Method and apparatus for detecting information loss of medical image, and electronic device

Abstract

The application discloses a medical image information loss detection method and device and electronic equipment. The method comprises the following steps: obtaining structure image data and functional image data corresponding to each scanning layer in a same scanning sequence, wherein the structure image data and the functional image data are obtained by layer-by-layer scanning of a target object; determining a corresponding ratio sequence between the structure image data and the functional image data, wherein the ratio sequence comprises a plurality of target area ratios, each scanning layer corresponds to a target area ratio, and the target area ratio is used to represent a ratio of an effective signal area in the functional image data of the same scanning layer to a human mask area in the structure image data; determining an abnormality detection threshold corresponding to the ratio sequence, and analyzing the ratio sequence by using the abnormality detection threshold, so as to determine a scanning layer in which the functional image data has an information loss abnormality. The application solves the technical problem that the detection effect of related art on functional image information loss abnormality is poor.

Description

Technical Field

[0001] This application relates to the field of medical image quality assessment technology, and more specifically, to a method, device, and electronic device for detecting missing information in medical images. Background Technology

[0002] In PET (Positron Emission Tomography) / CT (Computed Tomography) fusion imaging, PET provides metabolic function information, while CT provides anatomical structure information. Both are typically acquired sequentially on the same device and spatially registered and fused. However, due to factors such as long PET scan times, patient movement, equipment hardware malfunctions, or data transmission interruptions, some PET slices often suffer from localized signal loss, incomplete acquisition, or data corruption.

[0003] Currently, clinical practice primarily relies on manual visual examination, where radiologists meticulously review PET images layer by layer to identify any abnormally low-signal or blank areas. Additionally, there are PET-based single-modality detection methods that identify low-signal regions by setting fixed thresholds or statistical distribution characteristics (such as mean and standard deviation). Some research also employs deep learning models, training neural networks to learn the feature distribution of normal PET images to detect abnormal slices deviating from the normal pattern. However, these related technologies are not effective in detecting abnormalities in functional images, exhibiting problems such as low detection efficiency, difficulty in accurately distinguishing between normal low-uptake areas and abnormal signal loss areas in functional images, and poor interpretability.

[0004] There is currently no effective solution to the above problems. Summary of the Invention

[0005] This application provides a method, apparatus, and electronic device for detecting missing information in medical images, so as to at least solve the technical problem that the related technologies have poor detection effects on functional image information loss anomalies.

[0006] According to one aspect of the embodiments of this application, a method for detecting missing information in medical images is provided, comprising: acquiring structural image data and functional image data corresponding to each scanning layer in the same scanning sequence, wherein the structural image data and functional image data are obtained by scanning a target object layer by layer; determining a corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence contains multiple target area ratios, each scanning layer corresponds to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scanning layer; determining an anomaly detection threshold corresponding to the ratio sequence, and using the anomaly detection threshold to analyze the ratio sequence, thereby determining the scanning layer in the functional image data where there is an information missing anomaly.

[0007] Optionally, after acquiring structural image data and functional image data in the same scanning sequence, the method further includes: determining the first number of slices contained in the structural image data and the second number of slices contained in the functional image data; determining the smaller value between the first number of slices and the second number of slices as the effective number of slices, wherein the effective number of slices is the number of common layers between the structural image data and the functional image data; and sorting each slice in the structural image data and each slice in the functional image data according to the Z-axis position based on the effective number of slices, so as to ensure that the structural image data and the functional image data correspond layer by layer in space, wherein the Z-axis is the scanning axis along the head-to-foot direction of the human body.

[0008] Optionally, after acquiring structural image data and functional image data in the same scan sequence, the method further includes: determining a first spatial resolution corresponding to the structural image data and a second spatial resolution corresponding to the functional image data; and using an interpolation algorithm to convert the structural image data of the first spatial resolution to the second spatial resolution, or to convert the functional image data of the second spatial resolution to the first spatial resolution, so that the spatial resolution of the structural image data and the spatial resolution of the functional image data are consistent.

[0009] Optionally, determining the corresponding ratio sequence between structural image data and functional image data includes: for each scan layer, determining the area of ​​a first region of the human body mask in the structural image data, and determining the area of ​​a second region of the effective signal region in the functional image data, wherein the human body mask is used to characterize the spatial range of human tissue in the scan layer, and the effective signal region is used to characterize the coverage range of metabolic signals collected in the functional image data of the scan layer; the ratio between the area of ​​the second region and the area of ​​the first region corresponding to each scan layer is determined as the target area ratio corresponding to the scan layer, thereby obtaining the ratio sequence.

[0010] Optionally, the structural image data includes at least one of the following: computed tomography (CT) images; determining the area of ​​the first region of the human mask in the structural image data includes: converting the pixel values ​​of the CT image into HU values, and determining the pixel regions with HU values ​​greater than a first preset threshold as foreground regions; removing non-human regions from the foreground regions, and performing morphological operations on the foreground regions after removing non-human regions, wherein the non-human regions include at least one of the following: the scanning bed region, and the morphological operations are used to fill the holes inside the foreground regions and eliminate isolated noise points; determining the largest connected region in the foreground regions after morphological operations as the human mask, and calculating the area of ​​the human mask to obtain the area of ​​the first region.

