CN-121354838-B - Cardiovascular intervention catheter warehouse scheduling system based on demand analysis

CN121354838BCN 121354838 BCN121354838 BCN 121354838BCN-121354838-B

Abstract

The invention discloses a cardiovascular interventional catheter warehouse scheduling system based on demand analysis, which relates to the technical field of medical informatization and comprises a data acquisition module, a characteristic extraction module, a standard deduction module, a selection probability calculation module, a first judgment module and a second judgment module, wherein the data acquisition module is used for acquiring a surgery plan, a doctor identification, a step sequence and angiography images, the characteristic extraction module is used for generating a situation characteristic vector according to the angiography images, the standard deduction module is used for deducting a subsequent step and a corresponding catheter set, the selection probability calculation module is used for inquiring a doctor preference knowledge base by combining the doctor identification, a step code and the characteristic vector to obtain catheter selection probability distribution, the first judgment module is used for outputting a predicted catheter model if the catheter selection probability exceeds a first threshold value, otherwise, the use frequency of a catheter of historical data of a doctor is counted, the second judgment module is used for outputting a corresponding catheter model if the use frequency exceeds a second threshold value, otherwise, the predicted catheter model is determined according to rules from the standard catheter set, and the accuracy and clinical suitability of catheter prediction can be improved.

Inventors

TANG HAOXIN
ZHUANG JIAN
SHI CUICUI
LIU CHANGXING
ZHAO LINA

Assignees

中国人民武装警察部队特色医学中心

Dates

Publication Date: 20260512
Application Date: 20251023

Claims (9)

