Examination report template presentation method, nonvolatile readable storage medium and electronic equipment

Abstract

The invention relates to the technical field of medicine, and discloses a presentation method of an inspection report template, a nonvolatile readable storage medium and electronic equipment. The method comprises the steps of obtaining a medical ultrasonic image, determining a part feature vector and a prior feature vector of a target examination part according to the medical ultrasonic image, determining a target part type of the target examination part according to the part feature vector and the prior feature vector, and presenting a target examination report template corresponding to the target part type. According to the embodiment, the target part type of the target examination part can be automatically recognized from the medical ultrasonic image, the target examination report template corresponding to the target part type can be automatically called and presented, intervention operation of a doctor or other people is not needed, and therefore the tedious calling process is avoided, and the working efficiency can be improved. In addition, according to the embodiment of the invention, by fusing the part feature vector and the prior feature vector, the target inspection report template is more accurately and reliably called for presentation.

Description

Technical Field

[0001] The present disclosure relates to the field of medical technology, and in particular to a method for presenting an examination report template, a non-volatile readable storage medium, and an electronic device. Background Art

[0002] In daily medical ultrasound examinations, doctors need to retrieve corresponding examination items according to different examination parts of the patient and open the examination report template under the examination item. Doctors can edit information on the examination report template or not. Usually, doctors need to frequently switch workstations to the computer to retrieve and present the examination report template, or doctors arrange clerks to assist in retrieving and presenting the examination report template, or doctors issue a retrieval voice, and the computer processes the retrieval voice through voice software, parses the retrieval command of the examination report template, and retrieves and presents the examination report template according to the retrieval command. The above approach is cumbersome and is not conducive to improving the doctor's work efficiency. Summary of the invention

[0003] One purpose of the embodiments of the present disclosure is to provide a method for presenting an inspection report template, a non-volatile readable storage medium, and an electronic device, aiming to solve the technical problem that the inspection report template retrieval process in the related art is relatively complicated.

[0004] In a first aspect, an embodiment of the present disclosure provides a method for presenting an inspection report template, comprising:

[0005] Acquire a medical ultrasound image, wherein the medical ultrasound image includes an image area corresponding to the target examination site;

[0006] Determining a site feature vector and a priori feature vector of the target examination site according to the medical ultrasound image, wherein the priori feature vector is used to represent the characteristics of factors associated with the target examination site;

[0007] Determining a target part type of the target inspection part according to the part feature vector and the priori feature vector;

[0008] A target inspection report template corresponding to the target site type is presented.

[0009] Optionally, determining the site feature vector of the target examination site according to the medical ultrasound image includes:

[0010] The part feature vector of the target examination part is determined according to a preset ResNet model and the medical ultrasound image.

[0011] Optionally, the prior feature vector includes a probe feature vector for representing a probe shape, the medical ultrasound image includes a probe image region corresponding to the probe, and determining the prior feature vector according to the medical ultrasound image includes:

[0012] The probe feature vector is determined according to the preset ResNet model and the probe image area.

[0013] Optionally, determining the probe feature vector according to a preset ResNet model and the probe image area includes:

[0014] intercepting an image area including the probe on the medical ultrasound image as a probe image area;

[0015] A probe feature vector is determined according to the ResNet model and the probe image area.

[0016] Optionally, intercepting an image area including the probe on the medical ultrasound image as a probe image area comprises:

[0017] Determine a cutoff point on the width boundary of the medical ultrasound image, wherein a ratio of a distance from the cutoff point to an upper left corner point of the medical ultrasound image to a width of the medical ultrasound image is within a range of [1 / 4, 1 / 2];

[0018] The medical ultrasound image is horizontally intercepted starting from the interception point to obtain an intercepted image with a minimum area, and the intercepted image is the probe image area.

[0019] Optionally, the ResNet model includes multiple residual blocks, each of which includes a first convolutional layer, a second convolutional layer, a third convolutional layer and a fourth convolutional layer connected in sequence, wherein the second convolutional layer and the third convolutional layer are depth-separable convolutional layers.