[0011] Optionally, the functional image data includes at least one of the following: a positron emission tomography (PET) image; determining the area of ​​the second region of the effective signal region in the functional image data includes: normalizing the pixel values ​​of the PET image, and performing threshold segmentation on the normalized PET image according to a second preset threshold to obtain a PET mask, wherein the second preset threshold is used to distinguish between background regions and potential human signal regions in the PET image; taking the intersection of the PET mask and the human mask to obtain the effective signal region, and calculating the area of ​​the effective signal region to obtain the area of ​​the second region.

[0012] Optionally, the anomaly detection threshold includes at least one of the following: a first anomaly detection threshold and a second anomaly detection threshold; determining the anomaly detection threshold corresponding to the ratio sequence includes: determining the arithmetic mean of the target area ratios in the ratio sequence, and determining the first anomaly detection threshold based on the arithmetic mean and a first preset adjustment coefficient, wherein the first preset adjustment coefficient is used to set the normal decay degree of the target area ratio relative to the arithmetic mean; determining the first quartile and the third quartile of the target area ratios in the ratio sequence, and determining the interquartile range based on the first quartile and the third quartile; determining the second anomaly detection threshold based on the first quartile, the interquartile range, and the second preset adjustment coefficient, wherein the second preset adjustment coefficient is used to set the leniency of the anomaly judgment based on the interquartile range, and the second anomaly detection threshold is used for outliers in the ratio sequence that deviate from the normal range.

[0013] Optionally, the analysis of the anomaly detection threshold comparison value sequence to determine the scan layer with missing information in the functional image data includes: determining that the functional image data of the scan layer corresponding to the target area ratio has missing information when the target area ratio in the ratio sequence is less than a first anomaly detection threshold and / or the target area ratio is less than a second anomaly detection threshold.

[0014] According to another aspect of the embodiments of this application, a medical image information missing detection device is also provided, comprising: a data acquisition module, used to acquire structural image data and functional image data in the same scanning sequence, wherein the scanning sequence includes multiple scanning layers, and the structural image data and functional image data are obtained by scanning a target object layer by layer; a ratio calculation module, used to determine the corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence includes multiple target area ratios, each scanning layer corresponds to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scanning layer; and an anomaly identification module, used to determine the anomaly detection threshold corresponding to the ratio sequence, and use the anomaly detection threshold to analyze the ratio sequence, thereby determining the scanning layer in which the functional image data has information missing anomalies.

[0015] According to another aspect of the embodiments of this application, an electronic device is also provided, including: a memory and a processor, the processor being configured to run a program stored in the memory, wherein the program executes a method for detecting missing information in medical images during runtime.

[0016] According to another aspect of the embodiments of this application, a non-volatile storage medium is also provided, the non-volatile storage medium including a stored computer program, wherein the device where the non-volatile storage medium is located executes a method for detecting missing information in medical images by running the computer program.

[0017] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of a method for detecting missing information in medical images.

[0018] In this embodiment, the method involves acquiring structural and functional image data corresponding to each scan layer in the same scan sequence. The structural and functional image data are obtained by scanning the target object layer by layer. A ratio sequence is determined between the structural and functional image data, containing multiple target area ratios. Each scan layer corresponds to one target area ratio, which characterizes the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data within the same scan layer. An anomaly detection threshold corresponding to the ratio sequence is determined, and this threshold is used to analyze the ratio sequence, thereby identifying scan layers with information loss anomalies in the functional image data. By using structural images (such as CT) as a reference benchmark and analyzing the distribution of human body area ratios in each layer of functional images (such as PET) and structural images, the method automatically identifies local signal loss or acquisition failures in functional images. This achieves the goal of accurately distinguishing between normal low uptake and abnormal signal loss, and provides automated detection without complex training, thus solving the technical problem of poor detection performance for information loss anomalies in functional images in related technologies. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a hardware structure block diagram of a computer terminal (or electronic device) for implementing a method for detecting missing information in medical images, according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram of a method for detecting missing information in medical images according to an embodiment of this application;

[0022] Figure 3 This is a schematic diagram of the overall process of a dual-modal joint medical image information missing detection according to an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of an area ratio distribution and anomaly detection according to an embodiment of this application;

[0024] Figure 5 This is a schematic diagram of the structure of a medical image information missing detection device provided according to an embodiment of this application. Detailed Implementation

[0025] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0027] In related technologies, manual visual inspection relies on operator experience, resulting in low efficiency, high subjectivity, and a high likelihood of overlooking minor or localized signal deficiencies. PET-based single-modal detection methods struggle to distinguish between "physiologically low-uptake areas" (such as adipose tissue and lungs) and "pathological or equipment-related signal loss," leading to high false-positive rates and frequent misjudgments. While deep learning methods offer advantages in complex pattern recognition, they heavily rely on large amounts of well-labeled training data, resulting in high model training costs, limited generalization ability, and their "black box" nature leading to a lack of interpretability in the results, making them difficult to adopt in clinical quality control processes. Furthermore, none of the above methods utilize the stable anatomical benchmarks provided by structural imaging (CT), failing to establish objective and adaptive detection logic from the perspective of intermodal spatial consistency, making it difficult to achieve accurate, efficient, and verifiable identification of missing functional image information.

