A cloud area radiation information management system

Abstract

The present invention discloses a cloud area radiological information management system, which belongs to the field of data processing technology and includes: a registration module for receiving a medical image acquisition request from a patient and registering the patient's basic information, the basic information including priority information; an intelligent queuing module for directly arranging the patient to the end of the queue when the patient is an ordinary patient; when the patient is a priority patient, recalculating the priority value of each patient in the queue according to the patient category, the waiting time, and the current queue position; updating the queue according to the priority value of each patient; a report generation module for automatically generating a medical image report according to the patient's medical image; an encryption module for encrypting the medical image report; and an identity authentication module for downloading the medical image report when the user's identity authentication is passed. The average waiting time of patients can be reduced and the efficiency of the hospital's queuing for medical treatment can be improved.

Description

Technical Field

[0001] The invention belongs to the technical field of data processing, and in particular relates to a cloud area radiation information management system. Background Art

[0002] With the development of modern medicine, the diagnosis and treatment work in hospitals is increasingly dependent on medical imaging examinations (such as CT, DR, CR, MR, digital breast, interventional angiography, ultrasound, endoscopy, pathology, etc.), while traditional imaging materials (such as film, paper reports) have problems such as poor clarity of imaging films, difficulty in carrying, and difficulty in storage. With the development of computer communication technology, digital imaging has become an important breakthrough in solving the problem of sending and receiving images. The Radiology Information System (RIS) has been further developed, and the realization of filmless radiology departments and digital hospitals has become an unstoppable trend in modern medicine.

[0003] However, when arranging patients for medical image collection, existing radiological information management systems often only collect images in the order registered in the system. Some existing technologies consider giving priority to collecting images of special groups such as follow-up patients, the elderly, soldiers, and patients with appointments as a new queue. This will lead to frequent queue jumping by special patients, making it difficult for ordinary patients to have their medical images collected in a timely manner. The average waiting time for ordinary patients is too long, which generates negative emotions of unfairness, thereby affecting the efficiency of the hospital's medical queuing. Summary of the invention

[0004] In order to solve the technical problems of the average waiting time of ordinary patients being too long and the efficiency of hospital consultation queuing in the prior art, the present invention provides a cloud area radiation information management system.

[0005] The present invention provides a cloud region radiology information management system (Radiology Information System, RIS), comprising:

[0006] A registration module, used to receive a medical image acquisition request from a patient and register the patient's basic information, including priority information;

[0007] Intelligent queuing module, used to add patients to the queue according to their priority information and collect medical images of patients according to the queue order;

[0008] Among them, adding patients to the queue according to the patient's priority information specifically includes:

[0009] If the patient is a common patient, the patient will be directly placed at the end of the queue;

[0010] When the patient is a priority patient, the priority value of each patient in the queue is recalculated according to the patient category, the waiting time, and the current queue position;

[0011] Update the queue according to the priority value of each patient;

[0012] A report generation module is used to automatically generate medical imaging reports based on the patient's medical imaging;

[0013] An encryption module, used for encrypting medical imaging reports;

[0014] The identity authentication module is used to download the medical imaging report when the user's identity authentication is passed.

[0015] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0016] In the present invention, every time a follow-up patient, the elderly, soldiers and other special population patients, as well as appointment patients and other priority patients enter the queue, the priority value of each patient will be recalculated according to the patient category, the waiting time and the current queue position, and the queue will be updated according to the priority value of each patient to minimize the number of ordinary patients who have been waiting for too long and are currently at the front of the queue from being cut in line. At the same time, the special circumstances of priority patients are taken into account to reduce the average waiting time of patients and improve the hospital's queuing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The preferred implementation modes will be described below in a clear and understandable manner with reference to the accompanying drawings to further illustrate the above-mentioned characteristics, technical features, advantages and implementation methods of the present invention.

[0018] Figure 1 It is a structural schematic diagram of a cloud area radiation information management system provided by the present invention.

[0019] Figure 2 It is a flow chart of a method for calculating a patient's priority value provided by the present invention. DETAILED DESCRIPTION

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.