1. A demand-based analysis cardiovascular interventional catheter depot scheduling system, comprising: the data acquisition module is used for acquiring a surgical plan, a doctor mark of a main knife, a completed step sequence and a current angiography image; A feature extraction module for extracting a context feature vector from the current angiographic image, the context feature vector describing an anatomical state of a blood vessel; the standard deduction module is used for deducting through the operation plan knowledge graph according to the completed step sequence to obtain a possible subsequent standard step, a possible subsequent step code and a corresponding standard catheter set; The selection probability calculation module is used for inquiring a doctor preference knowledge base according to the doctor identification of the main cutter, the follow-up step codes and the situation feature vector to obtain catheter selection probability distribution; The first judging module is used for judging whether the selection probability of a certain catheter model in the catheter selection probability distribution exceeds a first preset threshold value, if yes, determining the catheter model as a predicted catheter model, otherwise, inquiring all historical operation data corresponding to the doctor identification of the main cutter and the current step code in the doctor preference knowledge base, and counting the use frequency of various catheter models; The second judging module is used for judging whether the using frequency of a certain conduit model exceeds a second preset threshold value, if yes, determining the conduit model as a predicted conduit model, otherwise, selecting a standard conduit model from the standard conduit model set as the predicted conduit model according to a preset rule; the replacement cost evaluation module is used for executing the following processing when the second judgment module judges that the catheter needs to be selected from the standard catheter set: Acquiring a history substitution record corresponding to the doctor mark of the main knife under the current operation situation; Extracting difference information of a catheter actually used in each replacement operation and a standard catheter; calculating one or more of average additional operation time, auxiliary instrument use increment and contrast agent additional consumption caused by substitution based on the difference information to form a substitution cost vector; Weighting and fusing the substitution cost vector and a real-time pressure index of the current operation to generate contextualized substitution cost scores; and taking the contextualized substitution cost score as a negative weighting factor in a preset rule to participate in the weighted scoring process of each model in the standard catheter set.
2. The on-demand analysis cardiovascular interventional catheter library scheduling system of claim 1, wherein the process of extracting contextual feature vectors from the current angiographic image comprises: acquiring a plurality of continuous angiography image frames in a time sequence, carrying out image feature extraction processing on each image frame to obtain a group of situation feature vector sequences, and calculating variance values of specific features in the situation feature vector sequences; Judging whether the variance value is lower than a stability threshold value, if so, using the situation characteristic vector of the current frame or the last frame of the sequence as an effective situation characteristic vector for subsequent inquiry, otherwise, re-acquiring a new image frame and re-calculating until the effective situation characteristic vector is obtained; the dimensions of the contextual feature vector include tortuosity, degree of calcification, stenosis length, and thrombus loading of the target vessel.
3. The cardiovascular interventional catheter depot scheduling system of claim 1, wherein the physician preference repository is configured by machine learning algorithms to analyze historical surgical data including surgical procedures, catheter model used, corresponding angiographic images, and physician identification.
4. The on-demand analysis cardiovascular interventional catheter depot scheduling system of claim 1, wherein in determining whether a selection probability of a catheter model in the catheter selection probability distribution exceeds a first predetermined threshold: If the highest value of the selection probability is in the neighborhood range taking the first preset threshold value as the center, judging the situation as a fuzzy situation; under the fuzzy situation, a confirmation request containing N candidate catheter models before the sorting from big to small in selection probability is sent to an operating room terminal, wherein N is a positive integer; Selection feedback from the operating room terminal is received and the catheter model selected by the feedback is determined to be the final predicted catheter model.
5. The cardiovascular interventional catheter depot scheduling system based on demand analysis of claim 1, wherein the preset rules comprise: weighting and scoring according to at least one factor of inventory turnover rate, effective period and purchasing cost of the conduit model; And selecting the standard catheter model with the highest score as the predicted catheter model.
6. The cardiovascular interventional catheter depot scheduling system of claim 1, wherein the replacement cost assessment module extracts difference information comprising: Calculating the difference value of the time used by the alternative catheter and the standard catheter in the same operation step through the operation time stamp sequence recorded in the doctor preference knowledge base; if the replacement catheter is used with other instruments, recording the use quantity and types of the additional instruments; Extracting contrast agent using amount difference through the operation image and report; all the difference information is stored in a substitution cost database in a numerical form and organized according to doctor-step-lesion type three-level index.
7. The cardiovascular interventional catheter depot scheduling system of claim 1, wherein the contextualized surrogate cost score is calculated by: Wherein, the For the time-cost normalized value, For the instrument consumption increment, In order to increase the amount of contrast agent, Is a weight coefficient.
8. The demand-based cardiovascular interventional catheter depot scheduling system of claim 1, further comprising: the situation deviation calculating module is used for executing the following operations after the probability distribution is output by the selection probability calculating module: Extracting cosine similarity of the current situation feature vector and a similar operation situation vector in a doctor preference knowledge base; If the cosine similarity is lower than a preset similarity threshold, judging that the current situation deviates from the typical preference situation of the doctor, automatically reducing the weight of the preference probability of the doctor, and improving the priority of the general model in the standard catheter set.
9. A method for demand-based analysis of cardiovascular interventional catheter depot scheduling comprising the steps of: Acquiring a surgical plan, a doctor identification of a main cutter, a completed step sequence and a current angiographic image; Extracting a situation feature vector according to the current angiography image, wherein the situation feature vector is used for describing the anatomical state of a blood vessel; Deducing through a surgical plan knowledge graph according to the completed step sequence to obtain a possible subsequent standard step, a possible subsequent step code and a corresponding standard catheter set; inquiring a doctor preference knowledge base according to the doctor identification of the main cutter, the follow-up step codes and the situation feature vector to obtain catheter selection probability distribution; If yes, determining the catheter model as a predicted catheter model, otherwise, inquiring all historical operation data corresponding to the doctor identification of the main cutter and the current step code in the doctor preference knowledge base, and counting the use frequency of various catheter models; if yes, determining the catheter model as a predicted catheter model, otherwise, selecting a standard catheter model from the standard catheter model set as the predicted catheter model according to a preset rule; When the catheter is judged to be selected from a standard catheter set, a history substitution record corresponding to the main doctor identifier under the current operation situation is acquired, difference information of the catheter actually used in each substitution operation and the standard catheter is extracted, one or more of average extra operation time length, auxiliary instrument use increment and contrast agent extra consumption caused by substitution are calculated based on the difference information to form a substitution cost vector, the substitution cost vector and a real-time pressure index of the current operation are subjected to weighted fusion to generate contextualized substitution cost scores, the contextualized substitution cost scores are used as negative weight factors in preset rules, and the contextualized substitution cost scores participate in the weighted scoring process of each model in the standard catheter set.