[0028] To address the aforementioned issues, this application provides a solution that proposes a method for detecting missing information in dual-modal medical images. This method utilizes structural images (CT) with clear anatomical structures as a reference benchmark. By analyzing the distribution curves of the ratio of human body area in each layer between functional images (PET) and structural images, it automatically identifies local signal loss or acquisition failures in functional images. The following is a detailed description of this method.

[0029] According to an embodiment of this application, a method for detecting missing information in medical images is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0030] The methods and embodiments provided in this application can be executed on mobile terminals, computer terminals, or similar computing devices. Figure 1 A hardware block diagram of a computer terminal (or electronic device) for implementing a method for detecting missing information in medical images is shown. Figure 1 As shown, the computer terminal 10 (or electronic device) may include one or more processors 102 (shown as 102a, 102b, ..., 102n in the figure) 102 (processor 102 may include, but is not limited to, a microprocessor MCU or a programmable logic device FPGA, etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of a BUS bus), a network interface, a power supply, and / or a camera. Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the aforementioned electronic device. For example, computer terminal 10 may also include... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0031] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits are generally referred to herein as "data processing circuits". These data processing circuits may be embodied, in whole or in part, in software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuits may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the computer terminal 10 (or electronic device). As involved in the embodiments of this application, the data processing circuits serve as a processor control mechanism (e.g., selection of a variable resistor termination path connected to an interface).

[0032] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the medical image information missing detection method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the aforementioned medical image information missing detection method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to the computer terminal 10 via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0033] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the computer terminal 10. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0034] The display can be, for example, a touchscreen liquid crystal display (LCD) that allows the user to interact with the user interface of the computer terminal 10 (or electronic device).

[0035] Under the above operating environment, this application provides a method for detecting missing information in medical images. Figure 2 This is a schematic diagram of a method for detecting missing information in medical images according to an embodiment of this application, as shown below. Figure 2 As shown, the method includes the following steps:

[0036] Step S202: Obtain structural image data and functional image data corresponding to each scan layer in the same scan sequence, wherein the structural image data and functional image data are obtained by scanning the target object layer by layer;

[0037] Step S204: Determine the corresponding ratio sequence between structural image data and functional image data. The ratio sequence contains multiple target area ratios, with each scanning layer corresponding to a target area ratio. The target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scanning layer.

[0038] Step S206: Determine the anomaly detection threshold corresponding to the ratio sequence, and use the anomaly detection threshold to analyze the ratio sequence, thereby identifying the scanning layer where the functional image data has information missing anomalies.

[0039] By using the above steps, structural images (such as CT) are used as a reference benchmark. By analyzing the ratio distribution of human body area in each layer of functional images (such as PET) and structural images, local signal loss or acquisition failure in functional images can be automatically identified. This achieves the goal of accurately distinguishing between normal low uptake and abnormal signal loss, and is an automated detection method that does not require complex training. This solves the technical problem of poor detection effect of related technologies for abnormal missing information in functional images.

[0040] The method for detecting missing information in medical images in steps S202 to S206 of the embodiments of this application will be further described below.

[0041] Figure 3 This is a schematic diagram of the overall process for dual-modal joint medical image information missing detection according to an embodiment of this application, as shown below. Figure 3 As shown in this embodiment, structural (CT) and functional (PET) image data from the same scanning sequence can be acquired first. For example, the DICOM sequences of PET and CT can be read and sorted separately, and then reference alignment can be performed to ensure that the two are registered in spatial geometry. If necessary, spatial resolution can also be unified. Subsequently, a CT human body mask (i.e., the human body mask in the structural image data) and a PET signal region mask (i.e., the effective signal region in the functional image data) are generated in parallel, and the ratio of the area of ​​the effective signal region in the functional image to the area of ​​the human body mask in the structural image is calculated layer by layer. Construct an area ratio distribution sequence Then, statistical analysis is performed on the sequence to calculate the statistical mean, interquartile range, and other statistical measures of the ratio sequence, and an adaptive anomaly detection threshold (such as the Tukey lower bound and the lower bound of the mean ratio) is constructed. Finally, layers in the sequence whose ratios are significantly lower than the anomaly detection threshold can be identified according to the judgment rules, and an anomaly layer list and visualization curve are output. The above process is described in detail below.

[0042] First, obtain the structural and functional image data corresponding to each scan layer in the same scan sequence. For example, CT and PET DICOM files in the same scan sequence can be read separately. Both types of data are derived from the raw information generated by scanning the target object layer by layer. In order to ensure the validity of the data in the subsequent bimodal analysis, the two need to be aligned to ensure that they are registered in spatial geometry. The specific steps are as follows.