[0021] In order to simplify the drawings, only the parts related to the invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".

[0022] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0023] In this document, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0024] In addition, in the description of the present invention, the terms "first", "second", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0025] In one embodiment, referring to Figure 1 , showing a schematic diagram of the structure of the cloud area radiation information management system provided by the present invention.

[0026] The present invention provides a cloud area radiology information management system (Radiology Information System, RIS), comprising: a registration module 1, an intelligent queuing module 2, a report generation module 3, an encryption module 4 and an identity authentication module 5.

[0027] Among them, the cloud regional radiology information management system refers to a regional information system based on cloud computing technology, which usually has the following functions: (1) Image data management: including image acquisition, storage, retrieval and archiving. (2) Patient information management: recording the patient's identity, medical history, examination records and diagnosis results. (3) Scheduling and appointment management: arranging the patient's radiology examination time and equipment resource allocation. (4) Workflow management: coordinating the workflow within the imaging department, including image review, report writing and result distribution. (5) Data analysis and reporting: providing statistical analysis, report generation and data visualization functions for quality control and business decision-making.

[0028] The registration module 1 is used to receive a medical image acquisition request from a patient and register the patient's basic information, including priority information.

[0029] Among them, the registration module 1 is mainly used to record the patient's relevant information for subsequent processing and management.

[0030] The intelligent queuing module 2 is used to add patients to the queue according to their priority information and collect medical images of the patients according to the queue order.

[0031] Among them, adding patients to the queue according to the patient's priority information specifically includes:

[0032] If the patient is a common patient, the patient will be placed directly at the end of the queue.

[0033] When the patient is a priority patient, the priority value of each patient in the queue is recalculated according to the patient category, waiting time, and current queue position.

[0034] Among them, priority patients include: patients who come back for a same-day visit, patients from special groups such as the elderly and military personnel, patients who have made appointments, and special patients identified by physicians. Patients who come back for a same-day visit usually require further medical imaging examinations or treatments in a shorter period of time because they have already received previous diagnosis and treatment. In order to maintain their medical plans and continuity, making them priority patients can ensure that they receive timely medical imaging services. As special groups, the elderly, military personnel, etc. deserve preferential treatment. Patients who have made appointments have already arranged specific medical imaging examination times in advance. In order to avoid their waiting time being too long and to ensure the quality of service, making them priority patients can give priority to their medical imaging examinations. Physicians may identify certain patients as special patients who need priority medical imaging examinations based on their specific conditions and clinical needs.

[0035] Among them, the priority value can be understood as the urgency of the patient's need for medical imaging examination in the queue. The higher the priority value, the front of the patient in the queue.

[0036] Reference Figure 2 , showing a flow chart of a method for calculating a patient's priority value provided by the present invention.

[0037] In a possible implementation, the intelligent queuing module 2 is specifically configured to perform the following steps 201 to 205:

[0038] Step 201: Construct a decision matrix A based on patient category, waiting time, and current queue position:

[0039]

[0040] Among them, ai1 represents the waiting time of the i-th patient, a i2 represents the current queue position of the i-th patient, a i3 represents the category of the i-th patient.

[0041] Furthermore, the indicator value of the patient category is determined as follows:

[0042] When the patient is a normal patient, the indicator value of the patient category is 1.

[0043] When the patient is a follow-up patient on the same day or a patient from a special population such as the elderly or military personnel, the indicator value of the patient category is 2.

[0044] When the patient is a scheduled patient, the indicator value for the patient category is 3.

[0045] When the patient is a special patient determined by the physician, the indicator value for the patient category is 4.

[0046] In other words, different types of patients receive different degrees of preferential treatment in the queue. Physicians may identify certain patients as special patients based on their specific conditions and clinical needs, and they need to be given the highest priority for medical imaging examinations. Patients who have made appointments have already arranged a specific medical imaging examination time in advance. If they cannot be guaranteed to receive a medical examination within the appointment time range, the appointment will lose its meaning, which will further lead to a further decline in queuing efficiency. The elderly, military personnel, and other special groups should be given preferential treatment.

