated by the white curves in each projection frame. The first MLO projection frame 202 includes a first boundary representation 212 of the pectoral muscle boundary identified in the first MLO projection frame 202. The second MLO projection frame 204 includes a second boundary representation 214 for the pectoral muscle boundary identified in the second MLO projection frame 206. The third MLO projection frame 206 includes a third boundary representation 216 of the pectoral muscle boundary identified in the third MLO projection frame 206. The boundary representations 212 - 216 follow the boundary of the pectoral muscle, which extends from the upper boundary of the respective projection frames 202 - 206 to approximately the middle or lower half of the chest wall. The pectoral muscle boundary crosses most of the internal breast tissue within the projection frames. Thus, the movement of the pectoral muscle boundary across the acquired projection frames is a good approximation of the movement of the internal breast tissue.

[0046] In some examples, the generated boundary representation may not be displayed. Rather, the boundary representation can be a curve defined by a set of points within a function or image. For example, the boundary representation can be a mathematical or plot-based representation that can be used to perform the calculations discussed herein. The representation can be shown in a plot, but not shown as an overlay on the projection frame. The boundary representation is generally a two-dimensional representation based on two-dimensional image data in the projection frame. However, in some examples, a three-dimensional boundary representation can be generated if three-dimensional image data is available.

[0047] Figure 3A Example drawing 300A depicts boundary representations 301-315 of the pectoral muscles during a tomosynthesis imaging procedure in which no patient movement occurred. The example drawing includes boundary representations 301-315 of the pectoral muscles for an example MLO tomosynthesis imaging procedure, where the projection frames were acquired at fifteen different angular positions. Each of the pectoral muscle boundary representations 301-315 is from one of the corresponding projection frames acquired during that example imaging procedure.

[0048] As discussed above with respect to Figure 1 Since the x-ray source moves around an arc, specific features of the patient appear at different positions on the receiver and thus in different positions in the resulting image. Thus, as can be seen from Figure 3AAs can be seen, the boundary positions and their corresponding representations appear at different positions in each projection frame, even when the breast does not move during the imaging procedure. In the case where the breast does not move during the imaging procedure, based on the position of the first boundary muscle representation 301, the positions of each of the other pectoral muscle boundary representations 302 - 315 can be predicted through mathematical or geometric calculations and / or derivations. For example, for each angular position of the x-ray source that emits x-rays, the x-ray path that passes through the breast from the x-ray source and reaches the detector is known. Using that known path and the initial position of the first boundary muscle representation 301, the positions of the remaining boundary representations 302 - 315 can be predicted. Although the first boundary muscle representation 301 is most often used as the basis or starting point for predicting the remaining boundaries 302 - 315, any of the other boundary representations 301 - 315 can be used as the starting point for predicting the remaining boundary representations 301 - 315. For example, the seventh boundary representation 307 can be used as the basis or starting point for predicting the positions of the remaining boundary representations 301 - 306, 308 - 315. The corresponding positions of the boundary representations 301 - 315 can be characterized or defined by the spacing or distance between the corresponding boundary representations 301 - 315. Even in the example where there is no movement, the spacing between each of the boundary representations 301 - 315 can be different. However, the differences in the spacing between each of the boundary representations 301 - 315 will be smooth and predictable. Thus, in the presence of breast movement, the spacing between at least two of the boundary representations 301 - 315 will be different from the predicted spacing or distance between the boundary representations 301 - 315.

[0049] Figure 3B Example drawing 300B depicts the pectoral muscle boundary representations 301 - 315, where patient movement occurred during the tomosynthesis imaging procedure. In particular, patient movement occurred between the time of acquisition of the eleventh projection frame and the time of acquisition of the twelfth projection frame during the tomosynthesis imaging procedure. Movement is identified based on the abnormal spacing between the eleventh boundary representation 311 and the twelfth boundary representation 312.

[0050] To determine a spacing anomaly and indicate movement, a distance (D) between each pair of boundary representations 301 - 315 can be determined. The distance (D) can be measured in a direction orthogonal to at least one of the boundary representations for which it is calculated. For example, a reference line can be generated that is substantially orthogonal or perpendicular to at least one of the boundary representations 301 - 315. Then the distance (D) can be calculated along that reference line. For each pair of boundary representations 301 - 315, the distance (D) can also be measured starting from the midpoint on the vertical axis. In some examples, the distance (D) is measured at multiple points along each of the boundary representations 301 - 315. The distance (D) can be measured along the boundary representations, and the minimum distance (D) can be used for further determination and calculation. In other examples, the area between each boundary representation can be calculated. The area can be calculated by a plotting or image analysis algorithm and / or the area can be calculated by computing the integral between two boundary representations. Then the calculated distance(s) (D) and / or the calculated area between the two boundary representations can be compared to a predicted value based on the predicted position of an ideal boundary representation without patient movement. If the calculated distance(s) (D) and / or the calculated area differ from the predicted value, then movement can be determined to have occurred. The difference between the determined distance(s) (D) and / or the calculated area and the predicted value can be used to calculate a movement score, and if the difference is large enough, then a movement warning can be generated.

[0051] The movement score can be based on the magnitude of the positional difference between a first boundary representation and a second generated boundary representation. For example, a higher movement score can be generated in cases where the determined distance (D) is large and / or the determined area between the first and second boundary representations is large. Additionally, the movement score can be based on the differences between each pair of boundary representations used for a tomosynthesis imaging protocol. For example, in the example depicted in Figure 3B the differences in the form of distance (D) and / or determined area can be calculated or determined for the following fourteen pairs of representations: (1) boundary representation 301 and boundary representation 302; (2) boundary representation 302 and boundary representation 303; (3) boundary representation 303 and boundary representation 304; (4) boundary representation 304 and boundary representation 305; (5) boundary representation 305 and boundary representation 306; (6) boundary representation 306 and boundary representation 307;

[0052] (7)Boundary representations 307 and 308; (8)Boundary representations 308 and 309; (9)Boundary representations 309 and 310; (10)Boundary representations 310 and 311; (11)Boundary representations 311 and 312; (12)Boundary representations 312 and 313; (13)Boundary representations 313 and 314; and (14)Boundary representations 314 and 315. The motion score can then be based on the sum of the absolute values of the determined differences between the fourteen pairs of boundary representations. The motion score can also be based on the average of the determined differences. Additionally, the motion score can also be based on the single largest difference. For example, the largest determined difference for a pair of boundary representations can be used for or as the motion score. In some examples, fewer than all pairs of boundary representations can be analyzed and / or used when generating the motion score.