Description

Cardiovascular intervention catheter warehouse scheduling system based on demand analysis Technical Field The invention relates to the technical field of medical informatization, in particular to a cardiovascular intervention catheter warehouse scheduling system based on demand analysis. Background In the cardiovascular interventional operation process, the accurate scheduling of the catheter directly influences the operation efficiency and success rate, the traditional technology mainly adopts a static inventory management mode, the catheter requirement of dynamic change in the operation process cannot be met, along with the development of medical informatization, the prior art starts to dynamically deduct by combining with an operation plan map, and attempts to introduce doctor historical preference as a correction factor. However, these methods have obvious drawbacks, firstly, the combination mode of standard deduction results and doctor preferences is too simple, and a sequential processing mode is generally adopted, namely, the standard requirements are deducted and then replaced by preference models, and the intrinsic correlation between the two is not deeply mined, secondly, the existing system ignores the situation dependence of the doctor preferences, considers the preferences as static parameters which are fixed, and leads to serious deviation between the predicted results and the actual requirements in complex operation scenes, for example, the selection of a specific catheter by a doctor is often closely related to the characteristics of vascular lesions, the same doctor may show completely different catheter selection trends in different lesion situations, and the cleavage processing mode leads to the fact that the system cannot accurately simulate the decision logic of the doctor in the real operation environment, and often generates the predicted results which do not accord with reality. In addition, when the condition that the standard catheter is inconsistent with the preference catheter is faced, the prior art lacks a scientific evaluation mechanism to balance selection, so that the reduction of operation efficiency or the waste of medical resources are easily caused, and therefore, the development of an intelligent scheduling system capable of deeply fusing the real-time operation situation with the individual preference of doctors is needed to realize the real situation awareness prediction. Disclosure of Invention Aiming at the defects existing in the prior art, the invention provides a cardiovascular interventional catheter warehouse scheduling system based on demand analysis. In order to achieve the above object, the technical scheme of the present invention is as follows: in a first aspect, the present invention discloses a demand-based analysis cardiovascular interventional catheter depot scheduling system comprising: the data acquisition module is used for acquiring a surgical plan, a doctor mark of a main knife, a completed step sequence and a current angiography image; A feature extraction module for extracting a context feature vector from the current angiographic image, the context feature vector describing an anatomical state of a blood vessel; the standard deduction module is used for deducting through the operation plan knowledge graph according to the completed step sequence to obtain a possible subsequent standard step, a possible subsequent step code and a corresponding standard catheter set; The selection probability calculation module is used for inquiring a doctor preference knowledge base according to the doctor identification of the main cutter, the follow-up step codes and the situation feature vector to obtain catheter selection probability distribution; The first judging module is used for judging whether the selection probability of a certain catheter model in the catheter selection probability distribution exceeds a first preset threshold value, if yes, determining the catheter model as a predicted catheter model, otherwise, inquiring all historical operation data corresponding to the doctor identification of the main cutter and the current step code in the doctor preference knowledge base, and counting the use frequency of various catheter models; and the second judging module is used for judging whether the using frequency of a certain conduit model exceeds a second preset threshold value, if so, determining the conduit model as a predicted conduit model, and if not, selecting a standard conduit model from the standard conduit model set as the predicted conduit model according to a preset rule. In a second aspect, the invention discloses a cardiovascular interventional catheter depot scheduling method based on demand analysis, comprising the steps of obtaining a surgical plan, a doctor's identification of a primary knife, a sequence of completed steps, and a current angiographic image; Extracting a situation feature vector according to the current angiography