[0020] Optionally, the size of the convolution kernel of the first convolution layer is 1*1, the size of the convolution kernel of the second convolution layer is 3*3, the size of the convolution kernel of the third convolution layer is 1*1, and the size of the convolution kernel of the fourth convolution layer is 1*1.

[0021] Optionally, determining the target part type of the target inspection part according to the part feature vector and the priori feature vector includes:

[0022] splicing the part feature vector and the prior feature vector to obtain a spliced ​​feature vector;

[0023] The target part type of the target inspection part is determined according to a preset classification model and the spliced ​​feature vector.

[0024] Optionally, presenting a target inspection report template corresponding to the target part type includes:

[0025] Determine an inspection item interface corresponding to the target part type;

[0026] Determining, in a preset report library, an inspection report template corresponding to the target part type as a target inspection report template;

[0027] Control the target inspection report to be presented on the inspection item interface.

[0028] In a second aspect, an embodiment of the present disclosure provides a non-volatile readable storage medium storing computer executable instructions, wherein the computer executable instructions are used to enable an electronic device to execute the above-mentioned method for presenting an inspection report template.

[0029] In a third aspect, an embodiment of the present disclosure provides an electronic device, including:

[0030] at least one processor; and,

[0031] a memory communicatively connected to the at least one processor; wherein,

[0032] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned method for presenting the inspection report template.

[0033] In the method for presenting the inspection report template provided in the embodiment of the present disclosure, a medical ultrasound image is obtained, wherein the medical ultrasound image includes an image area corresponding to the target inspection part, a part feature vector and a priori feature vector of the target inspection part are determined according to the medical ultrasound image, wherein the priori feature vector is used to represent the characteristics of the factors associated with the target inspection part, the target part type of the target inspection part is determined according to the part feature vector and the priori feature vector, and a target inspection report template corresponding to the target part type is presented. This embodiment can automatically identify the target part type of the target inspection part from the medical ultrasound image, automatically retrieve and present the target inspection report template corresponding to the target part type, without the intervention of a doctor or other person, thereby avoiding a tedious retrieval process, which is conducive to improving work efficiency. In addition, this embodiment can more reliably and accurately determine the target part type of the target inspection part by fusing the part feature vector and the priori feature vector, so that the target inspection report template can be retrieved more accurately and reliably for presentation. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0035] Figure 1 A schematic diagram of the structure of a medical examination system provided by an embodiment of the present disclosure;

[0036] Figure 2a to Figure 2d Schematic diagram of various probes provided by the prior art;

[0037] Figure 3 A flowchart of a method for presenting an inspection report template provided in an embodiment of the present disclosure;

[0038] Figure 4 A model architecture diagram of a ResNet model provided in an embodiment of the present disclosure;

[0039] Figure 5a to Figure 5d Schematic diagram of medical ultrasound images acquired using various probes provided in an embodiment of the present disclosure;

[0040] Figure 6 A schematic diagram of the structure of a device for presenting an inspection report template provided in an embodiment of the present disclosure;

[0041] Figure 7 A schematic diagram of a circuit structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the present disclosure more clearly understood, the present disclosure is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present disclosure and are not intended to limit the present disclosure. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present disclosure.

[0043] It should be noted that, if there is no conflict, the various features in the embodiments of the present disclosure can be combined with each other, all within the scope of protection of the present disclosure. In addition, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flow chart, in some cases, the steps shown or described can be performed in a different order than the module division in the device or the flow chart. Furthermore, the words "first", "second", "third", etc. used in the present disclosure do not limit the data and execution order, but only distinguish the same items or similar items with basically the same functions and effects.

[0044] The present disclosure provides a medical examination system. Figure 1 The medical examination system 100 includes an ultrasonic examination instrument 11 and a computer 12, and the ultrasonic examination instrument 11 is in communication connection with the computer 12. The ultrasonic examination instrument 11 is equipped with different probes to collect medical ultrasonic images of the examinee.