[0043] In some embodiments of this application, after acquiring structural image data and functional image data in the same scanning sequence, the method further includes: determining the first number of slices contained in the structural image data and the second number of slices contained in the functional image data; determining the smaller value between the first number of slices and the second number of slices as the effective number of slices, wherein the effective number of slices is the number of common layers between the structural image data and the functional image data; and sorting each slice in the structural image data and each slice in the functional image data according to the Z-axis position based on the effective number of slices, so as to ensure that the structural image data and the functional image data correspond layer by layer in space, wherein the Z-axis is the scanning axis along the head-to-foot direction of the human body.

[0044] Specifically, the number of the first and second slices contained in both images can be determined first, and the smaller of the two values ​​can be taken as the effective slice number N. This accurately defines the effective number of layers that can actually correspond between the structural image and the functional image. Layers exceeding the range of N (e.g., if CT has 300 layers and PET only has 280 layers, then layers 281-300) are not included in the calculation to avoid incorrect correspondence due to missing data on one side. Then, based on the effective slice number, the slices in the structural image and the functional image are precisely sorted according to their Z-axis positions along the head-to-foot direction of the human body to ensure that the two achieve a one-to-one correspondence layer by layer in the spatial dimension, avoiding inter-slice mismatch caused by inconsistent slice numbers or scanning position offsets.

[0045] In addition, since the spatial resolution of structural CT and functional PET is usually different, it is often necessary to unify the spatial resolution. The specific steps are as follows.

[0046] In some embodiments of this application, after acquiring structural image data and functional image data in the same scanning sequence, the method further includes: determining a first spatial resolution corresponding to the structural image data and a second spatial resolution corresponding to the functional image data; and using an interpolation algorithm to convert the structural image data of the first spatial resolution to the second spatial resolution, or to convert the functional image data of the second spatial resolution to the first spatial resolution, so that the spatial resolution of the structural image data and the spatial resolution of the functional image data are consistent.

[0047] Specifically, interpolation algorithms can be used to perform resolution conversion on one type of data. For example, nearest neighbor interpolation (preserving binary characteristics and avoiding grayscale aliasing) can be used to resample the mask of one modality to the resolution of another modality, so that the mask... Comparability within the same pixel space ensures spatial alignment of the two at a unified pixel scale, thereby guaranteeing accurate matching of the target area ratio calculated subsequently based on the same spatial precision. This avoids area estimation deviations and ratio sequence distortions caused by resolution differences, thus ensuring reasonable setting of anomaly detection thresholds and accurate identification of anomaly scanning layers.

[0048] After completing the benchmark alignment of the structural and functional image data, for each layer of matched structural and functional image data, the ratio of the effective signal region area in the functional image to the human body mask area in the structural image can be calculated, forming a ratio sequence arranged in the order of scanning layers, as follows.

[0049] In some embodiments of this application, determining the corresponding ratio sequence between structural image data and functional image data includes: for each scan layer, determining the area of ​​a first region of the human body mask in the structural image data, and determining the area of ​​a second region of the effective signal region in the functional image data, wherein the human body mask is used to characterize the spatial range of human tissue in the scan layer, and the effective signal region is used to characterize the coverage range of metabolic signals collected in the functional image data of the scan layer; the ratio between the area of ​​the second region and the area of ​​the first region corresponding to each scan layer is determined as the target area ratio corresponding to the scan layer, thereby obtaining the ratio sequence.

[0050] In this embodiment, the area of ​​the first region of the human body mask (i.e., the CT human body mask) in the structural image data can be determined first. The specific steps are as follows.

[0051] In some embodiments of this application, the structural image data includes at least one of the following: computed tomography (CT) images; determining the area of ​​a first region of the human body mask in the structural image data includes the following steps: converting the pixel values ​​of the CT image into HU values, and determining the pixel regions with HU values ​​greater than a first preset threshold as foreground regions; removing non-human body regions from the foreground regions, and performing morphological operations on the foreground regions after removing non-human body regions, wherein the non-human body regions include at least one of the following: the scanning bed region, and the morphological operations are used to fill the holes inside the foreground regions and eliminate isolated noise points; determining the largest connected region in the foreground regions after morphological operations as the human body mask, and calculating the area of ​​the human body mask to obtain the area of ​​the first region.

[0052] Specifically, the pixel values ​​of the CT image are converted to Hounsfield units (HU values), and HU values ​​greater than a first preset threshold are selected. Pixel regions with a resolution of -500 HU are marked as foreground regions. Non-human areas, including the scan bed, are then precisely removed from these foreground regions, effectively avoiding interference from non-tissue areas in mask extraction. Morphological operations are then performed on the foreground regions after removing non-human areas, significantly improving region connectivity and structural integrity by filling internal holes and eliminating isolated noise points. Finally, the largest connected region is extracted as the human mask. And calculate its area. This ensures that the calculated human body mask area only reflects the actual range of human tissue.

[0053] At the same time, it is also necessary to determine the area of ​​the second region of the effective signal region in the functional image data (i.e., the PET effective signal region mask). The specific steps are as follows.

[0054] In some embodiments of this application, the functional image data includes at least one of the following: a positron emission tomography (PET) image; determining the area of ​​a second region of effective signal area in the functional image data includes the following steps: normalizing the pixel values ​​of the PET image, and performing threshold segmentation on the normalized PET image according to a second preset threshold to obtain a PET mask, wherein the second preset threshold is used to distinguish between background areas and potential human signal areas in the PET image; taking the intersection of the PET mask and the human mask to obtain the effective signal area, and calculating the area of ​​the effective signal area to obtain the second region area.

