Invention content
The purpose of the present invention is to provide a kind of search method, device and terminals, user are facilitated to find more accurate similarThe case history of user improves user experience effect.
To achieve these goals, a kind of search method provided in an embodiment of the present invention, including：
At least one of active user's case history patient's condition project and at least one treatment project are obtained, wherein the patient's condition itemMesh is used to describe the diagnostic message and physiologic information of individual subscriber, and the treatment project is used to describe the diagnosis and treatment letter of individual subscriberBreath；
By active user's case history, matched with the case history of existing subscriber each in database, determine active user's case historyWith the similar value of patient's condition project in existing subscriber's case history in the database and the similar value of project is treated, and based on the diseaseThe similar value of the similar value of condition project and the treatment project, obtain in active user and the database it is one or moreThere is the matching degree of user；
According to the matching degree of active user and each existing subscriber, matching degree having more than predetermined threshold value is selectedUser.
Further, determining active user's case history and the similar value of existing subscriber's patient's condition project in the databaseStep includes：
Determine all patient's condition projects that active user's case history includes；
According to preset score value proportionate relationship between patient's condition project each in the database, will have in the databaseThe corresponding first default total score of all patient's condition projects of user, distributes to each patient's condition project of active user's case history, obtainsThe score value of each patient's condition project of active user's case history；
It determines the patient's condition project to match in existing subscriber's case history in active user's case history and the database, and calculates phaseThe sum of score value of patient's condition project is matched, it is similar to the patient's condition project of existing subscriber in the database to obtain active user's case historyValue.
Further, the patient's condition project includes one or more sub- patient's condition projects, wherein the sub- patient's condition project is usesThe detailed description project of the diagnostic message of family individual and the detailed description project of physiologic information,
The patient's condition project packet to match in the case history of existing subscriber in determining active user's case history and the databaseIt includes：
Judge corresponding patient's condition item in the patient's condition project of active user's case history and the case history of existing subscriber in the databaseBetween mesh whether have matched sub- patient's condition project, if so, in active user's case history and the database existing subscriber diseaseCondition project matches.
Further, the step of active user's case history and the similar value of the treatment project of existing subscriber in the database is determinedSuddenly include：
It determines the sum of all treatment projects that active user's case history includes, and judges whether the sum is less than the numberAccording to the project sum of all treatment projects of existing subscriber in library, the first result is obtained；
When first result is to be, by all treatment projects of existing subscriber in the database corresponding second in advanceIf total score, each treatment project of active user's case history is distributed to, obtains point of each treatment project of active user's case historyValue；
It determines the treatment project to match in the case history of existing subscriber in active user's case history and the database, and calculatesMatch the sum of score value for the treatment of project, and it is similar to the treatment project of existing subscriber in the database to obtain active user's case historyValue.
Further, determining active user's case history in the database treatment project of existing subscriber it is similarDuring the step of value, all treatment projects for judging active user's case history project sum, if less than in the databaseThe step of having the project sum of all treatment projects of user, obtaining after the first result, further includes：
First result for it is no when, judge active user's case history any treatment project and the database in haveIn the case history of user between corresponding treatment project, if there is matched at least one sub- treatment project, obtain the second result；
If second result is when being, by all treatment projects corresponding second of existing subscriber in the databaseDefault score value of the total score as the treatment project of active user's case history.
Further, the treatment project includes one or more sub- treatment projects, wherein the sub- treatment project is oneKind drug or a kind of therapeutic modality,
The treatment project to match in the case history of existing subscriber in determining active user's case history and the databaseStep includes：
Judge corresponding treatment item in the treatment project of active user's case history and the case history of existing subscriber in the databaseBetween mesh, if there is the matched first sub- treatment project, if so, existing subscriber in active user's case history and the databaseTreatment project match.
Further, the treatment item to match in the case history of existing subscriber in active user's case history and the database is determinedPurpose step further includes：
Judge corresponding treatment item in the treatment project of active user's case history and the case history of existing subscriber in the databaseBetween mesh, if having except the first sub- treatment project of sub- matched residue for the treatment of project foreign minister, if so, active user's case history withThe treatment project of existing subscriber matches in the database.
