TWM594253U - Intelligent analysis device for liver tumor - Google Patents

Abstract

A liver tumor intelligent analysis device includes a connected ultrasonic detection module and an analysis module equipped with a machine learning algorithm to judge the benign and malignant risks of liver tumors, wherein the analysis module further includes connected image acquisition A unit, a reference storage unit, a control unit, a liver tumor labeling unit, a classification unit, a comparison unit, and a prediction report generation unit for benign and malignant risk of liver tumors. The liver tumor intelligent analysis device adopts an ultrasound abdominal image for experienced abdominal specialists to mark the ultrasound images with liver tumor image points. The parameters and coefficients of these empirical data are trained by machine learning algorithms to establish an accuracy rate Up to 86% of inferential models. Therefore, the scan image can be detected through ultrasound, and the intelligent analysis device for liver tumors can immediately help the doctor or ultrasound technician to interpret the benign and malignant risks of liver tumors, so as to provide a reference for the doctor to diagnose liver tumor types.

Description

Liver tumor intelligent analysis device

This creation is about a smart analysis device for liver tumors, especially a The sonic detection module and the analysis module equipped with machine learning algorithms are used to judge the benign and malignant risks of liver tumors, in particular, it can scan images through ultrasound detection, and through this liver tumor intelligent analysis device, it can immediately help the doctor or ultrasound technician to interpret the liver. The risk of benign and malignant tumors, thus providing doctors with a reference basis for the diagnosis of liver tumor types.

Liver cancer ranks the fourth leading cause of cancer death in the world and the second leading cause of cancer death among Chinese Bit. Common causes of liver cancer in Asia are mostly caused by hepatitis B virus, hepatitis C virus and aflatoxin. European and American countries are commonly caused by hepatitis C virus, and liver cancer caused by steatohepatitis, diabetes, and high triglycerides is also produced. It is getting worse.

Surgery is currently the most direct treatment for liver cancer, but early diagnosis of liver cancer And the indicators of postoperative patient-related prognosis are also very important topics. Patients with liver cancer diagnosed early usually have more treatment options, and the efficacy of treatment is often reflected in the patient's survival rate. Therefore, regular inspection and early diagnosis and early treatment is the key to improving the quality of life of patients and prolonging their survival.

Early diagnosis in addition to blood tests to check liver function, hepatitis B virus and hepatitis A virus Bai Wai, the study pointed out that abdominal ultrasound is one of the important examinations for intact liver disease. According to the early research of Professor Xu Jinchuan, an emeritus professor of National Taiwan University, he pointed out that one-third of patients with small liver cancer had a blood test for liver cancer index-alpha fetal protein. To be normal, it must be supplemented by ultrasound examination to detect liver cancer early, plus abdominal ultrasound examination Features, fast, convenient and no radiation, become an important tool for liver cancer screening.

The diagnosis of liver cancer is different from other cancers. The film can be directly diagnosed by imaging examination, such as: abdominal ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI), etc., and its sensitivity (sensitivity) And specificity (specificity): 0.78 ~ 0.73 and 0.89 ~ 0.93, 0.84 ~ 0.83 and 0.99 ~ 0.91, and 0.83 and 0.88.

Although ultrasound inspection has its convenience, it also has limitations, such as: the operator’s experience Factors such as test accuracy, patient obesity, liver fibrosis or liver cirrhosis will affect the accuracy of ultrasound. Therefore, when ultrasound examination is suspected to be a malignant tumor, most of the second imaging examinations such as CT or MRI are used to confirm the diagnosis In addition to the high medical costs and lengthy inspection schedules, the CT examinations also have more radiation exposure considerations.

In view of this, an artificial intelligence analysis technology has been urgently needed in related technical fields A device that can effectively analyze the benign and malignant liver tumors by ultrasonic testing; compared with MRI and/or computed tomography, the intelligent analysis device of liver tumors will assist the accuracy of ultrasound image analysis and the accuracy of CT and MRI Rather, to help physicians use ultrasound to quickly diagnose liver tumor types accurately.

The main purpose of this creation is to overcome the above-mentioned problems encountered by conventional skills and Provide an ultrasonic detection module and an analysis module equipped with a machine learning algorithm to determine the nature of liver tumors, the accuracy rate can be as high as 86% and CT or MRI accuracy is similar, thereby assisting physicians to use radiation-free and safe Ultrasonic examination to quickly and accurately diagnose liver tumor types.

In order to achieve the above purpose, this creation is a smart analysis device for liver tumor, which is a package Including: an ultrasound detection module, which includes an ultrasound probe for transmitting ultrasound to scan the area corresponding to the outside of a subject and obtain an ultrasound image of a target liver tumor of the subject; and An analysis module, which is connected to the ultrasound detection module, acquires an ultrasound image of the target liver tumor of the subject formed by imaging by the ultrasound detection module by an image acquisition unit, and uses a reference storage unit There are several reference ultrasound images of liver tumors, which include ultrasound images of benign and malignant liver tumors. The analysis module stores a program, where when the program is executed by a control unit, the program can provide clinical personnel to judge the affected The type of liver tumor in one of the subjects and the probability of benign and malignant risk of liver tumor in this subject.