[0045] See also Figure 2a to Figure 2d Usually, when doctors perform ultrasound examinations on different parts of the body, they need to use an appropriate probe to scan. Figure 2a As shown in the figure, when doctors perform ultrasound examinations on the abdomen, such as the liver, gallbladder, pancreas, urinary system, uterus and other parts of the body, they need to use a convex array probe for scanning. Figure 2b As shown in the figure, when doctors perform ultrasound examinations on superficial organs, such as the carotid artery, thyroid gland, breast, etc., they need to use a linear array probe for scanning. Figure 2c When doctors do an ultrasound of the heart, they use a heart probe to scan the heart. Figure 2d As shown, when a doctor performs an ultrasound examination of the vagina, the doctor needs to use an intracavitary probe to scan.

[0046] The doctor uses the ultrasound examination instrument 11 to examine the examinee at the examination station. It is understandable that the examinee may have one or more examination items, such as the hepatobiliary and pancreatic items, uterine items, or thyroid items. Usually, in order to reflect the examination conclusion information of the examination item, the doctor needs to retrieve the examination report template corresponding to the examination item from the computer 12. The examination report template is used to record the examination conclusion information. For example, when the doctor examines the examinee's liver, the hepatobiliary and pancreatic items corresponding to the liver are called on the computer 12, and the examination report template corresponding to the liver is retrieved and presented under the hepatobiliary and pancreatic items, wherein the medical ultrasound image corresponding to the liver will be saved under the hepatobiliary and pancreatic items.

[0047] Next, when the doctor examines the examinee's heart again, he calls the heart project corresponding to the heart on the computer 12, and calls and presents the examination report template corresponding to the heart under the heart project, wherein the medical ultrasound image corresponding to the heart will be saved under the heart project.

[0048] As mentioned in the background technology, the related technology requires the doctor to switch between different examination items at the examination station and the computer and retrieve the examination report template under the switched examination item. This process is relatively cumbersome and greatly reduces the doctor's work efficiency.

[0049] Or, if Figure 1 As shown, the doctor notifies the nurse to switch different examination items and retrieves the examination report template under the switched examination items. This process is also not conducive to improving the doctor's work efficiency.

[0050] Alternatively, the related technology can install a voice collection device, which collects the call voice issued by the doctor and sends the call voice to the computer 12. The computer 12 processes the call voice through voice software, parses the call command of the examination report template, and calls and presents the examination report template according to the call command. However, the hospital is noisy, and the voice collection device also collects noise, which is not conducive to the computer 12 to accurately and reliably parse the call command, and easily reduces the success rate of calling the examination report template.

[0051] Based on this, the present disclosure provides a method for presenting an inspection report template. Figure 3 , the method for presenting the inspection report template includes the following steps:

[0052] S31: Acquire a medical ultrasound image, wherein the medical ultrasound image includes an image area corresponding to the target examination site.

[0053] In this step, the medical ultrasound image is an ultrasound image corresponding to the target examination site of the examiner acquired by the ultrasound examination instrument through the probe. The target examination site is one of various examination sites, wherein the examination site includes various devices located in the abdomen, superficial organs, heart, etc. Acquiring the medical ultrasound image includes the following steps: acquiring the medical ultrasound image acquired by the ultrasound examination instrument. The computer polls the ultrasound examination instrument at a preset frequency, so that the ultrasound examination instrument sends the medical ultrasound image to the computer.

[0054] S32: Determine a site feature vector and a priori feature vector of the target examination site according to the medical ultrasound image.

[0055] In this step, the part feature vector is used to represent the features of the target inspection part, and the prior feature vector is used to represent the features of the factors associated with the target inspection part, such as the factors including other factors such as the probe factor. As mentioned above, when the doctor uses a convex array probe to scan the target inspection part of the examinee, the target inspection part may be the liver, gallbladder, pancreas, urinary system or uterus, but the target inspection part is usually not the carotid artery, thyroid gland, breast or heart. Similarly, when the doctor uses a linear array probe to scan the target inspection part of the examinee, the target inspection part may be the carotid artery, thyroid gland or breast, but the target inspection part is usually not the liver, gallbladder, pancreas, urinary system, uterus or heart.

[0056] S33: Determine the target part type of the target inspection part according to the part feature vector and the priori feature vector.

[0057] In this step, the target part type is the part type of the target inspection part, and the part type is used to indicate the name of the inspection part, for example, the target part type is a heart type, a uterus type, a kidney type, or a liver type.