Further, the treatment item to match in the case history of existing subscriber in active user's case history and the database is determinedPurpose step further includes：
Judge whether all treatment projects in active user's case history have the time being correspondingly arranged, if so, according to it is described whenBetween sequencing the treatment project in active user's case history is ranked up, obtain the first ranking results；
Judge the second sequence that first ranking results sort with existing subscriber in the database according to time order and functionAs a result between, if there is the treatment project that time sequencing matches, if so, in active user's case history and the databaseThe treatment project for having user matches.
The embodiment of the present invention also provides a kind of retrieval device, including：
Acquisition module, for obtaining at least one of active user's case history patient's condition project and at least one treatment project,Wherein described patient's condition project is used to describe the diagnostic message and physiologic information of individual subscriber, and the treatment project is used for descriptionThe medical information of family individual；
Matching module for by active user's case history, being matched with the case history of existing subscriber each in database, determinesActive user's case history and the similar value of patient's condition project in existing subscriber's case history in the database and the similar value for the treatment of project,And similar value and the similar value of the treatment project based on the patient's condition project, it obtains in active user and the databaseThe matching degree of one or more existing subscribers；
Selecting module for the matching degree according to active user and each existing subscriber, is selected matching degree and is more thanThe existing subscriber of predetermined threshold value.
Further, the matching module includes：
First determination sub-module, for determining all patient's condition projects that active user's case history includes；
First distribution sub module, for being closed according to score value ratio preset between patient's condition project each in the databaseSystem by the corresponding first default total score of all patient's condition projects of existing subscriber in the database, distributes to active user's diseaseThe each patient's condition project gone through obtains the score value of each patient's condition project of active user's case history；
Second determination sub-module, for determining in existing subscriber's case history to match in active user's case history and the databasePatient's condition project, and calculate the sum of the score value for the patient's condition project that matches, obtain active user's case history with having in the databaseThe patient's condition project similar value of user.
Further, the patient's condition project includes one or more sub- patient's condition projects, wherein the sub- patient's condition project is usesThe detailed description project of the diagnostic message of family individual and the detailed description project of physiologic information,
Second determination sub-module includes：
First judges matching unit, for judging existing subscriber in the patient's condition project of active user's case history and the databaseCase history in whether there is matched sub- patient's condition project between corresponding patient's condition project, if so, active user's case history and describedThe patient's condition project of existing subscriber matches in database.
Further, the matching module includes：
It determines judging submodule, for determining the sum of all treatment projects that active user's case history includes, and judges instituteThe project sum whether sum is less than all treatment projects of existing subscriber in the database is stated, obtains the first result；
Second distribution sub module, in first result for when being, by the database existing subscriber it is allThe corresponding second default total score for the treatment of project distributes to each treatment project of active user's case history, obtains active user's diseaseThe score value of each treatment project gone through；
Third determination sub-module, for determining in active user's case history and the database phase in the case history of existing subscriberThe sum of the treatment project matched, and calculate the score value for the treatment project that matches, it obtains in active user's case history and the databaseThere is the treatment project similar value of user.
Further, the matching module further includes：
First judging submodule, for when first result is no, judging any treatment item of active user's case historyIn mesh and the database in the case history of existing subscriber between corresponding treatment project, if there is matched at least one son to controlTreatment project obtains the second result；
Third distribution sub module, if being when being, by the institute of existing subscriber in the database for second resultThere is score value of the corresponding second default total score for the treatment of project as the treatment project of active user's case history.
Further, the treatment project includes one or more sub- treatment projects, wherein the sub- treatment project is oneKind drug or a kind of therapeutic modality,
The third determination sub-module includes：
Second judges matching unit, for judging existing subscriber in the treatment project of active user's case history and the databaseCase history between corresponding treatment project, if there is the matched first sub- treatment project, if so, active user's case history andThe treatment project of existing subscriber matches in the database.
Further, the third determination sub-module further includes：
Third judges matching unit, for judging existing subscriber in the treatment project of active user's case history and the databaseCase history between corresponding treatment project, if having except the first sub- treatment project of the sub- matched residue for the treatment of project foreign minister,If so, the treatment project of existing subscriber matches in active user's case history and the database.
Further, the third determination sub-module further includes：
Judge sequencing unit, all treatment projects for judging in active user's case history whether have be correspondingly arranged whenBetween, if so, the treatment project in active user's case history is ranked up according to time order and function sequence, obtain the first sequence knotFruit；
4th judges matching unit, for judge in first ranking results and the database existing subscriber according to whenBetween between the second ranking results for successively sorting, if there is the treatment project that time sequencing matches, if so, active userThe treatment project of existing subscriber matches in case history and the database.