[0053] Figure 3C A plot 318 depicts the distances between the pectoral muscle boundary representations identified in the projection frames. As discussed above, a distance (D) can be calculated for each pair of boundary representations. The y-axis of plot 318 represents the distance (D) and the x-axis represents the projection frame pair number. The units of the y-axis can be pixels, although other units can be used and / or the axis can be normalized or unitless. Plot 318 shows those calculated distances (D) for 14 pairs of boundary representations, such as the fourteen pairs described above. For example, in plot 318, there are data points shown for a first distance 321, a second distance 322, a third distance 323, a fourth distance 324, a fifth distance 325, a sixth difference 326, a seventh distance 327, an eighth distance 328, a ninth distance 329, a tenth distance 330, an eleventh distance 331, a twelfth distance 332, a thirteenth distance 333, and a fourteenth distance 334. A curve 320 can be fit to the plotted data points representing the distance (D). Curve 320 can be a polynomial curve, such as a second-order polynomial curve. Curve 320 can also be generated by other interpolation and / or regression methods. Curve 320 can also be generated as a line representing the average (median or mean) of all measured distances (D). Curve 320 can also represent the expected distance value for each projection frame. Thus, an expected distance can be determined from a series of projection frames. Determining the expected distance based on the fit curve allows the expected distance to be determined even when the geometry of the imaging system is unknown or not used for such calculations. In some examples, both the fit curve 320 and the geometry of the system can be used to determine the expected distance.

[0054] The distance or difference between the data points representing the distance (D) for the projection frames and the fitted curve 320 is referred to as the shift variance and is denoted by 'S' in plot 318. The shift variance (S) represents the measured distance (D) for the corresponding data points and the expected value of the distance (D). As an example, the shift variance (S) represented in plot 318 is the distance between the data point for the eleventh distance 331 and curve 320. The shift variance (S) between the corresponding data point and curve 320 can be calculated based on the measured distance (D) for the projection frame pair number and the distance (D) of curve 320 at the projection frame pair number. For example, the shift variance (S) for projection pair number eleven represented at data point 331 is approximately 2 pixels and is based on the measured distance (D) of eight pixels at projection pair number eleven and the expected value of six pixels based on the position of curve 320 at projection pair number eleven (e.g., the y coordinate of curve 320 is six pixels at the x coordinate of eleven in plot 318). A pixel is generally equal to approximately 0.140 mm. In some examples, the shift variance (S) between the corresponding data point and curve 320 can also be calculated based on the normal to curve 320 from the corresponding data point or other minimization algorithms. If there is no patient motion during the imaging procedure, then the data points will overlap curve 320 and the shift variance (S) value will be zero or close to zero. Thus, if the shift variance (S) for any data point is greater than a predetermined threshold, then internal motion of the breast is likely to have occurred during the imaging procedure. The visual representation of the data points and curve 320 also provides insight into the amount of patient motion that may have occurred and the type of patient motion that may have occurred.

[0055] A motion score can be generated from one or more shift variance (S) values. For example, in the case of determining large shift variance (S) values, a high motion score can be generated. Additionally, if there are multiple calculated large shift variance (S) values (e.g., large shift variance (S) values for multiple data points), then a high motion score can be calculated. Conversely, in the case of small shift variance (S) values, the motion score can also be small.

[0056] Figure 3D Another example plot 300D depicts the pectoral muscle boundary representations 301 - 315 during which patient motion occurred during a tomosynthesis imaging procedure. Similar to Figure 3A-3B plots 300A - B depicted therein, the pectoral muscle boundary representations 301 - 315 are from an example MLO tomosynthesis imaging procedure where projection frames were acquired at fifteen different angular positions. Each of the pectoral muscle boundary representations 301 - 315 is from one of the corresponding projection frames acquired during that example imaging procedure.

[0057] The drawing 300D also includes a reference line 335. The reference line 335 is a line that is substantially perpendicular to the boundary representations 301-315. Each of the boundary representations 301-315 intersects the reference line 335. Each intersection point is indicated by a point in the drawing. The intersection distance (I) between each intersection point and another reference point along the reference line can be determined. As an example, the intersection point of the eleventh boundary representation 311 from the eleventh projection frame with the reference line 335 can be used as a reference point. The boundary representation 311 is bolded in the drawing 300D as a visual identifier for such an example. Any other point along the reference line 335, even a non-intersection point, can also be used as a reference point. The intersection distance (I) is the distance along the reference line from the corresponding intersection point of the boundary representation to the reference point.

[0058] Figure 3E An example intersection distance (I) measurement or calculation is depicted. Continue Figure 3D In the example of the drawing 300D in, the intersection point between the boundary representation 311 and the reference line 335 has been selected as the reference point. Thus, the intersection distance (I) is determined for each boundary representation from the selected reference point. Since there are fifteen projection frames in the example imaging protocol, fifteen intersection distances (I) can be calculated or measured. For example, the following intersection distances (I) can be calculated or measured: (1) the intersection distance (I1) between the first boundary representation 301 and the reference point; (2) the intersection distance (I2) between the second boundary representation 302 and the reference point; (3) the intersection distance (I3) between the third boundary representation 303 and the reference point; (4) the intersection distance (I4) between the fourth boundary representation 304 and the reference point; (5) the intersection distance (I5) between the fifth boundary representation 305 and the reference point;