[0058] In some embodiments, determining the target part type of the target inspection part according to the part feature vector and the prior feature vector includes: splicing the part feature vector and the prior feature vector to obtain a spliced ​​feature vector, and determining the target part type of the target inspection part according to a preset classification model and the spliced ​​feature vector. This embodiment fuses the part feature vector and the prior feature vector, so that the target part type of the target inspection part can be determined more reliably and accurately.

[0059] It can be understood that, in this embodiment, the part feature vector and the priori feature vector can be concatenated into a concatenated feature vector in the order of the part feature vector and the priori feature vector.

[0060] It can also be understood that, in this embodiment, the priori feature vector and the part feature vector can be concatenated into a concatenated feature vector in the order of the priori feature vector and the part feature vector.

[0061] In some embodiments, the classification model may be a logit model (Logit model, assessment model) or a model that supports other classification algorithms.

[0062] S34: Presenting a target inspection report template corresponding to the target part type.

[0063] In this step, the target inspection report template is an inspection report template corresponding to the target part type, and the inspection report template is used to record the inspection description information associated with the inspection part. In some embodiments, presenting the target inspection report template corresponding to the target part type includes the following steps: determining an inspection item interface corresponding to the target part type, determining the inspection report template corresponding to the target part type as the target inspection report template in a preset report library, and controlling the target inspection report template to be presented on the inspection item interface.

[0064] The inspection item interface is used to present the target inspection report template under the inspection item corresponding to the target part type. For example, if the target part type is heart type, the inspection item interface corresponding to the heart type will present the target inspection report template corresponding to the heart type under the heart inspection item.

[0065] The report library is used to store multiple inspection report templates, each inspection report template corresponds to a corresponding part type, wherein the inspection report template can be prepared in advance by a designer and pre-stored in the report library. Determining the inspection report template corresponding to the target part type in the preset report library as the target inspection report template includes: traversing the preset report library with the target part type as an index to obtain the inspection report template corresponding to the target part type, wherein the inspection report template corresponding to the target part type is the target inspection report template.

[0066] This embodiment retrieves the target inspection report template from the report library and automatically opens the target inspection report template on the inspection item interface. This embodiment can automatically identify the target part type of the target inspection part from the medical ultrasound image, and automatically retrieve and present the target inspection report template corresponding to the target part type without the intervention of a doctor or other person, thereby avoiding the cumbersome retrieval process and improving work efficiency. In addition, this embodiment can more reliably and accurately determine the target part type of the target inspection part by fusing the part feature vector with the prior feature vector, so that the target inspection report template can be retrieved and presented more accurately and reliably.

[0067] In some embodiments, determining a site feature vector of a target examination site based on a medical ultrasound image includes: determining a site feature vector of a target examination site based on a preset ResNet model and the medical ultrasound image.

[0068] The ResNet model uses the residual connection method to deepen the number of layers of the neural network, so that the ResNet model can achieve higher model accuracy under limited computing power. Among them, the ResNet model selected in this embodiment can be one of the ResNet model series, for example, the ResNet model can be ResNet-18 or ResNet-34 or ResNet-50 or ResNet-101 or ResNet-152.

[0069] See also Figure 4 , the ResNet model includes a plurality of residual blocks, namely a first residual block 41, a second residual block 42, a third residual block 43 and a fourth residual block 44, wherein each residual block includes a first convolutional layer, a second convolutional layer, a third convolutional layer and a fourth convolutional layer connected in sequence, wherein the second convolutional layer and the third convolutional layer are depth-separable convolutional layers. In some embodiments, the second convolutional layer is a channel-by-channel convolutional layer, and the third convolutional layer is a point-by-point convolutional layer. In some embodiments, the size of the convolution kernel of the first convolutional layer is 1*1, the size of the convolution kernel of the second convolutional layer is 3*3, the size of the convolution kernel of the third convolutional layer is 1*1, and the size of the convolution kernel of the fourth convolutional layer is 1*1.

