TWM551477U - An ultrasound diagnostic apparatus suitable for artificial intelligence imaging analysis - Google Patents

Abstract

An ultrasound diagnostic apparatus is provided. The ultrasound diagnostic apparatus comprises an ultrasound signal transceiver, a plurality of motion and/or rotation detectors (e.g., gyroscope) and a networking system. When the ultrasound signal capture function of this apparatus is enabled, the internal built-in motion and/or rotation detectors (e.g., gyroscope) will provide their own motion and/or rotation information for each captured two-dimension ultrasound image. Afterwards, a networking system will upload these two-dimension ultrasound images with motion and/or rotation information to a cloud computing server. The cloud computing server has an algorithm of artificial intelligence imaging analysis which can put all these two-dimension ultrasound images with motion and/or rotation information into a tree dimension image buffer. Further, the algorithm can deliver brand-new image cross sections with medical diagnostic values based on the guideline of medical ultrasound diagnosis. At last, the delivered images can be used for the clinical medical diagnosis or the artificial intelligence and machine learning.

Description

Ultrasonic diagnostic device suitable for artificial intelligence image analysis

The novel is a new type of ultrasonic diagnostic device, in particular, relates to an image displacement and rotation angle detecting system for comparing original two-dimensional ultrasonic images through image matching, scaling, and spatial conversion. The algorithm produces a new two-dimensional ultrasound image that is different from the original two-dimensional ultrasound image and provides more medical diagnostic information.

The medical ultrasonic diagnostic device has many characteristics such as convenience, non-aggressive, non-radioactive, easy to obtain, etc., so that the current ultrasonic diagnosis is widely used in every clinical department (except the psychiatric department), and according to each department The clinical guidelines of the Sonic Medical Association suggest that the application of ultrasound is based on two-dimensional planar images for the examination of each aspect. The three-dimensional images can only provide reference purposes and cannot be used as a basis for physician diagnosis. Therefore, as shown in Figure 2 The prior device (US 2009/0306509 A1) combines an ultrasonic diagnostic device with an image displacement and rotation angle detection system, the main purpose of which is that imaging of three-dimensional images does not effectively enhance the clinical diagnostic value.

Therefore, the main scope of the present invention is to apply this image displacement and rotation angle detection system to generate a new two-dimensional ultrasound image profile to increase the judgment of clinical diagnosis.

The new problem to be solved is that the current common two-dimensional ultrasonic scanning application is applied to the examination of human organs, and often requires operators (such as doctors, ultrasonic technicians) to make ten or even dozens of organs according to clinical guidelines. The observation of the profile can be completed. This process often leads to inconsistency in the inspection results due to the experience of different operators or the differences of organs in different patients. Therefore, the novel system transmits an image displacement and rotation angle detection system. And algorithms for image matching, scaling, and spatial conversion to allow the operator to generate a profile of the human organ referred to in each clinical guidelines through a simple linear scan, which effectively reduces the operator's The instrument operation time can also eliminate the operator's manual misjudgment.

In order to solve the manual misjudgment of the operator of the ultrasonic diagnostic device, the present invention solves this problem by an image displacement and rotation angle detection system, and an image comparison, scaling, and spatial conversion algorithm, image displacement and rotation angle detection. The system can accurately know the relative position of each two-dimensional image generated by the operator in a simple linear scan, and then the novel device will network the collected two-dimensional images and their relative positions with each other. The device is passed to the cloud computing server, and the cloud computing server places the two-dimensional images and their relative positions in a three-dimensional image storage area, and finally generates images through image matching, scaling, and spatial conversion algorithms. A section of the human organ referred to in the clinical guidelines.

The previous technology also has an image displacement and rotation angle detection system, but its purpose It is mainly used for imaging 3D images. However, the general recognition in the medical field is that the imaging of 3D images can not effectively improve the value of clinical diagnosis. Therefore, the effect of the novel is to effectively reduce the operator's instrument operation time and eliminate the operator. Due to unfamiliarity with the operation of the instrument, or lack of experience, or the special organ characterization of the subject, there is no complete manual inspection of the human organ profile as mentioned in the clinical guidelines.