[0059] (6) the intersection distance (I6) between the sixth boundary representation 306 and the reference point; (7) the intersection distance (I7) between the seventh boundary representation 307 and the reference point; (8) the intersection distance (I8) between the eighth boundary representation 308 and the reference point; (9) the intersection distance (I9) between the ninth boundary representation 309 and the reference point; (10) the intersection distance (I10) between the tenth boundary representation 310 and the reference point;

[0060] (11) the intersection distance (I11) between the eleventh boundary representation 311 and the reference point; (12) the intersection distance (I12) between the twelfth boundary representation 312 and the reference point; (13) the intersection distance (I13) between the thirteenth boundary representation 313 and the reference point; (14) the intersection distance (I14) between the fourteenth boundary representation 314 and the reference point; and (15) the intersection distance (I15) between the fifteenth boundary representation 315 and the reference point. It is noted that, Figure 3EThe intersection distance I11 is not depicted because the reference point in this example has been chosen to be the intersection of the eleventh boundary representation with the reference line 335. Thus, in this example, the intersection distance I11 is zero. The intersection distance (I) can be used to determine whether patient motion has occurred during the imaging procedure, as discussed below. In an example where a small number of projection frames are captured during the imaging procedure, a small number of intersection distances (I) are calculated or measured. Similarly, in an example where more projection frames are captured during the imaging procedure, more intersection distances (I) can be calculated or measured.

[0061] Figure 3F Example drawing 340 depicting the intersection distance (I) of the boundary representations 301 - 315 with the reference line 335. In example drawing 340, fifteen data points are plotted representing the intersection distances (I) for fifteen intersections. The y-axis of drawing 340 represents the intercept (I) values and the x-axis of the drawing represents the projection frame or boundary representation number. The units of the y-axis can be pixels, although other units can be used and / or the axis can be normalized or unitless. Due to breast motion during the imaging procedure, the intersection distances are non-uniform and the data points in the drawing do not follow a straight line. Curve 342 can be fit to the data points. Curve 342 can be a polynomial curve, such as a second-order polynomial. Curve 342 can also be generated by other interpolation and / or regression methods. Curve 342 can represent the expected intersection distance (I) for each intersection.

[0062] The distance or difference between the data points representing the intersection distance (I) for the projection frames and the fit curve 342 is referred to as the intersection shift variance and is denoted by "IS" in drawing 340. In the intersection, the shift variance (IS) can be calculated similarly to the shift variance (S) discussed above. For example, the intersection shift variance (IS) between the corresponding data point and curve 342 can be calculated based on the intersection distance (I) measured for the intersection and the distance (I) of curve 342 at the intersection. If there is no patient motion during the imaging procedure, then the data points will overlap with curve 342 and the intersection shift variance (IS) value will be zero or close to zero. Thus, if the intersection shift variance (IS) for any data point is greater than a predetermined threshold, then internal motion of the breast is likely to have occurred during the imaging procedure. The visual representation of the data points and curve 342 also provides insight into the amount of patient motion that may have occurred and the type of patient motion that may have occurred.

[0063] The motion score can be generated from one or more intersecting shift variance (IS) values. For example, in the case of determining a large intersecting shift variance (IS) value, a high motion score can be generated. Additionally, if there are multiple large intersecting shift variance (IS) values calculated (e.g., large intersecting shift variance (IS) values for multiple data points), then a high motion score can be calculated. Conversely, in the case of a small intersecting shift variance (IS) value, the motion score will also be small.

[0064] Figure 4A An example method 400 for approximating or identifying the motion of internal breast tissue during a tomosynthesis procedure is depicted. At operation 402, the patient's breast is compressed in the MLO position. As discussed above, this position is often approximately 45 degrees from the vertical direction, but in some examples and depending on the needs of a particular patient, it can be between approximately 30 - 60 degrees. In some examples, compressing the breast includes placing the breast on a breast platform and moving a compression paddle towards the breast platform until the breast is compressed therebetween. During the compression of the breast, at operation 404, a first tomosynthesis MLO projection frame is acquired. The first tomosynthesis MLO projection frame can be for a first angle relative to the breast. For example, the first tomosynthesis MLO projection frame can be acquired by x-ray radiation emitted from an x-ray source along an arc at the first angle, as Figure 1 depicted. At operation 406, during the compression of the breast, a second tomosynthesis MLO projection frame is acquired. The second tomosynthesis MLO projection frame can be for a second angle relative to the breast. For example, the second tomosynthesis MLO projection frame can be acquired by x-ray radiation emitted from an x-ray source along an arc at the second angle, as Figure 1 depicted. Operations 404 and 406 can be performed during the same compression of the breast. That is, the breast is continuously compressed while the projection frames are acquired. Although not depicted in method 400, method 400 can also include acquiring additional tomosynthesis MLO projection frames for additional angles.

[0065] After acquiring the MLO projection frames, the boundaries of the pectoral muscle are identified. At operation 408, the first boundary of the pectoral muscle in the first MLO projection frame is identified. As discussed above, the first boundary of the pectoral muscle in the first projection frame can be identified by using CAD techniques. At operation 410, a first representation of the identified first boundary of the pectoral muscle is generated. The generated first boundary representation can be one of the boundary representations depicted above Figure 2B and / or Figure 3A-3B 、in 3D. At operation 412, the second boundary of the pectoral muscle in the second MLO projection frame is identified. As discussed above, the second boundary of the pectoral muscle in the projection frame can be identified by using CAD techniques. At operation 414, a second representation of the identified second boundary of the pectoral muscle is generated. The generated second boundary representation can be one of the aboveFigure 2B and / or Figure 3A-3B one of the boundary representations depicted in 3D. As an example, the first generated boundary representation can be Figure 3A-3B and / or boundary representation 301 in 3D, and the second generated boundary representation can be Figure 3A-3B and / or boundary representation 302 in 3D. Any other combination from Figure 3A -B and / or boundary representations 301 - 315 in 3D can also be the first and second generated boundary representations.

[0066] At operation 416, the difference between the first boundary representation and the second boundary representation is determined. The difference can be the difference in position between the first boundary representation and the second boundary representation in the corresponding projection frame. For example, the difference can be the distance (D) discussed above and depicted in Figure 3B . The difference can also be the area between the two generated boundary representations, as discussed above. The difference can also be the minimum distance between the first boundary representation and the second boundary representation.