[0070] The task of classifying the target examination part of medical ultrasound images usually relies on high-dimensional features. High-dimensional features usually refer to specific visual features with more pixels and discrimination that can be extracted from medical ultrasound images to describe the properties of the target examination part. Processing high-dimensional features requires more computing resources and time, and usually requires a larger model storage space to store the weights and parameters of each layer of the ResNet model.

[0071] The ResNet model designs the first convolution layer as a regular convolution layer, but designs the second, third and fourth convolution layers as depth-separable convolution layers. This helps to reduce the number of parameters required for convolution calculation by splitting the correlation between spatial dimensions and channel (depth) dimensions while ensuring that features are not lost. It also reduces the number of parameters of the ResNet model, efficiently extracts high-dimensional features of medical ultrasound images, and improves model calculation efficiency.

[0072] In some embodiments, the prior feature vector includes a probe feature vector for representing the shape of the probe, the medical ultrasound image includes a probe image area corresponding to the probe, and determining the prior feature vector based on the medical ultrasound image includes: determining the probe feature vector based on a preset ResNet model and the probe image area. Therefore, this embodiment determines the target part type of the target inspection part based on the part feature vector and the probe feature vector, and presents a target inspection report template corresponding to the target part type.

[0073] In this embodiment, the probe image area is input into the ResNet model to obtain a probe feature vector, and then the probe feature vector is used as prior knowledge and combined with the part feature vector to determine the target part type, thereby improving the reliability and accuracy of classification.

[0074] It can be understood that, in this embodiment, the entire medical ultrasound image can be input into the ResNet model to extract the probe image area, and then the probe image area can be analyzed to obtain the probe feature vector.

[0075] It can also be understood that, in this embodiment, before inputting into the ResNet model, the probe image area can be cut out in advance from the medical ultrasound image, and then the probe image area can be input into the ResNet model. This can improve the efficiency of determining the probe feature vector, reduce the amount of calculation, and also help to improve accuracy.

[0076] In some embodiments, determining the probe feature vector according to a preset ResNet model and a probe image area includes: intercepting an image area including the probe on a medical ultrasound image as a probe image area, and determining the probe feature vector according to the ResNet model and the probe image area.

[0077] In some embodiments, intercepting an image area including a probe on a medical ultrasound image as a probe image area includes the following steps: determining an interception point on a width boundary of the medical ultrasound image, wherein the ratio of a distance from the interception point to an upper left corner point of the medical ultrasound image to the width of the medical ultrasound image is in the range of [1 / 4, 1 / 2], and horizontally intercepting the medical ultrasound image starting from the interception point to obtain a intercepted image with a minimum area, wherein the intercepted image is the probe image area.

[0078] For example, this embodiment takes a position point that is h*1 / 3 away from the upper left corner point of the medical ultrasound image as the interception point, where h is the width of the medical ultrasound image, and then horizontally divides the medical ultrasound image with the interception point as the starting point to obtain two parts of the intercepted image. This embodiment selects the intercepted image with the smallest area as the probe image area.

[0079] See also Figure 5a When a doctor uses a convex array probe to perform an ultrasound examination on the abdomen, the medical ultrasound image includes a convex array probe image region 5a0 corresponding to the convex array probe, wherein the area of ​​the convex array probe image region 5a0 is one third of the area of ​​the medical ultrasound image. In this embodiment, the convex array probe image region 5a0 may be input into the ResNet model to obtain a probe feature vector of the convex array probe.

[0080] See also Figure 5b When a doctor uses a linear array probe to perform an ultrasonic examination on a superficial organ, the medical ultrasonic image includes a linear array probe image region 5b0 corresponding to the linear array probe, wherein the area of ​​the linear array probe image region 5b0 is one third of the area of ​​the medical ultrasonic image. In this embodiment, the linear array probe image region 5b0 may be input into the ResNet model to obtain a probe feature vector of the linear array probe.

[0081] See also Figure 5c When a doctor uses a cardiac probe to perform an ultrasonic examination on the heart, the medical ultrasonic image includes a cardiac probe image region 5c0 corresponding to the cardiac probe, wherein the area of ​​the cardiac probe image region 5c0 is one third of the area of ​​the medical ultrasonic image. In this embodiment, the cardiac probe image region 5c0 may be input into the ResNet model to obtain a probe feature vector of the cardiac probe.