1‧‧‧Specific embodiment device of the present invention

11,21‧‧‧Ultrasonic Transducer

12,22‧‧‧Supersonic Signal Transceiver

13,23‧‧‧Two-dimensional ultrasonic image generator

14,24‧‧‧Image Displacement and Rotation Angle Detection System

15‧‧‧Networking system

16‧‧‧Cloud computing server

17‧‧‧Two-dimensional ultrasonic image profile generation module

25‧‧‧3D Ultrasonic Image Generator

Horizontal displacement of X.‧‧ front and rear 2D images in the image plane

Vertical displacement of Z‧‧‧ front and rear 2D images in the image plane

Y‧‧‧ Displacement of front and rear 2D images

The angle of rotation of the α·‧‧ front and rear 2D images on the X+Z axis

The angle of rotation of the β·‧‧ front and rear 2D images on the Y+Z axis

R‧‧‧ Rotation angle of the front and rear 2D images on the X+Y axis

51‧‧‧Two-dimensional images acquired in the Y direction in linear scanning mode

52‧‧‧Three-dimensional space image storage area

53‧‧‧Image calculated by the 2D ultrasonic image profile generation module on the X and Y planes

54‧‧‧The position where the ultrasound transducer needs to be placed in the X and Y planes

55‧‧‧Image rotated by β angle calculated by the 2D ultrasonic image profile generation module

56‧‧‧The position where the ultrasound transducer is required to obtain an image with a β-angle rotation

Figure 1: Schematic diagram showing an ultrasonic diagnostic apparatus according to the specific implementation of the present invention

Figure 2: Schematic diagram showing a conventional ultrasonic diagnostic device with image displacement and rotation angle detection system

Figure 3: shows all possible image displacements and rotation angles

Figure 4: shows the scanning method of the common handheld ultrasonic diagnostic device

Figure 5: Schematic diagram showing the new two-dimensional ultrasonic image generated by the new model

According to a specific embodiment 1 of the present invention, an ultrasonic diagnostic device having a networking function and a cloud computing server are provided. As shown in FIG. 1 , the device 1 of the specific embodiment of the present invention includes an ultrasonic transducer 11 and a supersonic signal transmitting and receiving device. The device 12, the two-dimensional ultrasonic image generator 13, the image displacement and rotation angle detecting system 14, the networking system 15, the cloud computing server 16, and the two-dimensional ultrasonic image profile generating module 17.

In this embodiment, the ultrasonic transducer 11 is combined with the ultrasonic signal transceiver 12 and the two-dimensional ultrasonic image generator 13 to constitute a conventional ultrasonic diagnostic device. The image displacement and rotation angle detecting system 14 can be a gyro sensor commonly used in mobile phone devices, and can be in multiple forms to provide more accurate displacement and rotation angle detection. When the operator scans the human organ in a hand-held manner, the displacement or rotation angle between the images will be as shown in Figure 3: the horizontal displacement of the front and back 2D images in the image plane X, the front and back 2D images in the image The vertical displacement Z in the plane, the displacement of the front and back two-dimensional images with each other, the rotation angle α of the front and back two-dimensional images on the X+Z axis, the rotation angle of the front and rear two-dimensional images on the Y+Z axis β, before, The rotation angle r of the rear 2D image on the X+Y axis. As for the common scanning method shown in FIG. 4, the linear scanning is a method commonly used for abdominal ultrasound scanning, and the scanning direction is perpendicular to the direction of the ultrasonic transducer 11, that is, the front and rear two-dimensional images shown in FIG. The displacement between the front and back two-dimensional images is the displacement of the front and back two-dimensional images in the displacement Y direction. The difference between the sector scanning and the linear scanning is that the amplitude can be smaller. A large range of areas is observed, and the amplitude of the swing can be represented by the rotation angle β of the front and back two-dimensional images of FIG. 3 in the Y+Z axis, representing the rotation angle of the front and back two-dimensional images in the Y+Z axis, and the arrow type scanning. It is often used for gynecological ultrasound diagnosis. The scanning method is to rotate the probe at the angle of rotation r of the X+Y axis along the front and back 2D images of Figure 3. It also represents the rotation angle of the front and back 2D images in the X+Y axis. .

In practical applications, taking the linear scan most commonly in Figure 4 as an example, when the operator moves along the displacement Y direction of the front and back two-dimensional images, it may be due to the patient's respiratory vibration or operation. The jitter caused by the handheld device itself may cause the horizontal displacement X, the front and back two-dimensional images of the front and back two-dimensional images in the image plane except the displacement of the front and rear two-dimensional images. The vertical displacement Z in the image plane, the rotation angle α of the front and back 2D images in the X+Z axis, the rotation angle β of the front and back 2D images in the Y+Z axis, and The displacement of the 2D image in the undesired image plane such as the rotation angle r of the X+Y axis or the rotation angle between the image planes, so the relative positional relationship between the front and back 2D images requires an image displacement and rotation angle detection system. These displacements or angles are presented. With these displacement and angle information, each of the captured two-dimensional ultrasonic images can be correctly placed one after another in the three-dimensional image storage area 52.