[0055] Specifically, firstly, intensity deviations caused by different scanning devices or parameters can be eliminated by normalizing the pixel values ​​of the PET image. For example, percentile window normalization can be performed on the PET pixel values. Then, based on a second preset threshold (such as a body threshold)... and core threshold (etc.) Perform precise threshold segmentation on the normalized image to identify potential signal regions and generate a PET mask containing only potential metabolic signals; then perform spatial intersection operation on this PET mask and the human body mask extracted from structural image data (CT human body mask) to exclude regions where no human body exists in the CT scan. This effectively eliminates non-physiological signal regions that extend beyond the boundaries of human tissue, retaining only metabolically active regions that overlap with actual anatomical structures as valid signal regions, and calculating their areas accordingly. This ensures that the calculation basis for the target area ratio is completely limited to the range of physiological signals within human tissue.

[0056] After obtaining the area of ​​the first region Second region area After that, the target area ratio can be calculated. For each layer... Calculate the target area ratio This yields a distribution sequence (ratio sequence) along the body axis (Z-axis). Each ratio precisely characterizes the relative coverage of the effective signal region of the functional image relative to the human body contour of the structural image. This ratio sequence fully reflects the area matching relationship between the functional image and the structural image at each anatomical level, providing a quantitative basis for subsequent anomaly identification.

[0057] Next, statistical distribution analysis can be performed on the ratio sequence to determine the anomaly detection threshold corresponding to the ratio sequence. This anomaly detection threshold is calculated based on the statistical characteristics of the sequence itself and is used to establish an objective standard for judging whether functional image data is abnormal. The specific determination steps are as follows.

[0058] In some embodiments of this application, the anomaly detection threshold includes at least one of the following: a first anomaly detection threshold and a second anomaly detection threshold; determining the anomaly detection threshold corresponding to the ratio sequence includes the following steps: determining the arithmetic mean of the target area ratios in the ratio sequence, and determining the first anomaly detection threshold based on the arithmetic mean and a first preset adjustment coefficient, wherein the first preset adjustment coefficient is used to set the normal decay degree of the target area ratio relative to the arithmetic mean; determining the first quartile and the third quartile of the target area ratios in the ratio sequence, and determining the interquartile range based on the first quartile and the third quartile; determining the second anomaly detection threshold based on the first quartile, the interquartile range, and the second preset adjustment coefficient, wherein the second preset adjustment coefficient is used to set the leniency of the anomaly judgment based on the interquartile range, and the second anomaly detection threshold is used for outliers in the ratio sequence that deviate from the normal range.

[0059] Specifically, it can be based on this ratio sequence arithmetic mean With the first preset adjustment coefficient (For example, the default value can be 0.6) Construct the first anomaly detection threshold (lower bound of the mean ratio). The following formula represents the normal decay range of the target area ratio relative to the central trend:

[0060]

[0061] At the same time, it can be combined with the first quartile. With the third quartile Calculate interquartile range And based on the interquartile range and the second preset adjustment coefficient (For example, the default value can be 1.5) Construct a second anomaly detection threshold (Tukey lower bound) This method is used to identify outliers that deviate from the normal distribution range, thereby achieving dual adaptive judgment of information-deficient regions in functional image data: on the one hand, it filters low-uptake regions that are generally low but still within the physiologically normal range through the central trend threshold (first anomaly detection threshold); on the other hand, it accurately identifies anomalous layers with local signal loss through the dispersion threshold (second anomaly detection threshold). This effectively overcomes the problem of insufficient sensitivity and poor robustness of a single threshold under different scanning sequence distribution patterns, significantly improves the accuracy of identifying real anomalies and the ability to resist interference, and ultimately achieves efficient, reliable, and interpretable detection of missing regions in functional image data.

[0062] Determining the first anomaly detection threshold Second anomaly detection threshold Then, the abnormal layers can be identified by comparing the value sequences, as follows.

[0063] In some embodiments of this application, the analysis of anomaly detection threshold comparison value sequence to determine that the functional image data of the scan layer has information missing anomaly includes: when the target area ratio in the ratio sequence is less than a first anomaly detection threshold and / or the target area ratio is less than a second anomaly detection threshold, determining that the functional image data of the scan layer corresponding to the target area ratio has information missing anomaly.

[0064] Specifically, when the target area ratio of any scanned layer is lower than either of the two thresholds mentioned above, it is determined that the functional image of that layer has an information loss anomaly. For example, for each layer... If any of the following conditions are met, it is marked as an abnormal layer:

[0065] Tukey Rules: ;

[0066] Mean ratio rule: ;

[0067] Finally, the identification results can be output, and the output information may include, but is not limited to: the index of the anomaly layer, the anomaly type (such as "tukey", "mean_frac", or "both"), etc. A ratio distribution curve can also be plotted for manual verification. For example, ... Figure 4 The diagram shown illustrates an area ratio distribution and anomaly detection. Figure 4 The middle section shows the area ratio with the slice index as the horizontal axis (X-axis). This is a scatter plot showing the distribution of points along the vertical axis (Y-axis). The dense clusters at the top represent normal scan slices, with a high and stable PET / CT area ratio. The mean line is also marked on the plot. ), lower bound of the mean proportion ( ) and Tukey's lower bound ( Discrete points located below the lower boundary (e.g.) Points with values ​​close to 0 are marked as outliers, visually indicating the specific layer where signal loss occurred.