[0082] See also Figure 5d When a doctor uses an intracavity probe to perform an ultrasound examination on the vagina, the medical ultrasound image includes an intracavity probe image region 5d0 corresponding to the intracavity probe, wherein the area of ​​the intracavity probe image region 5d0 is one third of the area of ​​the medical ultrasound image. In this embodiment, the intracavity probe image region 5d0 may be input into the ResNet model to obtain a probe feature vector of the intracavity probe.

[0083] This embodiment intercepts the image area containing the probe on the medical ultrasound image as the probe image area, which is beneficial for the ResNet model to quickly and efficiently determine the probe feature vector, thereby improving the efficiency and accuracy of retrieving the target inspection report template.

[0084] This embodiment processes the medical ultrasound image and the probe image area based on the ResNet-50 model. The processing process is as follows:

[0085] Firstly, a medical ultrasound image or a probe image region is used as input data and first passes through a preprocessing layer to obtain preprocessed input data, wherein the preprocessing includes image normalization processing and zero filling processing.

[0086] Secondly, the preprocessed input data is input into the residual block, wherein the residual block includes a depth-separable convolution operation and a batch normalization operation. The preprocessed input data is first processed by the pooling layer, and then input into the first convolution layer. The first convolution layer performs the first convolution on the pooled input data, and then inputs into the batch normalization layer. The batch normalization layer performs a batch normalization on the input data processed by the first convolution, and then inputs into the relu activation function. The relu activation function activates the input data, and the activated input data is input into the second convolution layer. The operation of the second convolution layer can be carried out in sequence according to the above method, which will not be repeated here.

[0087] Then, by processing the input data through multiple residual blocks, the part feature vector and the probe feature vector can be obtained. On the one hand, the use of residual connection can enhance the expression ability of features and extract more reliable and accurate part feature vectors and probe feature vectors. On the other hand, since the residual block uses a deep separable convolutional layer, it is beneficial to reduce the amount of data calculation and improve the efficiency of retrieving the target inspection report template.

[0088] Finally, this embodiment inputs the part feature vector and the probe feature vector into the classification model to obtain the target part type.

[0089] It should be noted that, in each of the above-mentioned embodiments, there is not necessarily a certain order between the above-mentioned steps. A person skilled in the art can understand, based on the description of the embodiments of the present disclosure, that in different embodiments, the above-mentioned steps may have different execution orders, that is, they may be executed in parallel, may be executed interchangeably, and so on.

[0090] As another aspect of the embodiments of the present disclosure, the embodiments of the present disclosure provide a device for presenting an inspection report template. The device for presenting an inspection report template may be a software module, which includes several instructions stored in a memory, and a processor may access the memory and call the instructions for execution to complete the method for presenting an inspection report template described in each of the above embodiments.

[0091] In some embodiments, the presentation device of the inspection report template can also be constructed by hardware devices. For example, the presentation device of the inspection report template can be constructed by one or more chips, and each chip can work in coordination with each other to complete the presentation method of the inspection report template described in each of the above embodiments. For another example, the presentation device of the inspection report template can also be constructed by various logic devices, such as a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a single-chip microcomputer, an ARM (Acorn RISC Machine) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination of these components.

[0092] See also Figure 6 The inspection report template presentation device 600 includes an image acquisition module 61 , a feature determination module 62 , a type determination module 63 and a template presentation module 64 .

[0093] The image acquisition module 61 is used to acquire a medical ultrasound image, wherein the medical ultrasound image includes an image area corresponding to the target inspection part. The feature determination module 62 is used to determine the part feature vector and the prior feature vector of the target inspection part according to the medical ultrasound image, wherein the prior feature vector is used to represent the features of the factors associated with the target inspection part. The type determination module 63 is used to determine the target part type of the target inspection part according to the part feature vector and the prior feature vector. The template presentation module 64 is used to present a target inspection report template corresponding to the target part type.