[0067] At operation 418, a motion score is generated based at least on the difference between the first generated boundary representation and the second generated boundary representation. The motion score can be based on the magnitude of the position difference between the first generated boundary representation and the second generated boundary representation. For example, a higher motion score can be generated if the determined distance (D) is large and / or the determined area between the first generated boundary representation and the second generated boundary representation is large. Additionally, the motion score can be based on the difference between each pair of generated boundary representations for a tomosynthesis imaging protocol. The motion score can be based on the aggregation of the absolute values of the determined differences between possible pairs of boundary representations. The motion score can also be based on the average of the determined differences. Additionally, the motion score can also be based on the single largest difference. For example, the largest determined difference for a pair of boundary representations can be used as or for the motion score. The motion score can be automatically used to adjust or dispose of the projection frames most affected by patient motion. For example, if a subset of projection frames exhibits motion, then image reconstruction can be performed without the subset of projection frames not affected by motion, or image reconstruction can be performed on all projection frames after applying corrections to the subset of affected projection frames. This motion score-based processing can include appropriate global and local adjustments, transformations, and shifting back to correct the amount of motion. Additionally, the motion score can be used to prompt and perform filtering to suppress high-frequency components to prevent any final image from being contaminated (blurred), while passing low-frequency components to improve the signal-to-noise ratio of the final image. The motion score can also be compared to a predetermined threshold, and if the motion score is greater than the predetermined threshold, then a motion warning can be generated.

[0068] Figure 4BDepicts another example method 420 for approximating or identifying the motion of internal breast tissue during a tomosynthesis procedure. At operation 422, the patient's breast is compressed in the MLO position. The compression can be the same as operation 402 in method 400 depicted in Figure 4A At operation 424, a plurality of tomosynthesis MLO projection frames are acquired during the compression of the breast. For example, while the patient's breast is in a compressed state, during a tomosynthesis imaging procedure, a series of projection frames are acquired as the x-ray source moves around an arc, as depicted in Figure 1 .

[0069] At operation 426, the boundaries of the pectoralis muscle are identified in at least two of the projection frames acquired in operation 424. As discussed above, the boundaries of the pectoralis muscle in the projection frames can be identified by using CAD techniques. At operation 428, a representation of the identified boundaries of the pectoralis muscle is generated for each of the at least two projection frames for which the boundaries of the pectoralis muscle have been identified. The boundary representation generated in operation 428 can be any of the boundary representations discussed above.

[0070] At operation 430, a first difference between the generated boundary representations is determined. For example, the first difference can be any positional difference between two boundary representations discussed above, such as the distance (D) between two boundary representations and / or the area between two boundary representations. At operation 432, a difference between the first difference and an expected value of the first difference is determined. The expected value can be based on a curve fitted to a plurality of differences calculated for pairs of projection frames, such as the curve 320 depicted in Figure 3C . In this example, the difference between the first difference and the expected value is the shift variance (S) for the first pair of projection frames. Additionally, as discussed above, in the case where there is no motion of the breast during the imaging procedure, the positions of the various boundary representations are predictable. Therefore, the difference between two boundary representations is also predictable and determinable. Based on the predicted positions of the boundary representations, the expected value of the difference between any two boundary representations can be determined based on the geometry of the imaging system. Thus, in some examples, the expected value can be calculated based in part on the x-ray angle of the source for each corresponding projection frame.

[0071] If the first difference determined at operation 430 is different from the expected value of the first difference, then motion is likely to have occurred during the time between the acquisition of the two corresponding projection frames. The magnitude of the difference determined at operation 432 (e.g., the shift variance (S)) generally indicates the amount of motion that has occurred between the two projection frames for which the boundary representations were generated and used for the calculation and determination.

[0072] At operation 434, the difference determined in operation 432 (e.g., shift variance (S)) is compared with a predetermined threshold. The predetermined threshold can be a threshold for the amount of acceptable movement. For example, in some cases, a small amount of movement during an imaging protocol can be acceptable. Thus, the predetermined threshold can be set to an amount of movement that does not result in a degradation of image quality and / or that still results in clinically acceptable reconstructed and tomosynthesis slices. At operation 436, a motion warning can be generated based on the comparison performed at operation 434. For example, if the difference between the expected value and the first value determined at operation 432 is greater than the predetermined threshold, then a motion warning can be generated. The motion warning can indicate to the reviewer that internal breast tissue movement has occurred during the tomosynthesis imaging protocol. The warning can also indicate between which projection frames the movement occurred and the severity of the movement. The motion warning can also be an audible warning, such as a sound emitted, to alert that movement has occurred during the imaging protocol. When a motion warning is provided, the technician can then immediately re-image the patient, which prevents the patient from having to return to the imaging facility at a later time.

[0073] Figure 4C Another example method 440 for approximating or identifying movement of internal breast tissue during a tomosynthesis protocol is depicted. At operation 442, the patient's breast is compressed, and at operation 444, a plurality of tomosynthesis projection frames are acquired during the compression of the breast. For example, fifteen projection frames can be acquired. In some examples, at least three projection frames are acquired. Operations 442 and 444 can be the same as operations 422 and 424 of method 420 depicted in Figure 4B At operation 446, the boundaries of the pectoral muscles are identified in the plurality of projection frames acquired in operation 444. In some examples, the boundaries of the pectoral muscles can be identified for each projection frame or a subset thereof in the plurality of projection frames. As discussed above, the boundaries of the pectoral muscles in the projection frames can be identified by using CAD techniques. At operation 448, a representation is generated for each boundary identified in operation 448. The boundary representation generated in operation 448 can be any of the boundary representations discussed above. In some examples, boundary representations can be generated for less than all of the projection frames acquired in operation 444. For example, boundary representations can be generated for at least a subset of the boundaries identified in operation 446.