[0068] The above method enables parallel and collaborative identification of two types of abnormal patterns: overall attenuation and local outlier. It effectively distinguishes between normal low-intake areas and areas of real signal loss, improves the accuracy and robustness of detection, and overcomes the defect of easy misjudgment by single threshold determination.

[0069] It should be noted that this method is not only applicable to CT / PET images, but also to industrial X-ray inspection, visible light surface defect inspection, and other scenarios.

[0070] This application's solution acquires structural and functional image data corresponding to each scan layer in the same scanning sequence, constructs a target area ratio sequence with the ratio of the effective signal region area of ​​each functional image to the human body mask area of ​​the structural image as its core, and uses the stable anatomical mask in the structural image as an objective benchmark to normalize the expression of the functional image signal intensity. This effectively distinguishes between low ratio regions caused by normal physiological low uptake and abnormally low ratio regions caused by acquisition failure or signal loss, overcoming the shortcomings of traditional methods in accurately identifying true abnormalities. Furthermore, based on the statistical distribution characteristics of this ratio sequence, it automatically calculates anomaly detection... By measuring thresholds and using them to perform consistency analysis on the ratios of each layer, the system achieves automated and high-precision identification of abnormal layers with missing functional image information. This avoids the subjectivity and inefficiency of manual interpretation, requires no deep learning training, is simple to deploy, and is computationally efficient. At the same time, it gives the abnormality judgment process clear physical meaning and interpretability, and can adapt to the differences between different patients and different scanning protocols. This effectively solves the technical problems of low efficiency, high misjudgment rate, and poor interpretability in existing technologies for detecting missing functional image data. It achieves the effect of improving the automation level and diagnostic reliability of medical image quality control, and facilitates clinical understanding and integration into quality control processes.

[0071] According to an embodiment of this application, an embodiment of a medical image information missing detection device is also provided. Figure 5 This is a schematic diagram of a medical image information missing detection device according to an embodiment of this application. Figure 5 As shown, the device includes:

[0072] The data acquisition module 50 is used to acquire structural image data and functional image data in the same scan sequence. The scan sequence contains multiple scan layers, and the structural image data and functional image data are obtained by scanning the target object layer by layer.

[0073] The ratio calculation module 52 is used to determine the corresponding ratio sequence between structural image data and functional image data. The ratio sequence contains multiple target area ratios, and each scanning layer corresponds to a target area ratio. The target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scanning layer.

[0074] The anomaly identification module 54 is used to determine the anomaly detection threshold corresponding to the ratio sequence, and to analyze the ratio sequence using the anomaly detection threshold, thereby identifying the scanning layer where the functional image data has information missing anomalies.

[0075] Optionally, after acquiring structural image data and functional image data in the same scanning sequence, the method further includes: determining the first number of slices contained in the structural image data and the second number of slices contained in the functional image data; determining the smaller value between the first number of slices and the second number of slices as the effective number of slices, wherein the effective number of slices is the number of common layers between the structural image data and the functional image data; and sorting each slice in the structural image data and each slice in the functional image data according to the Z-axis position based on the effective number of slices to ensure that the structural image data and the functional image data correspond layer by layer in space, wherein the Z-axis is the scanning axis along the head-to-foot direction of the human body.

[0076] Optionally, after acquiring structural image data and functional image data in the same scanning sequence, the method further includes: determining a first spatial resolution corresponding to the structural image data and a second spatial resolution corresponding to the functional image data; and using an interpolation algorithm to convert the structural image data of the first spatial resolution to the second spatial resolution, or to convert the functional image data of the second spatial resolution to the first spatial resolution, so that the spatial resolution of the structural image data and the spatial resolution of the functional image data are consistent.

[0077] Optionally, determining the corresponding ratio sequence between structural image data and functional image data includes: for each scan layer, determining the area of ​​a first region of the human body mask in the structural image data, and determining the area of ​​a second region of the effective signal region in the functional image data, wherein the human body mask is used to characterize the spatial range of human tissue in the scan layer, and the effective signal region is used to characterize the coverage range of metabolic signals collected in the functional image data of the scan layer; the ratio between the area of ​​the second region and the area of ​​the first region corresponding to each scan layer is determined as the target area ratio corresponding to the scan layer, thereby obtaining the ratio sequence.