[0047] It should be noted that the longer the patient has been waiting and the closer they are to the front of the queue, the less willing they are to be cut in line and the more likely they are to have negative emotions. We should consider increasing the priority value of patients who have been waiting too long and are at the front of the queue, reducing the number of patients who have been waiting too long and are at the front of the queue being cut in line, reducing the average waiting time of patients, and improving the hospital's queuing efficiency.

[0048] In the present invention, by comprehensively considering factors such as patient category, remaining waiting time and current queue position, the priority of patients can be determined more fairly, and the queuing and scheduling process of medical image acquisition can be optimized. By rationally allocating resources and optimizing processes, patient satisfaction can be improved, and the efficiency and service quality of medical institutions can be improved.

[0049] Step 202: normalize the index values ​​of each decision index in the decision matrix A to obtain a normalized decision matrix B:

[0050]

[0051]

[0052] Among them, bij represents the normalized index value of the jth decision index of the i-th patient, j = 1 or 2 or 3, a ij represents the index value of the jth decision index of the i-th patient before normalization, a maxj represents the maximum value of the j-th decision indicator, a minj Represents the minimum value of the j-th decision indicator.

[0053] It should be noted that the purpose of normalizing the index values ​​of each decision index in the decision matrix A is to eliminate the dimensions and differences between the indicators so as to enable comparison and trade-off.

[0054] Step 203: Take the maximum value of each column index of the normalized decision matrix B to form a column vector as the positive ideal solution B + , take the minimum value of each column index of the normalized decision matrix B to form a column vector as the negative ideal solution B - .

[0055] Among them, the positive ideal solution B + represents the ideal situation when the maximum value is achieved on each decision indicator, while the negative ideal solution B - It represents the ideal situation when the minimum value is achieved on each decision indicator.

[0056] Step 204: Calculate the index value of each patient to the positive ideal solution B + and negative ideal solution B - Distance between:

[0057]

[0058]

[0059] Among them, the index value of each patient is the positive ideal solution B + and negative ideal solution B - The distance between them reflects the degree of difference between the patient's index value and the ideal situation in each decision-making index.

[0060] Step 205: Get the positive ideal solution B according to the index value of each patient + and negative ideal solution B - The distance between them is used to calculate the priority value P of each patient. i :

[0061]

[0062] The priority value of the patient is determined according to the degree of difference between the patient's index value and the ideal situation. The smaller the difference with the ideal situation, the higher the priority value.

[0063] Update the queue based on the priority value of each patient.

[0064] Among them, according to the priority value P i The patients are sorted by the size of the queue and the order of the queue is updated. Patients with higher priority values ​​can be placed at the front of the queue so that medical imaging can be collected earlier.

[0065] In the present invention, multiple decision-making indicators are comprehensively considered to achieve personalized priority sorting, reduce the average waiting time of patients, improve the efficiency of hospital consultation queuing, and thus optimize the process of medical image acquisition.

[0066] The report generation module 3 is used to automatically generate a medical imaging report based on the patient's medical imaging.

[0067] Specifically, first, relevant features are extracted from the patient's medical images, and the extracted features are converted into a representation that can be used to automatically generate reports. After that, natural language processing (NLP) technology is used, combined with specific models and algorithms, to convert the feature representation into natural language text. The generated text report is structured to meet the specifications and standards of medical reports. Finally, the final generated medical imaging report is output in a visual form for doctors and other medical professionals to view and interpret. This can include displaying the report in a medical imaging system or exporting it as a document in PDF or other formats.

[0068] In a possible implementation, the report generation module 3 is specifically configured to perform the following steps 301 to 308:

[0069] Step 301: Obtain a sample data set, where the sample data set includes a plurality of medical training images with known conclusions.

[0070] Step 302: Input the medical training image into an encoder module, which includes an image performance encoder, an image conclusion encoder, and a clinical diagnosis encoder.