The program includes: a liver and liver tumor labeling unit, used to build based on empirical rules Establish specific rules to automatically mark the ultrasound images of liver tumors with tumor image points from the reference ultrasound images of the liver tumors and identify the type of liver tumors; a classification unit is used to classify these Ultrasound images of liver tumors in the tumor image spot area are trained using a machine learning algorithm to build an inference model; and a comparison unit for analyzing the ultrasound images of the target liver tumor with the inference model to provide The clinical staff judges the type of liver tumor in one of the subjects and predicts the risk of benign and malignant liver tumor in the subject.

In the above embodiment of this creation, the image acquisition unit is a digital video interface (Digital Visual Interface, DVI).

In the above embodiment of the present invention, the control unit is a central processor (Central Processing Unit, CPU).

In the above embodiment of this writing, the liver tumor labeling unit is based on Coefficients and/or parameters based on empirical data are automatically marked from these reference ultrasound images of liver tumors Ultrasound images of liver tumors with tumor image points are generated.

In the above example of this writing, the liver tumor type includes benign tumor and malignant Tumor.

In the above embodiment of this writing, the analysis module further includes a prediction of liver tumors Sexual risk report generating unit connected to the control unit, the control unit can provide the clinical unit to determine the type of liver tumor of the subject and predict the benign and malignant risk of liver tumor of the subject The probability is input into the generating unit for predicting the benign and malignant risk of liver tumors to produce a diagnostic report that assists physicians in judging the nature of liver tumors.

Please refer to "Picture 1", it is a better way to create a liver tumor intelligent analysis device Block diagram of the embodiment. As shown in the figure: This creation is a liver tumor intelligent analysis device, which includes an ultrasound detection module 1 and an analysis module 2. The ultrasonic detection module 1 mentioned above includes an ultrasonic probe 11. The analysis module 2 is connected to the ultrasonic detection module 1 and includes an image acquisition unit 21, a reference storage unit 22, a control unit 23, a liver tumor marking unit 24, a classification unit 25, and a comparison unit 26 And a report generation unit 27 for predicting the benign and malignant risk of liver tumors. Wherein, the control unit 23 may be a central processing unit (Central Processing Unit, CPU), which is used as the image acquisition unit 21, the reference storage unit 22, the liver tumor labeling unit 24, the classification unit 25, the comparison unit 26 and the predicted hepatic tumor benign and malignant risk report generation unit 27, operation, control, processing, encoding, decoding and various driving instructions are issued during operation. If so, a new liver tumor intelligent analysis device is constituted by the device disclosed above.

When this work is used, the liver tumor intelligent analysis device of this work can be implemented in a In the computer, and the control unit 23 is the central processor of the computer, the liver tumor labeling unit 24, the classification unit 25, the comparison unit 26 and the predicted liver tumor benign and malignant risk report generation unit 27 may be a program in the computer , And stored in the hard disk or memory, and the image acquisition unit 21 is a computer's digital visual interface (Digital Visual Interface, DVI), and the reference storage unit 22 can be a hard disk, and further has a screen, a mouse and The keyboard serves as related output and operation. In addition, the intelligent liver tumor analysis device of the present invention can also be implemented in a server.

When used, the ultrasonic probe 11 of the ultrasonic detection module 1 provides transmission Ultrasound scans the area outside the subject corresponding to the liver to obtain ultrasound images of the liver tumors of the subject. During the scan, the doctor may also detect at least one ultrasound image of a suspicious tumor and make a selection It is an ultrasound image of a target liver tumor.

The analysis module 2 can be acquired by the image acquisition unit 21 through the ultrasonic detection The subject's target liver tumor ultrasound image formed by imaging module 1 is formed, and the reference storage unit 22 stores several reference liver tumor ultrasound images, which include benign and malignant ultrasound images of liver tumors. The analysis module 2 stores a program, where when the program is executed by the control unit 23, the program can provide clinical personnel to determine the type of liver tumor of the subject and predict the benign and malignant risk of liver tumor of the subject Probability, the program includes the liver tumor labeling unit 24, the classification unit 25, the comparison unit 26 and the predicted liver tumor benign and malignant risk report generation unit 27.

The liver tumor labeling unit 24 is based on coefficients based on data with matching experience and/or Or parameters, automatically mark the ultrasound image of the liver tumor with the tumor image point area from the several reference liver tumor ultrasound images and identify the type of liver tumor. For example: the liver tumor labeling unit 24 can be labeled based on the experience of a physician. The classification unit 25 then trains the ultrasound images of liver tumors with tumor image points using a machine learning algorithm to build an inference model. The comparison unit 26 analyzes the target liver tumor ultrasound image with the inferential model to provide clinical staff with a judgment on the nature of one of the subject’s liver tumors and further predicts the benign and malignant risk of one of the subject’s liver tumors Probability. Finally, through the predicted liver tumor benign and malignant risk report generating unit 27, the clinical unit judged by the comparison unit 26 to provide a clinical staff to determine the type of a liver tumor of the subject and predict the probability of benign and malignant risk of a liver tumor of the subject Up to the 27th report generation unit for predicting the benign and malignant risk of liver tumors, a diagnosis report to help physicians judge the nature of liver tumors is produced.