[0078] Optionally, the structural image data includes at least one of the following: computed tomography (CT) images; determining the area of ​​the first region of the human mask in the structural image data includes: converting the pixel values ​​of the CT image into HU values, and determining the pixel regions with HU values ​​greater than a first preset threshold as foreground regions; removing non-human regions from the foreground regions, and performing morphological operations on the foreground regions after removing non-human regions, wherein the non-human regions include at least one of the following: the scanning bed region, and the morphological operations are used to fill the holes inside the foreground regions and eliminate isolated noise points; determining the largest connected region in the foreground regions after morphological operations as the human mask, and calculating the area of ​​the human mask to obtain the area of ​​the first region.

[0079] Optionally, the functional image data includes at least one of the following: a positron emission tomography (PET) image; determining the area of ​​the second region of the effective signal region in the functional image data includes: normalizing the pixel values ​​of the PET image, and performing threshold segmentation on the normalized PET image according to a second preset threshold to obtain a PET mask, wherein the second preset threshold is used to distinguish between background regions and potential human signal regions in the PET image; taking the intersection of the PET mask and the human mask to obtain the effective signal region, and calculating the area of ​​the effective signal region to obtain the area of ​​the second region.

[0080] Optionally, the anomaly detection threshold includes at least one of the following: a first anomaly detection threshold and a second anomaly detection threshold; determining the anomaly detection threshold corresponding to the ratio sequence includes: determining the arithmetic mean of the target area ratios in the ratio sequence, and determining the first anomaly detection threshold based on the arithmetic mean and a first preset adjustment coefficient, wherein the first preset adjustment coefficient is used to set the normal decay degree of the target area ratio relative to the arithmetic mean; determining the first quartile and the third quartile of the target area ratios in the ratio sequence, and determining the interquartile range based on the first quartile and the third quartile; determining the second anomaly detection threshold based on the first quartile, the interquartile range, and the second preset adjustment coefficient, wherein the second preset adjustment coefficient is used to set the leniency of the anomaly judgment based on the interquartile range, and the second anomaly detection threshold is used for outliers in the ratio sequence that deviate from the normal range.

[0081] Optionally, the analysis of the anomaly detection threshold comparison value sequence to determine the scan layer with missing information in the functional image data includes: determining that the functional image data of the scan layer corresponding to the target area ratio has missing information when the target area ratio in the ratio sequence is less than a first anomaly detection threshold and / or the target area ratio is less than a second anomaly detection threshold.

[0082] It should be noted that the modules in the above-mentioned medical image information missing detection device can be program modules (e.g., a set of program instructions to implement a certain function) or hardware modules. For the latter, they can be in the following forms, but are not limited to these: each of the above modules is in the form of a processor, or the functions of each of the above modules are implemented by a processor.

[0083] It should be noted that the medical image information missing detection device provided in this embodiment can be used to perform... Figure 2 The method for detecting missing information in medical images shown above is also applicable to the embodiments of this application, and will not be repeated here.

[0084] This application embodiment also provides a non-volatile storage medium, which includes a stored computer program. The device containing the non-volatile storage medium executes the following method for detecting missing information in medical images by running the computer program: acquiring structural image data and functional image data corresponding to each scan layer in the same scan sequence, wherein the structural image data and functional image data are obtained by scanning the target object layer by layer; determining a corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence contains multiple target area ratios, with each scan layer corresponding to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scan layer; determining an anomaly detection threshold corresponding to the ratio sequence, and using the anomaly detection threshold to analyze the ratio sequence, thereby determining the scan layer in the functional image data where there is an information missing anomaly.

[0085] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the medical image information missing detection method described in various embodiments of this application: acquiring structural image data and functional image data corresponding to each scanning layer in the same scanning sequence, wherein the structural image data and functional image data are obtained by scanning the target object layer by layer; determining the corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence contains multiple target area ratios, each scanning layer corresponds to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data to the human body mask area in the structural image data of the same scanning layer; determining the anomaly detection threshold corresponding to the ratio sequence, and using the anomaly detection threshold to analyze the ratio sequence, thereby determining the scanning layer in which the functional image data has information missing anomalies.

[0086] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0087] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0088] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0089] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0090] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0091] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard drive, magnetic disk, or optical disk.

[0092] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for detecting missing information in medical images, characterized in that, include: Obtain structural image data and functional image data corresponding to each scan layer in the same scan sequence, wherein the structural image data and functional image data are obtained by scanning the target object layer by layer; Determine the corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence contains multiple target area ratios, each scanning layer corresponds to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data of the same scanning layer to the human body mask area in the structural image data; An anomaly detection threshold corresponding to the ratio sequence is determined, and the ratio sequence is analyzed using the anomaly detection threshold to determine the scanning layer where the functional image data has an information missing anomaly.

2. The method for detecting missing information in medical images according to claim 1, characterized in that, After acquiring structural and functional image data from the same scan sequence, the method further includes: Determine the first number of slices contained in the structural image data and the second number of slices contained in the functional image data; The smaller value between the first slice number and the second slice number is determined as the effective slice number, wherein the effective slice number is the number of common layers between the structural image data and the functional image data; Based on the number of effective slices, each slice in the structural image data and each slice in the functional image data are sorted according to the Z-axis position to ensure that the structural image data and functional image data correspond layer by layer in space, wherein the Z-axis is the scanning axis along the head-to-foot direction of the human body.