[0074] At operation 450, the distance (D) between each pair of boundary representations generated in operation 448 is measured or calculated. The distance (D) can be the distance (D) discussed above and / or Figure 3B depicted in

[0075] At operation 452, an expected distance value is determined based on the measured distance (D) of operation 450. Determining the expected distance value can include curve fitting to the measured distance, such as the curve 320 depicted above and in Figure 3C . In some examples, the measured distances (D) can be co-plotted and the fitted curve can be a polynomial curve, such as a second-order polynomial curve. The fitted curve can also be generated using any other technique discussed above. The fitted curve can then be used to generate an expected distance value for each boundary representation pair and / or projection frame pair.

[0076] At operation 454, a shift variance (S) value is determined for each boundary representation pair for which a distance (D) was measured or calculated in operation 452. The shift variance (S) value is the difference between the measured distance (D) of the boundary representation pair and the expected distance value of the boundary representation pair. In some examples, the shift variance (S) values are calculated for less than all possible boundary representation pairs for which a distance (D) was measured or calculated. For example, the shift variance (S) can be calculated for at least a subset of the boundary representation pairs for which a distance (D) was measured or calculated. The shift variance (S) value is the difference between the measured distance (D) of the boundary representation pair and the expected distance value of the boundary representation pair.

[0077] At operation 456, patient motion during the imaging protocol is identified based on the shift variance (S) determined in operation 454. Identification of the motion can be based on comparing the shift variance (S) to a predetermined threshold. If any shift variance (S) is greater than the predetermined threshold, then patient motion can be determined to have occurred. The average value of the shift variance (S) values can also be compared to the predetermined threshold to determine if patient motion has occurred. If patient motion is identified at operation 456, then a motion warning can be generated. The motion warning can indicate to a reviewer that internal breast tissue motion has occurred during the tomosynthesis imaging protocol. The warning can also indicate between which projection frames the motion has occurred based on which boundary representation pair produced a large shift variance (S) value. The motion warning can also include an indication of the motion severity based on the magnitude of the shift variance (S) value and / or the magnitude of the difference between the shift variance (S) value and the predetermined threshold. The motion warning can also be an audible warning, such as a sound emitted, to alert that motion has occurred during the imaging protocol. A motion score can also be generated from one or more of the shift variance (S) values. For example, a high motion score can be generated in the case of determining a large shift variance (S) value. Additionally, if there are multiple calculated large shift variance (S) values (e.g., large shift variance (S) values for multiple data points), then a high motion score can be calculated. Conversely, in the case of small shift variance (S) values, the motion score will also be small.

[0078] Figure 4DDepicts another example method 460 for approximating or identifying the movement of internal breast tissue during a tomosynthesis procedure. At operation 462, the patient's breast is compressed, and at operation 464, multiple tomosynthesis projection frames are acquired during the compression of the breast. For example, fifteen projection frames can be acquired. In some examples, at least three projection frames are acquired. Operations 462 and 464 can be the same as operations 442 and 444 of method 440 depicted in Figure 4C . At operation 466, the boundaries of the pectoral muscle are identified in the multiple projection frames acquired in operation 464, and at operation 468, representations are generated for those boundaries. Operations 462 - 468 can be the same as operations 442 - 448 of method 440 depicted in Figure 4C .

[0079] At operation 470, a reference line or curve that intersects the boundary representation is generated. The reference line can be the reference line 335 depicted in Figure 3D . Operation 470 can include co - drawing the boundary representation generated in operation 468. In some examples, less than all of the generated boundary representations can be used. For example, at least a subset of the boundary representations generated in operation 468 can be drawn and used for other calculations or measurements in method 460. Along with the co - drawn boundary representations, a reference line can be generated such that the reference line intersects each of the co - drawn boundary representations. In some examples, the reference line can be perpendicular or approximately perpendicular to one or more of the boundary representations. At operation 472, reference points are identified or selected along the reference line. The reference point can be any point along the reference line. In some examples, the reference point can be selected as one of the points where the reference line intersects the boundary representation.

[0080] At operation 474, the intersection distance (I) is calculated or measured. The intersection distance (I) is the distance between the reference point and the intersection point of the boundary representation and the reference line, as discussed above with reference to Figure 3D-3E . Thus, operation 474 can also include identifying the intersection point of the boundary representation and the reference line. The intersection distance (I) can be measured or calculated for each boundary representation or at least a subset of the boundary representations.

[0081] At operation 476, an expected intersection distance (I) value is determined. The expected intersection distance can be determined based on the measured intersection distance (I) measured in operation 474. For example, the measured intersection distance (I) can be co - drawn as data points and a curve can be fit to the data points, as discussed above with reference to Figure 3F . For example, the curve 342 depicted in Figure 3F can be generated based on the intersection distance calculated in operation 474. Then the curve can be used to generate or calculate the expected intersection distance value.

[0082] At operation 478, an intersection shift variance (IS) can be calculated for each of or at least a subset of the boundary representations. The intersection shift variance (IS) is the difference between the measured intersection distance (I) of the boundary representation (measured at operation 474) and the expected value of the intersection distance (determined at operation 476). Any of the methods or processes discussed above can be used to determine or calculate the intersection shift variance (IS). For example, the intersection shift variance (IS) can be calculated as the difference or distance between the respective data points representing the measured intersection distance (I) and a curve (such as curve 342 depicted in Figure 3F ).

[0083] At operation 480, patient motion during an imaging protocol is identified based on the intersection shift variance (IS) value determined at operation 478. The identification of motion can be based on comparing the intersection shift variance (IS) value to a predetermined threshold. If any intersection shift variance (IS) value is greater than the predetermined threshold, then it can be determined that patient motion has occurred. The average value of the intersection shift variance (IS) values can also be compared to the predetermined threshold to determine if patient motion has occurred. If patient motion is identified at operation 480, then a motion warning can be generated. The motion warning can indicate to a reviewer that internal breast tissue motion has occurred during a tomosynthesis imaging protocol. The warning can also indicate between which projection frames motion has occurred based on which boundary representation produced a large intersection shift variance (IS) value. The motion warning can also include an indication of the severity of the motion based on the magnitude of the intersection shift variance (IS) value and / or the magnitude of the difference between the intersection shift variance (IS) value and the predetermined threshold. The motion warning can also be an audible warning, such as a sound emitted, to alert that motion has occurred during the imaging protocol. A motion score can also be generated from one or more of the intersection shift variance (IS) values. For example, in the case of determining a large intersection shift variance (IS) value, a high motion score can be generated. Additionally, if multiple large intersection shift variance (IS) values are calculated (e.g., large intersection shift variance (IS) values for multiple data points), then a high motion score can be calculated. Conversely, in the case of small intersection shift variance (IS) values, the motion score will also be small.