[0071] Among them, the encoder module is used to convert medical training images into representation forms for subsequent processing and analysis.

[0072] In a possible implementation manner, the report generating module 3 is specifically configured to:

[0073] An image representation encoder is constructed through a convolutional neural network, and the image representation encoder is used to convert the visual representation information of medical training images into a vector representation of a fixed length as an embedded feature.

[0074] An image conclusion encoder is constructed through text embedding technology. The image conclusion encoder encodes the conclusion information of medical training images into a vector representation of a fixed length as an embedding feature.

[0075] A clinical diagnosis encoder is constructed through a recurrent neural network, and the clinical diagnosis information is converted into a fixed-length vector representation through the clinical diagnosis encoder as an embedding feature.

[0076] Step 303: Input the embedded features output by the encoder module into the recurrent neural network.

[0077] Among them, the recurrent neural network (RNN) is a type of neural network model widely used in sequence data processing. Unlike traditional feedforward neural networks, RNN has a self-looping structure that allows information to be transmitted and shared in the network. The recurrent neural network generates a new hidden state and output based on the hidden state of the previous time step and the input of the current time step.

[0078] Among them, the embedded features are high-level representations of the images after encoding and transformation, usually a vector or matrix. These embedded features contain important information of medical images.

[0079] Step 304: The hidden state output by the recurrent neural network and the embedded features output by the encoder module are input into the decoder module.

[0080] Step 305: Output the prediction result of the description word of the medical training image through the decoder module.

[0081] In the entire encoding and decoding process, the encoder module is responsible for converting images into embedded features, the recurrent neural network is responsible for modeling on the time series and generating hidden states, and the decoder module converts the hidden states and embedded features into the final output results.

[0082] Step 306: construct an objective function of the decoder module, and adjust the parameter θ of the decoder module through the objective function. The objective function f(θ) can be expressed as:

[0083]

[0084] Wherein, w represents the word obtained by Monte Carlo sampling, x represents the prediction result, and w is subject to the text probability p obtained according to the prediction result x. θ , r i (w) represents the current reward obtained under the i-th evaluation indicator, represents the reward for reasoning under the i-th evaluation metric, λ iRepresents the weight of the i-th evaluation indicator. The seven evaluation indicators include: BLEU-1, BLEU-2, BLEU-3, BLEU-4, METEOR, ROUGE-L and CIDEr.

[0085] Among them, BLEU-1, BLEU-2, BLEU-3 and BLEU-4 are different variants of BLEU. BLEU-1 is used to measure the match between the automatic translation result and the reference translation at the level of a single word (unigram). BLEU-2 is used to measure the match between the automatic translation result and the reference translation at the level of two consecutive words (bigram). BLEU-3 is used to measure the match between the automatic translation result and the reference translation at the level of three consecutive words (trigram). BLEU-4 is used to measure the match between the automatic translation result and the reference translation at the level of four consecutive words (4-gram).

[0086] Among them, METEOR combines comparisons at multiple levels such as exact matching, vocabulary, phrases and syntactic structure to measure the degree of match between the reference answer and the candidate translation.

[0087] Among them, ROUGE-L measures the similarity between candidate summaries and reference summaries by calculating the longest common subsequence between them.

[0088] Among them, CIDEr measures the quality of candidate descriptions by calculating the similarity between candidate descriptions and reference descriptions and combining diversity and relevance metrics.

[0089] Among them, the objective function is used to measure the performance of the decoder module in generating text.

[0090] Among them, the prediction result probability p θ Represents the probability of the text generated by the decoder module under given conditions. This probability can be estimated by Monte Carlo sampling.

[0091] It should be noted that by maximizing the objective function, the parameters θ of the decoder module can be adjusted to optimize the quality of the generated text. The parameter adjustment process is usually performed using optimization methods such as gradient descent. By continuously iterating and adjusting parameters, the decoder module can generate text that is more in line with expectations and obtain higher rewards under various evaluation indicators.