3. The method for detecting missing information in medical images according to claim 1, characterized in that, After acquiring structural and functional image data from the same scan sequence, the method further includes: Determine the first spatial resolution corresponding to the structural image data and the second spatial resolution corresponding to the functional image data; An interpolation algorithm is used to convert the structural image data at the first spatial resolution to the second spatial resolution, or to convert the functional image data at the second spatial resolution to the first spatial resolution, so that the spatial resolution of the structural image data and the spatial resolution of the functional image data are consistent.

4. The method for detecting missing information in medical images according to claim 1, characterized in that, Determining the corresponding ratio sequence between the structural image data and the functional image data includes: For each scanning layer, the area of ​​a first region of the human body mask in the structural image data is determined, and the area of ​​a second region of the effective signal region in the functional image data is determined, wherein the human body mask is used to characterize the spatial extent of human tissue in the scanning layer, and the effective signal region is used to characterize the coverage of metabolic signals collected in the functional image data of the scanning layer; The ratio between the area of ​​the second region and the area of ​​the first region corresponding to each scan layer is determined as the target area ratio corresponding to the scan layer, thereby obtaining the ratio sequence.

5. The method for detecting missing information in medical images according to claim 4, characterized in that, The structural imaging data includes at least one of the following: computed tomography (CT) images; Determining the area of ​​the first region of the human mask in the structural image data includes: The pixel values ​​of the CT image are converted into HU values, and the pixel regions with HU values ​​greater than a first preset threshold are determined as foreground regions; Non-human areas are removed from the foreground region, and morphological operations are performed on the foreground region after removing the non-human areas. The non-human areas include at least one of the following: the scanning bed region. The morphological operations are used to fill the holes inside the foreground region and eliminate isolated noise points. The largest connected region in the foreground region after the morphological operation is determined as the human body mask, and the area of ​​the human body mask is calculated to obtain the area of ​​the first region.

6. The method for detecting missing information in medical images according to claim 4, characterized in that, The functional image data includes at least one of the following: positron emission tomography (PET) images; determining the area of ​​a second region in the functional image data that represents a valid signal region includes: The pixel values ​​of the PET image are normalized, and the normalized PET image is segmented according to a second preset threshold to obtain a PET mask. The second preset threshold is used to distinguish between the background area and the potential human signal area in the PET image. The intersection of the PET mask and the human body mask is used to obtain the effective signal region, and the area of ​​the effective signal region is calculated to obtain the area of ​​the second region.

7. The method for detecting missing information in medical images according to claim 1, characterized in that, The anomaly detection threshold includes at least one of the following: a first anomaly detection threshold and a second anomaly detection threshold; determining the anomaly detection threshold corresponding to the ratio sequence includes: The arithmetic mean of the target area ratios in the ratio sequence is determined, and the first anomaly detection threshold is determined based on the arithmetic mean and a first preset adjustment coefficient, wherein the first preset adjustment coefficient is used to set the normal decay degree of the target area ratio relative to the arithmetic mean; Determine the first quartile and the third quartile of the target area ratio in the ratio sequence, and determine the interquartile range based on the first quartile and the third quartile; A second anomaly detection threshold is determined based on the first quartile, the interquartile range, and a second preset adjustment coefficient. The second preset adjustment coefficient is used to set the leniency level of anomaly judgment based on the interquartile range, and the second anomaly detection threshold is used for outliers in the ratio sequence that deviate from the normal range.

8. The method for detecting missing information in medical images according to claim 7, characterized in that, Analyzing the ratio sequence using the aforementioned anomaly detection threshold to determine the scan layers in which the functional image data exhibits information loss anomalies includes: If the target area ratio in the ratio sequence is less than the first anomaly detection threshold, and / or the target area ratio is less than the second anomaly detection threshold, it is determined that the functional image data of the scan layer corresponding to the target area ratio has an information missing anomaly.

9. A medical imaging information missing detection device, characterized in that, include: The data acquisition module is used to acquire structural image data and functional image data in the same scanning sequence, wherein the scanning sequence contains multiple scanning layers, and the structural image data and functional image data are obtained by scanning the target object layer by layer; The ratio calculation module is used to determine the corresponding ratio sequence between the structural image data and the functional image data, wherein the ratio sequence contains multiple target area ratios, each scanning layer corresponds to one target area ratio, and the target area ratio is used to characterize the ratio of the effective signal region area in the functional image data of the same scanning layer to the human body mask area in the structural image data; An anomaly identification module is used to determine the anomaly detection threshold corresponding to the ratio sequence, and to analyze the ratio sequence using the anomaly detection threshold, thereby determining that the functional image data has an anomaly of missing information in the scanning layer.

10. An electronic device, characterized in that, include: A memory and a processor, the processor being configured to run a program stored in the memory, wherein the program, when running, executes the information missing detection method for medical images according to any one of claims 1 to 8.

11. A non-volatile storage medium, characterized in that, The non-volatile storage medium includes a stored computer program, wherein the device containing the non-volatile storage medium executes the information missing detection method for medical images according to any one of claims 1 to 8 by running the computer program.

12. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the medical image information missing detection method according to any one of claims 1 to 8.

Images