[0084] Figure 5ADepicts an example medical image 502 of a breast with multiple motion guides 504. The medical image 502 can be a tomosynthesis projection frame acquired during a tomosynthesis imaging procedure. In the depicted example, the motion guides 504 are multiple parallel vertical lines that are evenly spaced from each other. The motion guides 504 provide a reference frame for the positions of the features of the breast. When two images are thus compared to each other, one can more easily discern whether the features of the breast have moved between the time the first image was captured and the time the second image was captured. The images can be compared to each other by concurrently or sequentially displaying the images, such as in a movie mode. In some examples, each individual motion guide 504 is selectable and configurable. For example, a motion guide 504 can be selected and dragged to a new position within the image. This feature is desirable in cases where a reviewing physician may want the motion guide 504 to be directly aligned with a feature or region of interest of the breast, such as the nipple or a lesion. The configurability of the motion guides 504 is also desirable when a reviewing physician wishes to move a particular motion guide 504 out of the field of view so that the motion guide 504 does not obscure the image of the breast itself. In some examples, each motion guide can be deleted and / or new motion guides can be added. The display of the motion guides can also be turned on and off by the reviewing physician via user interface features that are displayed concurrently with the medical image 502. Although the motion guides 504 are depicted as vertical, parallel, evenly spaced, and having fixed positions in the depicted example, in other examples the motion guides 504 can be non-parallel, non-vertical, and / or non-evenly spaced. The motion guides 504 can also have different positions. For example, the motion guides 504 can also include horizontal lines to provide a reference frame for vertical motion. Due to the compensation for the movement of the tube moving along an arc (as discussed above), the horizontal lines can gradually move downward for each increasing projection frame, which makes the breast in the projection frame appear to move downward.

[0085] Figure 5B Depicts an example series of projection frames 512 - 516 for a tomosynthesis imaging procedure of a stationary breast. More specifically, an example first projection frame 512, an example seventh projection frame 514, and an example fifteenth projection frame 516. The motion guides 504 are shown in each projection frame 512 - 516. As can be seen from the comparison of the projection frames 512 - 516, the breast appears to move downward during the tomosynthesis imaging procedure. However, this apparent movement is due to the movement of the X-ray tube along an arc, as discussed above with respect to Figure 1 But, during the example imaging procedure that produced the example projection frames 512 - 516, the breast remained stationary. The projection frames 512 - 516 can be displayed concurrently or sequentially, such as in a movie mode, where the projection frames 512 - 516 can be played as a video that includes the sequence of projection frames.

[0086] Figure 5C depicts example series of projection frames 522 - 526 during which breast motion occurs during a tomosynthesis imaging procedure. This series of projection frames 522 - 526 is substantially similar to the projection frames 512 - 516 depicted in Figure 5B , except that Figure 5B the breast in the projection frames 522 - 526 is in motion between the time of acquisition of the first projection frame 522 and the time of acquisition of the seventh projection frame 524. Due to the motion guide 504, the motion of the breast can be seen more easily, which is horizontal motion in the depicted example. As can be seen from the projection frames 522 - 526, the breast shifts across at least one of the motion guides. Additionally, when motion is detected or identified, such as by the methods or processes discussed herein, a motion indicator 506 can be displayed on one or more of the projection frames 522 - 526. The motion indicator 506 can indicate the direction of motion as well as the magnitude of the motion. For example, in an example where the motion indicator 506 is an arrow, the arrow points in the direction of motion, and the size or color of the arrow can indicate the magnitude of the motion. The magnitude of the motion can be based on a motion score. The motion indicator 506 can also be based on the motion of the breast compared to a previous projection frame. In some examples, the motion indicator 506 can be temporarily displayed and can be turned on and off. The motion indicator 506 can also be displayed in an area of the projection frame that does not overlap with the breast so as not to obscure the view of the breast. Although depicted as an arrow, the motion indicator 506 can be other visual indicators. For example, the motion indicator can be a representation of the motion score of the imaging procedure, which can be represented numerically, alphabetically, in different colors, or by other visual indications to indicate the magnitude of the motion score. The projection frames 522 - 526 can be displayed concurrently or sequentially, such as in a movie mode, where the projection frames 522 - 526 can be played as a video including a sequence of projection frames.

[0087] Figure 6Depicts an example method 600 for a motion guide to display medical images. At operation 602, a view of a projection frame with multiple motion guides is displayed on a display (such as a display at a workstation or a remote viewing station). The projected view can include displaying the projection frames concurrently or sequentially, such as in a cine mode. At operation 604, a selection of one of the motion guides is received. The selection can be made through any input mechanism, such as via a mouse, a trackball, or touch input. At operation 606, an input indicating that the selected motion guide is to be moved to a new position is received. The input can be in the form of a drag motion or other input means for indicating the new position of the action guide. Based on receiving the input to move the selected motion guide, at operation 608 the selected motion guide is displayed at the new position. Movement of one or more of the motion guides may be desired in cases where a motion guide obscures a portion of the breast or where a reviewing physician desires to align one or more of the motion guides with a particular anatomical structure or other landmark of the breast. When a motion guide has been moved, the moved motion guide can appear at the new position in each projection frame. Thus, during the cine mode, the motion guide does not change position as the projection frames are displayed.