For example, in the three-dimensional image storage area 52, as shown in FIG. 5, the two-dimensional image 51 in the Y direction is stored in the linear scanning mode, but it is assumed to be in the clinical guidelines of the Ultrasonic Medical Association. It is also necessary to observe the images on the X and Y planes. It is conventionally necessary to place the ultrasonic transducer 11 on the image in which the X and Y planes are conventionally required to be placed at the position 54 of the ultrasonic transducer. The observed image, however, can be extracted from the image on the X and Y planes through the image data in the three-dimensional image storage area 52, and rendered equivalent by appropriate image scaling, interpolation, and the like. The ultrasonic transducer 11 is placed on a conventional image in which the image on the X and Y planes needs to be placed at the position 54 of the ultrasonic transducer. This image is a cross-sectional view of the two-dimensional ultrasonic image. The image 53 calculated on the X and Y planes is generated by the module. Similarly, if the operator wants to observe the image rotated by the β angle, it is conventional to place the ultrasonic transducer 11 in the conventional angle to obtain the β angle. Rotating images require supersonic The position 56 of the transducer can be obtained to obtain the image to be observed, but the image can be extracted from the image rotated by the β angle through the image data in the three-dimensional image storage area 52. Figure 5 is an image 55 rotated by a two-dimensional ultrasonic image profile generation module. The image processing algorithm developed by the present invention can generate all clinical guidelines after a linear scan. The human organ profile diagram mentioned in the above can greatly reduce the operator's device operation time and enable the operator to serve more patients than the conventional method. It can reduce the amount of ultrasonic exposure of the subject, and also reduce the profile of some human organs that are missed by different operators. In addition, the cloud computing server mentioned in the present invention can be collected along with The more the amount of image data, the more the image can be cross-matched, that is, through the artificial intelligence machine learning method, how to determine the doctor's method of "problem image profile" or "normal image profile" into the machine In this way, the machine can gradually have the ability of the doctor to judge, such as providing health warnings or assisting physicians.

Finally, with regard to the part of the networked device, the traditional ultrasonic diagnostic device has a limited ability of hardware and software, so it cannot be processed for a long time, and the capacity for storing images is also limited. Therefore, the present invention uses a networked device. To solve the above problems, but if the software and hardware capabilities of the ultrasonic diagnostic device are broken and the capacity of the stored image is greatly improved, there is an opportunity to complete the original computing in the cloud in the ultrasonic diagnostic device. The networked device can be omitted from the functions on the device.

The features and spirit of the present invention are intended to be more apparent from the detailed description of the preferred embodiments. On the contrary, the intention is to cover various modifications and equivalent arrangements within the scope of the claimed invention.

11‧‧‧Ultrasonic Transducer

12‧‧‧Supersonic Signal Transceiver

13‧‧‧Two-dimensional ultrasonic image generator

14‧‧‧Image Displacement and Rotation Angle Detection System

15‧‧‧Networking system

16‧‧‧Cloud computing server

17‧‧‧Two-dimensional ultrasonic image profile generation module

Claims (3)

Ultrasonic diagnostic device suitable for artificial intelligence image analysis, including ultrasonic transducer, ultrasonic signal transceiver, two-dimensional ultrasonic image generator, image displacement and rotation angle detection system, networked system, cloud computing server And a two-dimensional ultrasonic image profile generation module, the two-dimensional ultrasonic image generator can generate a two-dimensional ultrasonic image through the ultrasonic transducer and the ultrasonic signal transceiver, and then the image displacement and rotation The angle detection system can assign each two-dimensional ultrasonic image to its corresponding image displacement and rotation angle information. Finally, the image data and the displacement and rotation angle data are transmitted to the cloud computing server through the networked system; the cloud computing servo The device has a two-dimensional ultrasonic image profile generation module, which can generate the human organ profile mentioned in each clinical guidance suggestion through image comparison, scaling, and spatial conversion algorithms. The ultrasonic diagnostic apparatus of claim 1, wherein the networked system is a wired or wireless connectable network system. The ultrasonic diagnostic apparatus according to claim 1, wherein the cloud computing server is a computing system capable of being responsible for image processing related algorithms.