In this way, the intelligent analysis device for liver tumors mentioned in this creation is based on rich experience The abdominal ultrasound specialist marks the ultrasound images with liver tumor image points. The parameters and coefficients of these empirical data are trained by machine learning algorithms to establish an inference model with an accuracy of up to 86%. Therefore, the scan image can be detected through ultrasound, and the intelligent analysis device for liver tumors can immediately help the doctor or ultrasound technician to interpret the benign and malignant risks of liver tumors, so as to provide a reference for the doctor to diagnose liver tumor types.

In summary, this creation is a smart analysis device for liver tumors, which can effectively improve The various shortcomings are used to determine the benign and malignant risk of liver tumors with ultrasonic detection modules and analysis modules equipped with machine learning algorithms. Based on coefficients and/or parameters based on matching experience data, tumor image points can be used. The ultrasound images of liver tumors are marked, and an inference model with an accuracy rate of up to 86% is established through the training of machine learning algorithms, thereby assisting physicians to use radiation-free and safe ultrasound examination to quickly diagnose liver tumor types accurately. In turn, the creation of this creation can be more advanced, more practical, and more in line with the needs of users, and it has indeed met the requirements for new type patent applications, and the patent application is filed according to law.

However, the above are only the preferred embodiments of this creation, and should not be limited to this The scope of the implementation of the original creation is fixed; therefore, all Simple equivalent changes and modifications should still fall within the scope of this creative patent.

1: Ultrasonic detection module 11: Ultrasonic probe 2: Analysis module 21: Image acquisition unit 22: Reference storage unit 23: Control unit 24: Liver tumor marking unit 25: taxon 26: Comparison unit 27: Generating report unit for predicting the risk of liver tumor

Fig. 1 is a block diagram of a preferred embodiment of the smart liver tumor analysis device.

1: Control unit

1: Ultrasonic detection module

11: Ultrasonic probe

2: Analysis module

21: Image acquisition unit

22: Reference storage unit

23: Control unit

24: Liver tumor marking unit

25: taxon

26: Comparison unit

27: Generating report unit for predicting the risk of liver tumors

Claims (6)

A liver tumor intelligent analysis device, including: An ultrasound detection module, which includes an ultrasound probe, is used to emit ultrasound to scan a region corresponding to the liver outside of a subject, and obtain an ultrasound image of a target liver tumor of the subject; and An analysis module, which is connected to the ultrasound detection module, acquires an ultrasound image of the target liver tumor of the subject formed by imaging by the ultrasound detection module by an image acquisition unit, and uses a reference storage unit There are several reference ultrasound images of liver tumors, which include ultrasound images of benign and malignant liver tumors. The analysis module stores a program, where when the program is executed by a control unit, the program can provide clinical personnel to judge the affected The liver tumor type of one of the test subjects and the probability of predicting the benign and malignant risk of one of the subject's liver tumors. The program includes: a liver tumor labeling unit, which is used to automatically refer to the several reference based on specific rules established based on empirical rules The ultrasound image of the liver tumor is marked with the ultrasound image of the liver tumor with the tumor image point area and the type of the liver tumor is identified; a classification unit is used to use the ultrasound image of the liver tumor with the tumor image point area. A machine learning algorithm training to establish an inference model; and a comparison unit to analyze the target liver tumor ultrasound image with the inference model to provide clinical staff to determine the liver tumor type of the subject And predict the probability of benign and malignant liver tumor risk in one of the subjects. According to the liver tumor intelligent analysis device described in item 1 of the patent application scope, The image acquisition unit is a digital visual interface (Digital Visual Interface, DVI). According to the liver tumor intelligent analysis device described in item 1 of the patent application scope, wherein the control The control unit is a central processing unit (Central Processing Unit, CPU). The liver tumor intelligent analysis device described in item 1 of the patent application scope, wherein the liver The tumor marking unit is based on the coefficients and/or parameters based on the matching experience data, and automatically marks the liver tumor ultrasound image with the tumor image point area from the several reference liver tumor ultrasound images. The liver tumor intelligent analysis device described in item 1 of the patent application scope, wherein the liver Types of dirty tumors include benign and malignant tumors. According to the liver tumor intelligent analysis device described in item 1 of the scope of patent application, where The analysis module further includes a generating unit for predicting the benign and malignant risk of liver tumors, which is connected to the control unit, and the control unit can provide the clinical staff to determine the liver tumor type of the subject and predict the The probability of liver cancer benign or unhealthy risk of one of the subjects is input into the predicted liver tumor benign and malignant risk report generation unit to produce a liver tumor nature diagnosis report.

Images