[0092] Furthermore, the weight λ of each evaluation index i The method of determining is:

[0093] The same weight value is set for each evaluation indicator, and the weight value of each evaluation indicator is increased by 1 one by one. By comparing the evaluation results after adding the weight value to each evaluation indicator, the most influential evaluation indicator is determined.

[0094] Starting from the most influential evaluation indicator, the weight value of the most influential evaluation indicator is gradually increased with c as the preset increment. By comparing the evaluation results after continuously increasing the weight value, the optimal weight of the most influential evaluation indicator is determined, and c is a positive integer.

[0095] The optimal weight of the most influential evaluation indicator is fixed, and the above steps are repeated to determine the remaining most influential indicators one by one and determine their optimal weights.

[0096] The optimal weights of each evaluation index are taken as the optimal weight combination, and the optimal weight combination is normalized to obtain the final weight λ of each evaluation index. i .

[0097] In the present invention, the importance of each evaluation index is comprehensively considered and their optimal weights are determined. Finally, the optimal weights of each evaluation index are normalized to obtain the final weight λ i , which is used to comprehensively evaluate the performance of the decoder module. This method can determine the weights of the evaluation indicators according to their actual influence, making the evaluation results more accurate and reliable.

[0098] Furthermore, the weight determination method provided by the present invention effectively reduces the computational complexity of weight search and can quickly and accurately determine the weight of each evaluation indicator.

[0099] Step 307: Obtain the patient's medical image, and obtain the image performance results, image conclusion results and clinical diagnosis results through the encoder module, recurrent neural network and decoder module.

[0100] Step 308: Automatically generate a medical imaging report based on the imaging performance results, imaging conclusion results and clinical diagnosis results.

[0101] The encryption module 4 is used to encrypt the medical imaging report.

[0102] It should be noted that encrypting medical imaging reports can provide benefits in terms of data privacy protection, data security, compliance, and transmission security, helping to ensure the confidentiality and security of medical data.

[0103] In a possible implementation, the encryption module 4 is specifically configured to perform the following steps 401 to 407:

[0104] Step 401: construct an elliptic curve, the parameters of the elliptic curve include: p, a, b, G, n, h, where p represents a prime number, a represents a first elliptic coefficient, b represents a second elliptic coefficient, G represents a base point, the base point G can be expressed as (X0, Y0), n represents the order of G, and h represents a cofactor.

[0105] Among them, in cryptography, elliptic curves provide a secure encryption and digital signature scheme that is more secure than traditional encryption algorithms and requires a shorter key length. This is because the discrete logarithm problem of elliptic curves over finite fields is more difficult to solve than traditional encryption algorithms, so attackers need longer time and higher computing power to crack the cryptographic system.

[0106] Step 402: Select a random number k, calculate kG=(kX0, kY0)=(X1, Y1), and convert the coordinate X1 into an integer.

[0107] Step 403: Calculate r=X1modn. If r=0, return to step 402; otherwise, proceed to step 404.

[0108] Step 404: Calculate l=SHA-1(M), and convert l into an integer, where M represents the original data, and SHA-1(M) represents the digital summary of M calculated using SHA-1.

[0109] Among them, SHA-1 is a secure hash algorithm that can convert a message of any length into a 160-bit digital fingerprint. The SHA-1 algorithm is widely used in fields such as digital signatures, authentication, and data integrity checks, and has high security and wide applicability. A digital summary is a message of any length (such as text, files, etc.) converted into a fixed-length digital string through a specific algorithm. The digital summary can be regarded as the "fingerprint" of the original data, which has the characteristics of uniqueness, irreversibility, and sensitivity. By comparing whether two digital summaries are the same, it can be determined whether the original data has been tampered with.

[0110] Step 405: Construct a signature equation, k=l+s+rdmodn, and use the private key d to calculate s=kl-rd(mod n). If s=0, return to step 402, otherwise proceed to step 406.

[0111] Step 406: Use (r, s) as the digital signature of the medical imaging report to encrypt the medical imaging report.