[0094] This embodiment can automatically identify the target part type of the target inspection part from the medical ultrasound image, and automatically retrieve and present the target inspection report template corresponding to the target part type, without the intervention of doctors or other people, thereby avoiding the cumbersome retrieval process and improving work efficiency. In addition, this embodiment can more reliably and accurately determine the target part type of the target inspection part by fusing the part feature vector with the prior feature vector, so that the target inspection report template can be retrieved and presented more accurately and reliably.

[0095] In some embodiments, the feature determination module 62 is specifically used to determine a site feature vector of a target examination site based on a preset ResNet model and a medical ultrasound image.

[0096] In some embodiments, the prior feature vector includes a probe feature vector for representing the probe shape, the medical ultrasound image includes a probe image area corresponding to the probe, and the feature determination module 62 is specifically used to determine the probe feature vector according to a preset ResNet model and the probe image area.

[0097] In some embodiments, the feature determination module 62 is further specifically used to: intercept an image area including the probe on the medical ultrasound image as a probe image area, and determine a probe feature vector according to the ResNet model and the probe image area.

[0098] In some embodiments, the feature determination module 62 is also specifically used to: determine a cutoff point on the width boundary of the medical ultrasound image, the ratio of the distance from the cutoff point to the upper left corner point of the medical ultrasound image to the width of the medical ultrasound image is in the range of [1 / 4, 1 / 2], and the medical ultrasound image is horizontally cut off from the cutoff point to obtain a cutoff image with the smallest area, and the cutoff image is the probe image area.

[0099] In some embodiments, the type determination module 63 is specifically used to: splice the part feature vector and the prior feature vector to obtain a spliced ​​feature vector, and determine the target part type of the target inspection part according to a preset classification model and the spliced ​​feature vector.

[0100] In some embodiments, the ResNet model includes multiple residual blocks, which include a pooling layer, a first convolutional layer, a second convolutional layer, a third convolutional layer and a fourth convolutional layer connected in sequence, wherein the second convolutional layer and the third convolutional layer are depth-separable convolutional layers.

[0101] In some embodiments, the size of the convolution kernel of the first convolution layer is 1*1, the size of the convolution kernel of the second convolution layer is 3*3, the size of the convolution kernel of the third convolution layer is 1*1, and the size of the convolution kernel of the fourth convolution layer is 1*1.

[0102] In some embodiments, the template presentation module 64 is specifically used to: determine the inspection item interface corresponding to the target part type, determine the inspection report template corresponding to the target part type in the preset report library as the target inspection report template, and control the presentation of the target inspection report on the inspection item interface.

[0103] It should be noted that the above-mentioned inspection report template presentation device can execute the inspection report template presentation method provided in the embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method. For technical details not fully described in the inspection report template presentation device implementation, please refer to the inspection report template presentation method provided in the embodiment of the present disclosure.

[0104] See also Figure 7 , Figure 7 The circuit structure diagram of an electronic device provided in an embodiment of the present disclosure is shown, wherein the electronic device can be any suitable type of device or electronic product, for example, the electronic device includes a computer or a mobile phone and other devices with logic calculation and analysis functions. Figure 7 As shown, the electronic device 700 includes one or more processors 71 and a memory 72. Figure 7 A processor 71 is taken as an example.

[0105] The processor 71 and the memory 72 may be connected via a bus or other means. Figure 7 The example of connecting through bus is taken in the following.

[0106] The memory 72, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions / modules corresponding to the method for presenting the inspection report template in the embodiment of the present disclosure. The processor 71 executes various functional applications and data processing of the inspection report template presentation device by running the non-volatile software programs, instructions and modules stored in the memory 72, that is, realizes the inspection report template presentation method provided by the above method embodiment and the functions of each module or unit of the above device embodiment.

[0107] The memory 72 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 72 may optionally include a memory remotely arranged relative to the processor 71, and these remote memories may be connected to the processor 71 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0108] The program instructions / modules are stored in the memory 72, and when executed by the one or more processors 71, the method for presenting the inspection report template in any of the above method embodiments is executed.

[0109] The present disclosure also provides a non-volatile computer storage medium, wherein the non-volatile computer storage medium stores computer executable instructions, and the computer executable instructions are executed by one or more processors, such as Figure 7 A processor 71 in the embodiment may enable the one or more processors to execute the method for presenting the inspection report template in any of the above method embodiments.