[0088] At operation 610, it is determined that breast motion occurs between at least two of the projection frames being displayed or to be displayed. The determination of the occurring breast motion can be performed by any of the techniques discussed herein. The determination of the breast motion in operation 610 can also include determination of the direction and / or magnitude of the motion. Based on the breast motion determined in operation 610, at operation 612 a motion indicator is displayed. The motion indicator can be Figure 5B the example motion indicator 506 depicted in

[0089] Figure 7 Depicts an example of a suitable tomosynthesis system 700 in which one or more of the present embodiments can be implemented. The tomosynthesis system 700 includes a gantry 702 and a workstation 704 in communication with the gantry 702. The workstation can include a display for displaying an indicator 706, such as a motion warning or other information. The display of the workstation 704 can also be used to display and review projection frames.

[0090] Figure 8illustrates an example of a suitable operating environment 800 in which one or more of the present embodiments may be implemented. The example operating environment may be incorporated in a workstation 704 or other computing device for reviewing medical images such as projection frames. In its most basic configuration, the operating environment 800 generally includes at least one processing unit 802 and a memory 804. The processing unit may be a processor, which is hardware. Depending on the exact configuration and type of the computing device, the memory 804 (which stores instructions for performing the motion detection techniques disclosed herein) may be volatile (such as RAM), non-volatile (such as ROM, flash memory, etc.), or some combination of both. This most basic configuration is illustrated by the dashed line 806 in Figure 8 which. The memory 804 stores instructions that, when executed by the processing unit(s) 802, perform the methods and operations described herein. Additionally, the environment 800 may also include storage devices (removable, 808, and / or non-removable, 810), including but not limited to magnetic or optical disks or tapes. Similarly, the environment 800 may also have one or more input devices 814, such as a keyboard, mouse, pen, voice input, etc., and / or one or more output devices 816, such as a display, speakers, printer, etc. Also included in the environment may be one or more communication connections 812, such as a LAN, WAN, point-to-point, etc. In an embodiment, the connection may be operable to enable point-to-point communication, connection-oriented communication, connectionless communication, etc.

[0091] The operating environment 800 generally includes at least some form of computer-readable medium. Computer-readable media can be any available media that can be accessed by the processing unit 802 or other devices including the operating environment. By way of example and not limitation, computer-readable media may include computer storage media and communication media. Computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disk (DVD) or other optical storage, magnetic cassettes, tapes, disk storage or other magnetic storage devices, or any other non-transitory media that can be used to store the desired information. Computer storage media is non-transitory and does not include communication media.

[0092] A communication medium implements computer-readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and includes any information delivery medium. The term "modulated data signal" refers to a signal having one or more characteristics set or changed in such a manner as to encode information as the signal. By way of example, and not limitation, communication media include wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, microwave, and other wireless media. Combinations of any of the foregoing are also intended to be included within the scope of computer-readable media.

[0093] The operating environment 800 can be a single computer operating in a network environment using a logical connection to one or more remote computers. The remote computers can be personal computers, servers, routers, network PCs, peer devices, or other common network nodes, and typically include many or all of the elements described above as well as other elements not recited. The logical connection can include any method supported by an available communication medium. Such networking environments are often used in medical offices, enterprise-wide computer networks, intranets, and the Internet.

[0094] Embodiments described herein can be implemented and executed using software, hardware, or a combination of software and hardware to implement and execute the systems and methods disclosed herein. Although specific devices have been described throughout the disclosure as performing specific functions, those skilled in the art will recognize that these devices are provided for illustrative purposes and that other devices can be employed to perform the functions disclosed herein without departing from the scope of the disclosure. Additionally, some aspects of the disclosure have been described above with reference to block diagrams and / or operational descriptions of systems and methods according to aspects of the disclosure. The functions, operations, and / or actions noted in the blocks may not occur in the order shown in any corresponding flowchart. For example, two blocks shown in succession may in fact be executed or performed substantially concurrently or in reverse order, depending upon the functionality involved and the implementation.

[0095] The present disclosure describes some embodiments of the present technology with reference to the accompanying drawings, in which only some possible embodiments are shown. For example, although the present technology is mainly discussed with reference to the pectoral muscles, the technology can also be applied to other internal features of the breast with distinguishable or recognizable boundaries, such as implants or chest wall muscles in the image. However, other aspects may be implemented in many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to make the present disclosure thorough and complete and to fully convey the scope of possible embodiments to those skilled in the art. Additionally, as used herein and in the claims, the phrase "at least one of element A, element B, or element C" is intended to convey any one of the following: element A, element B, element C, element A and B, element A and C, element B and C, and element A, B, and C. Additionally, those skilled in the art will understand that terms such as "about" or "substantially" convey a degree that depends on the measurement techniques used herein. To the extent that those skilled in the art may not be able to clearly define or understand these terms, the term "about" shall mean plus or minus ten percent.

[0096] Although specific embodiments are described herein, the scope of the present technology is not limited to those specific embodiments. Those skilled in the art will recognize other embodiments or improvements within the scope and spirit of the present technology. Accordingly, specific structures, acts, or media are disclosed only as illustrative embodiments. The scope of the technology is defined by the appended claims and any equivalents thereof.

Claims

1. A method for identifying internal movement of a patient's breast during an imaging procedure, the method comprising: Compressing the patient's breast in the mediolateral oblique (MLO) position; During the compression of the breast, acquiring a first tomosynthesis MLO projection frame for a first angle with respect to the breast; During the compression of the breast, acquiring a second tomosynthesis MLO projection frame for a second angle with respect to the breast; Identifying a first boundary of the pectoralis muscle in the first projection frame; Generating a first representation of the first boundary of the pectoralis muscle; Identifying a second boundary of the pectoralis muscle in the second projection frame; Generating a second representation of the second boundary of the pectoralis muscle; Determining a difference between the first representation and the second representation; and Generating a motion score based at least on the difference between the first representation and the second representation.

2. The method according to claim 1, wherein the generated first representation is a two-dimensional representation.

3. The method according to claim 1, wherein the difference is based on the area between the first representation and the second representation.

4. The method according to claim 1, wherein the difference is based on the minimum distance between the first representation and the second representation.