[0112] It should be noted that the digital signature equation involved in the present invention only has one multiplication operation, which reduces the multiplication operation compared to the digital signature algorithm in the prior art. While ensuring that the forgery of the digital signature can be detected and prevented, the signature equation is simpler, the calculation amount is smaller, and the communication cost is lower. In actual use, the speed of the digital signature is improved and less computing resources are occupied.

[0113] It should be noted that in the process of generating a digital signature, the present invention has a random number k, so that a third party cannot steal the private key d, because the random number k and the private key d are unknown. In an equation, there are two unknown numbers that cannot be solved, thereby ensuring that an attacker cannot forge a digital signature by stealing the sender's private key. In other words, the attacker forges a digital signature by forging a fake private key, and the forged digital signature needs to be verified by the receiver. There is only one public key corresponding to the private key, and the digital signature generated by the fake private key cannot be correct, thereby achieving the authentication of the user's identity. If the attacker tampers with the message sent by the sender, different digital summaries are obtained, and then different r values ​​are obtained, and the real r value is unique in the digital signature. The r value generated by the digital summary generated by the tampered message is different from the r value generated by the real digital summary. Therefore, when the digital signature is verified at the message receiving end, the fake r value is different from the real r value in the digital signature, and the message cannot pass, thereby achieving the verification of the authenticity and integrity of the message.

[0114] In a possible implementation manner, the encryption module 4 is further configured to perform the following steps:

[0115] Step 407: Construct a chaotic mapping relationship between the random number k and the private key d:

[0116]

[0117] Where n represents the number of downloads, λ1, λ2, λ3, λ4 and λ5 represent control parameters and are all constants.

[0118] It should be noted that the role of the chaotic mapping relation is to protect random numbers and private keys. The chaotic mapping relation in the prior art has a discontinuous chaotic parameter range in the existing two-dimensional chaotic mapping, and there are many periodic windows in the parameter space. The chaotic behavior is relatively fragile. When the parameters are disturbed, the chaotic behavior will easily disappear and the problem of chaotic degradation will occur. The present invention first initializes two parameter polynomials, then folds any value into a fixed range through modular operation, and finally generates a chaotic mapping from a nonlinear polynomial to generate a two-dimensional chaotic mapping with robust chaos, which makes up for the shortcomings of the existing chaotic mapping relation.

[0119] The identity authentication module 5 is used to download the medical imaging report when the user's identity authentication is passed.

[0120] In a possible implementation, the identity authentication module 5 is specifically configured to perform the following steps 501 to 505:

[0121] Step 501: Calculate the public key Q based on the user's private key d, where Q=dG, d represents the private key, and G represents the base point.

[0122] Step 502: Calculate l=SHA-1(M), where SHA-1(M) represents the digital summary of M calculated using SHA-1.

[0123] Step 503: Calculate P=(s+l)G+rQ=(X1,Y1). If X1=0 and Y1=0, the signature is invalid.

[0124] Step 504: Convert the coordinate X1 to an integer, and set r'=X1 mod n.

[0125] Step 505: If r=r', it is determined that the digital signature verification is successful, and the user is allowed to download the medical imaging report.

[0126] In a possible implementation, the cloud region radiology information management system further includes:

[0127] The retrieval module 6 is used to receive a retrieval request from a physician and retrieve a medical imaging report.

[0128] Specifically, when physicians need to view a patient's medical imaging report, they can submit a request through the retrieval module. After receiving the request, the retrieval module will retrieve the medical imaging report stored in the system based on the patient's identity information or other identifiers, and return the relevant report to the physician. Through the retrieval module, physicians can easily obtain the medical imaging reports they need for diagnosis, treatment or other medical decisions. The existence of this module can simplify the physician's workflow and reduce their time and effort in obtaining medical imaging reports. At the same time, the retrieval module can ensure the security and controllability of medical imaging reports, and only authorized physicians can obtain the corresponding reports.