[0110] An embodiment of the present disclosure also provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by an electronic device, the electronic device executes any one of the methods for presenting an inspection report template.

[0111] The above described device or equipment embodiments are merely illustrative, wherein the unit modules described as separate components may or may not be physically separated, and the components displayed as module units may or may not be physical units, that is, they may be located in one place, or may be distributed on multiple network module units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0112] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure, rather than to limit them. Under the concept of the present disclosure, the technical features in the above embodiments or different embodiments may also be combined, the steps may be implemented in any order, and there are many other changes in different aspects of the present disclosure as described above, which are not provided in detail for the sake of simplicity. Although the present disclosure has been described in detail with reference to the aforementioned embodiments, a person of ordinary skill in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present disclosure.

Claims

1. A method for presenting an inspection report template, characterized in that: include: Acquire a medical ultrasound image, wherein the medical ultrasound image includes an image area corresponding to the target examination site; Determining a site feature vector and a priori feature vector of the target examination site according to the medical ultrasound image, wherein the priori feature vector is used to represent the characteristics of factors associated with the target examination site; Determining a target part type of the target inspection part according to the part feature vector and the priori feature vector; A target inspection report template corresponding to the target site type is presented.

2. The presentation method according to claim 1, characterized in that: The step of determining the part feature vector of the target examination part according to the medical ultrasound image comprises: The part feature vector of the target examination part is determined according to a preset ResNet model and the medical ultrasound image.

3. The presentation method according to claim 1, characterized in that: The prior feature vector includes a probe feature vector for representing a probe shape, the medical ultrasound image includes a probe image region corresponding to the probe, and determining the prior feature vector according to the medical ultrasound image includes: The probe feature vector is determined according to the preset ResNet model and the probe image area.

4. The presentation method according to claim 3, characterized in that: Determining the probe feature vector according to the preset ResNet model and the probe image area includes: intercepting an image area including the probe on the medical ultrasound image as a probe image area; A probe feature vector is determined according to the ResNet model and the probe image area.

5. The presentation method according to claim 4, characterized in that: The method of intercepting an image area including the probe on the medical ultrasound image as a probe image area comprises: Determine a cutoff point on the width boundary of the medical ultrasound image, wherein a ratio of a distance from the cutoff point to an upper left corner point of the medical ultrasound image to a width of the medical ultrasound image is within a range of [1 / 4, 1 / 2]; The medical ultrasound image is horizontally intercepted starting from the interception point to obtain an intercepted image with a minimum area, and the intercepted image is the probe image area.

6. The presentation method according to any one of claims 2 to 5, characterized in that: The ResNet model includes multiple residual blocks, each of which includes a first convolutional layer, a second convolutional layer, a third convolutional layer, and a fourth convolutional layer connected in sequence, wherein the second convolutional layer and the third convolutional layer are depth-separable convolutional layers.

7. The presentation method according to claim 6, characterized in that: The size of the convolution kernel of the first convolution layer is 1*1, the size of the convolution kernel of the second convolution layer is 3*3, the size of the convolution kernel of the third convolution layer is 1*1, and the size of the convolution kernel of the fourth convolution layer is 1*1.

8. The presentation method according to any one of claims 1 to 5, characterized in that: Determining the target part type of the target inspection part according to the part feature vector and the prior feature vector includes: splicing the part feature vector and the prior feature vector to obtain a spliced ​​feature vector; The target part type of the target inspection part is determined according to a preset classification model and the spliced ​​feature vector.

9. The presentation method according to any one of claims 1 to 5, characterized in that: The target inspection report template corresponding to the target part type includes: Determine an inspection item interface corresponding to the target part type; Determining, in a preset report library, an inspection report template corresponding to the target part type as a target inspection report template; Control the target inspection report to be presented on the inspection item interface.

10. A non-volatile readable storage medium, characterized in that: The non-volatile readable storage medium stores computer executable instructions, and the computer executable instructions are used to enable an electronic device to execute the method for presenting an inspection report template as described in any one of claims 1 to 9.

11. An electronic device, characterized in that: include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for presenting an inspection report template as described in any one of claims 1 to 9.

Images