5. The method according to claim 1, further comprising: Comparing the difference with an expected value, wherein the expected value is based on at least one of: The x-ray angle of the x-ray source of the first projection frame and the x-ray angle of the x-ray source of the second projection frame, or A fitted curve based at least on the first tomosynthesis MLO projection frame and the second tomosynthesis MLO projection frame; and Generating a motion warning based on the comparison of the difference with the expected value.

6. The method according to claim 1, further comprising concurrently displaying at least a portion of the first projection frame and the second projection frame with a plurality of parallel motion guides in a cine view.

7. The method according to claim 6, further comprising: Receiving a selection of one of the plurality of parallel motion guides; Receiving an input to move the selected parallel motion guide to a new position; and And Displaying the selected parallel motion guide at the new position based on the received input to move the selected parallel motion guide.

8. The method according to claim 6, wherein the plurality of parallel motion guides are evenly spaced relative to each other.

9. A method for identifying internal movement of a patient's breast during an imaging procedure, the method comprising: Compressing the patient's breast in the mediolateral oblique (MLO) position; Acquiring a plurality of tomosynthesis MLO projection frames during the compression of the breast, wherein the plurality of tomosynthesis MLO projection frames include images of a portion of the patient's breast and a portion of the pectoralis muscle; For at least two of the plurality of tomosynthesis MLO projection frames, identifying the boundary of the pectoralis muscle; For the at least two of the plurality of tomosynthesis MLO projection frames, generating a representation of the boundary of the pectoralis muscle; For the at least two of the plurality of tomosynthesis MLO projection frames, determining a first difference between the generated representations; Determining a second difference between the first difference and an expected value of the first difference; Comparing the second difference with a predetermined threshold; And Generating a motion warning based on the comparison of the second difference with the predetermined threshold.

10. The method according to claim 9, wherein the generated representation is a two-dimensional representation.

11. The method according to claim 9, wherein the first difference is based on the area between the generated representations.

12. The method according to claim 9, wherein the first difference is based on the minimum distance between the generated representations.

13. The method according to claim 9, wherein the second difference is a shift variance value.

14. The method according to claim 9, further comprising continuously displaying at least a portion of the projected frames concurrently with a plurality of parallel motion guides in a cine view.

15. The method according to claim 14, further comprising: receiving a selection of one of the plurality of parallel motion guides; receiving an input to move the selected parallel motion guide to a new position; and displaying the selected parallel motion guide at the new position based on the received input to move the selected parallel motion guide.

16. The method according to claim 14, wherein the plurality of parallel motion guides are evenly spaced relative to each other.

17. A system for identifying internal motion of a patient's breast during an imaging procedure, the system comprising: an x-ray source configured to rotate around the breast; a compression paddle configured to compress the breast in a mediolateral oblique (MLO) position; an x-ray detector deployed opposite the compression paddle from the x-ray source; at least one processor; and a memory storing instructions that, when executed by the at least one processor, cause the system to perform a set of operations including: during compression of the breast in the MLO position: emitting a first x-ray emission from the x-ray source at a first angle relative to the breast; after the first x-ray emission passes through the breast, detecting the first x-ray emission from the x-ray source by the x-ray detector; emitting a second x-ray emission from the x-ray source at a second angle relative to the breast; and after the second x-ray emission passes through the breast, detecting the second x-ray emission by the x-ray detector; generating a first tomosynthesis MLO projection frame for the first angle based on the detected first x-ray emission; generating a second tomosynthesis MLO projection frame for the second angle based on the detected second x-ray emission; identifying a first boundary of the pectoral muscle in the first projection frame; generating a first representation of the first boundary of the pectoral muscle; identifying a second boundary of the pectoral muscle in the second projection frame; generating a second representation of the second boundary of the pectoral muscle; determining a difference between the first representation and the second representation; and generating a motion score based at least on the difference between the first representation and the second representation.

18. The system according to claim 17, wherein the generated first representation is a two-dimensional representation.

19. The system according to claim 17, wherein the difference is based on the area between the first representation and the second representation.

20. The system according to claim 17, wherein the difference is based on the minimum distance between the first representation and the second representation.

21. A method for identifying internal motion of a patient's breast during an imaging procedure, the method comprising: compressing the patient's breast in a mediolateral oblique (MLO) position; acquiring a plurality of tomosynthesis projection frames during compression of the breast, wherein the plurality of tomosynthesis projection frames include images of a portion of the patient's breast and a portion of the pectoral muscle; For at least a subset of the plurality of tomosynthesis projection frames, identify the boundaries of the pectoral muscles; For the identified boundaries of the pectoral muscles, generate a boundary representation for the identified boundaries of the pectoral muscles; Measure the distance between the generated boundary representations for at least a subset of all possible pairs of boundary representations; Determine an expected distance value for each boundary representation for which the distance is measured; Based on the measured distance and the expected distance value, determine a shift variance for each boundary pair for which the distance is measured; Compare the shift variance with a predetermined threshold; And Based on the comparison of the shift variance with the predetermined threshold, generate a motion warning.

22. A method for identifying internal motion of a patient's breast during an imaging procedure, the method comprising: Compress the patient's breast in the mediolateral oblique (MLO) position; Acquire a plurality of tomosynthesis projection frames during the compression of the breast, wherein the plurality of tomosynthesis projection frames include images of a portion of the patient's breast and a portion of the pectoral muscles; For at least a subset of the plurality of tomosynthesis projection frames, identify the boundaries of the pectoral muscles; For the identified boundaries of the pectoral muscles, generate a boundary representation for the identified boundaries of the pectoral muscles; Generate a reference line that intersects the generated boundary representation; Identify reference points along the reference line; For at least a subset of the generated boundary representations, calculate the intersection distance from the reference point to the intersection of the corresponding boundary and the reference line; Determine an expected intersection distance value based on the calculated intersection distance; Determine an intersection shift variance for each boundary representation for which the intersection distance is calculated; Compare the intersection shift variance with a predetermined threshold; And Based on the comparison of the intersection shift variance with the predetermined threshold, generate a motion warning.

Images