[0129] Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0130] In the present invention, every time a follow-up patient, the elderly, soldiers and other special population patients, as well as appointment patients and other priority patients enter the queue, the priority value of each patient will be recalculated according to the patient category, the waiting time and the current queue position, and the queue will be updated according to the priority value of each patient to minimize the number of ordinary patients who have been waiting for too long and are currently at the front of the queue from being cut in line. At the same time, the special circumstances of priority patients are taken into account to reduce the average waiting time of patients and improve the hospital's queuing efficiency.

[0131] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0132] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for those of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A cloud area radiation information management system, characterized in that: include: A registration module, used to receive a medical image acquisition request from a patient and register basic information of the patient, wherein the basic information includes priority information; Intelligent queuing module, used to add patients to the queue according to their priority information and collect medical images of patients according to the queue order; The step of adding the patient to the queue according to the patient's priority information specifically includes: If the patient is a common patient, the patient will be directly placed at the end of the queue; When the patient is a priority patient, the priority value of each patient in the queue is recalculated according to the patient category, the waiting time, and the current queue position; Update the queue according to the priority value of each patient; A report generation module is used to automatically generate medical imaging reports based on the patient's medical imaging; An encryption module, used for encrypting the medical imaging report; An identity authentication module, used to download the medical imaging report if the user's identity authentication is passed; The report generation module is specifically used to perform the following steps: Step 301: Acquire a sample data set, wherein the sample data set includes a plurality of medical training images with known conclusions; Step 302: inputting the medical training image into an encoder module, wherein the encoder module includes an image performance encoder, an image conclusion encoder and a clinical diagnosis encoder; Step 303: input the embedding features output by the encoder module into a recurrent neural network; Step 304: inputting the hidden state output by the recurrent neural network and the embedded features output by the encoder module into a decoder module; Step 305: outputting the prediction result of the description word of the medical training image through the decoder module; Step 306: construct an objective function of the decoder module, and adjust the parameter θ of the decoder module by the objective function. The objective function f(θ) can be expressed as: Wherein, w represents the word obtained by Monte Carlo sampling, x represents the prediction result, and w is subject to the text probability p obtained according to the prediction result x. θ , r i (w) represents the current reward obtained under the i-th evaluation indicator, represents the reward for reasoning under the i-th evaluation metric, λ i represents the weight of the i-th evaluation index; the seven evaluation indicators include: BLEU-1, BLEU-2, BLEU-3, BLEU-4, METEOR, ROUGE-L and CIDEr; Step 307: Obtain the patient's medical image, and obtain the image performance result, image conclusion result and clinical diagnosis result through the encoder module, the recurrent neural network and the decoder module; Step 308: Automatically generate a medical imaging report according to the imaging performance results, the imaging conclusion results and the clinical diagnosis results.

2. The cloud area radiation information management system according to claim 1, characterized in that: The intelligent queuing module is specifically used to perform the following steps: Step 201: Construct a decision matrix A based on patient category, waiting time, and current queue position: Among them, a i1 represents the waiting time of the i-th patient, a i2 represents the current queue position of the i-th patient, a i3 represents the category of the i-th patient; Step 202: normalize the index values ​​of each decision index in the decision matrix A to obtain a normalized decision matrix B: Among them, b ij represents the normalized index value of the jth decision index of the i-th patient, j = 1 or 2 or 3, a ij represents the index value of the jth decision index of the i-th patient before normalization, a maxj represents the maximum value of the jth decision indicator, a minj represents the minimum value of the j-th decision indicator; Step 203: Take the maximum value of each column index of the normalized decision matrix B to form a column vector as the positive ideal solution B + , take the minimum value of each column index of the normalized decision matrix B to form a column vector as the negative ideal solution B - ; Step 204: Calculate the index value of each patient to the positive ideal solution B + and the negative ideal solution B - Distance between: Step 205: According to the index value of each patient, the positive ideal solution B is obtained. + and the negative ideal solution B - The distance between them is used to calculate the priority value P of each patient. i :

3. The cloud area radiation information management system according to claim 1, characterized in that: The indicator value of the patient category is determined as follows: When the patient is a common patient, the index value of the patient category is 1; When the patient is a follow-up patient on the same day or a patient from a special population such as the elderly or military personnel, the index value of the patient category is 2; When the patient is a scheduled patient, the index value of the patient category is 3; When the patient is a special patient determined by a physician, the index value of the patient category is 4.

4. The cloud area radiation information management system according to claim 1, characterized in that: The weight λ of each evaluation indicator i The method of determining is: Set the same weight value for each evaluation indicator, and add 1 to the weight value of each evaluation indicator one by one. By comparing the evaluation results after adding weight values ​​to each evaluation indicator, determine the most influential evaluation indicator; Starting from the most influential evaluation indicator, the weight value of the most influential evaluation indicator is gradually increased with n as the preset increment, and the optimal weight of the most influential evaluation indicator is determined by comparing the evaluation results after the weight values ​​are continuously increased; Fix the optimal weight of the most influential evaluation indicator and repeat the above steps to determine the remaining most influential indicators one by one and determine their optimal weights; The optimal weights of each evaluation index are used as the optimal weight combination, and the optimal weight combination is normalized to obtain the final weight λ of each evaluation index. i .

5. The cloud area radiation information management system according to claim 1, characterized in that: The report generation module is specifically used for: The image representation encoder is constructed by a convolutional neural network, and the visual representation information of the medical training image is converted into a vector representation of a fixed length as an embedding feature by the image representation encoder; The image conclusion encoder is constructed by using text embedding technology, and the conclusion information of the medical training image is encoded and converted into a vector representation of a fixed length by using the image conclusion encoder to serve as an embedding feature; The clinical diagnosis encoder is constructed through a recurrent neural network, and the clinical diagnosis information is converted into a vector representation of a fixed length through the clinical diagnosis encoder to serve as an embedded feature.

6. The cloud area radiation information management system according to claim 1, characterized in that: The encryption module is specifically used to perform the following steps: Step 401: construct an elliptic curve, the parameters of which include: p, a, b, G, n, h, where p represents a prime number, a represents a first elliptic coefficient, b represents a second elliptic coefficient, G represents a base point, the base point G can be represented as (X0, Y0), n represents the order of G, and h represents a cofactor; Step 402: Select a random number k, calculate kG = (kX0, kY0) = (X1, Y1), and convert the coordinate X1 into an integer; Step 403: Calculate r = X1 mod n, if r = 0, return to step 402, otherwise go to step 404; Step 404: Calculate l=SHA-1(M), and convert l into an integer, where M represents the original data, and SHA-1(M) represents the digital summary of M calculated using SHA-1; Step 405: Construct a signature equation, k=l+s+rdmodn, and use the private key d to calculate s=kl-rd(mod n). If s=0, return to step 402, otherwise proceed to step 406; Step 406: Use (r, s) as the digital signature of the medical imaging report to encrypt the medical imaging report.

7. The cloud area radiation information management system according to claim 6, characterized in that: The encryption module is further used to perform the following steps: Step 407: Construct a chaotic mapping relationship between the random number k and the private key d: Where n represents the number of downloads, λ1, λ2, λ3, λ4 and λ5 represent control parameters and are all constants.

8. The cloud area radiation information management system according to claim 6, characterized in that: The identity verification module is specifically used to perform the following steps: Step 501: Calculate the public key Q according to the user's private key d, where Q = dG, d represents the private key, and G represents the base point; Step 502: Calculate l=SHA-1(M), where SHA-1(M) represents the digital summary of M calculated using SHA-1; Step 503: Calculate P = (s + l) G + rQ = (X1, Y1). If X1 = 0 and Y1 = 0, the signature is invalid. Step 504: Convert the coordinate X1 to an integer, and set r'=X1 modn; Step 505: If r=r', it is determined that the digital signature verification is passed, and the user is allowed to download the medical imaging report.

9. The cloud area radiation information management system according to claim 1, characterized in that: Also includes: The retrieval module is used to receive a retrieval request from a physician and retrieve the medical imaging report.

Images