Correspondingly, the embodiment of the present invention also provides a kind of terminal, including retrieving device as described above.
The above-mentioned technical proposal of the present invention has the beneficial effect that：
In the solution of the present invention, by the way that the information categorization in active user's case history is divided into disparity items, then in numberIt is matched according to search in library, can be obtained more accurate similar users by more careful project search, facilitate use in this wayThe user's case history to match with itself case history is found at family, improves user experience effect.
Specific embodiment
To make the technical problem to be solved in the present invention, technical solution and advantage clearer, below in conjunction with attached drawing and toolBody embodiment is described in detail.
The problem of present invention is difficult to find that accurate similar ward mate for user in the prior art, provides a kind of retrieval sideMethod, device and terminal facilitate user search out more accurate similar ward mate, improve user experience effect.
As shown in Figure 1, a kind of search method of the embodiment of the present invention, including：
Step 11, at least one of active user's case history patient's condition project and at least one treatment project, wherein institute are obtainedPatient's condition project is stated for describing the diagnostic message and physiologic information of individual subscriber, the treatment project is used to describe individual subscriberMedical information；
Step 12, it by active user's case history, is matched with the case history of existing subscriber each in database, determines current useFamily case history and the similar value of patient's condition project in existing subscriber's case history in the database and the similar value for the treatment of project, and be based onThe similar value of the similar value of the patient's condition project and the treatment project, obtain in active user and the database one orThe matching degree of multiple existing subscribers；
Step 13, it according to the matching degree of active user and each existing subscriber, selects matching degree and is more than default thresholdThe existing subscriber of value.
In above-mentioned steps 11 to step 13, by the way that the information categorization in active user's case history is divided into disparity items,Then search is matched in the database, can obtain more accurate similar users by more careful project search in this way,It facilitates user and finds the user's case history to match with itself case history, improve user experience effect.
It should be noted that：Wherein described diagnostic message refers to for each user, the implant treatment of diagnosis, illnessRelevant medical history, information, the physiologic information such as disease progression stage, diagnosis information refer to for each user, useThe height at family, the information such as weight, age, certain diagnostic message and physiologic information are not illustrated one by one herein, any to belong to thisThe similar information of invention, all belongs to the scope of protection of the present invention.
In addition, it is described matching refer in information identical with active user in the database or the database withThe information of active user is in same preset range, for example the height difference of user is less than in the height of active user and databaseThe age gap of user is less than 5 years old in 5 centimetres, such as the age of active user and database, certainly above to be merely illustrative,It is any to match similar mode with the present invention, it all belongs to the scope of protection of the present invention, differ a citing herein.
In addition, above-mentioned predetermined threshold value situation can be set according to demand, for example it is the precision of user for needing to search forIt is required that when relatively low, the predetermined threshold value can be set as to 70 percent, than user's claimed accuracy if desired for search compared withThe predetermined threshold value can be then set as 90 percent by height.It can thus be searched out and with the preferable user of matching degree.
Operative treatment can be referred to for the medical information, chemotherapy, endocrine therapy, targeted therapy, radiotherapy replacesOne or more therapeutic modalities of generation treatment.
Since active user is when inputting patient's condition project, the information of input is not necessarily patient's condition item all in databaseMesh, therefore in active user when input patient's condition project scans for, it is ill in order to guarantee the institute for inputting active userCondition project can be compared with the data in database, and in the search method of the embodiment of the present invention, step 12 includes：
Step 121, all patient's condition projects that active user's case history includes are determined；
Step 122, according to preset score value proportionate relationship between patient's condition project each in the database, by the dataThe corresponding first default total score of all patient's condition projects of existing subscriber in library distributes to each patient's condition item of active user's case historyMesh obtains the score value of each patient's condition project of active user's case history；
Step 123, the patient's condition project to match in existing subscriber's case history in active user's case history and the database is determined,The sum of and calculate the score value for the patient's condition project that matches, obtain active user's case history and the patient's condition item of existing subscriber in the databaseMesh similar value.
It should be noted that：The score value proportionate relationship is the importance of the patient's condition project in case history, and to eachA identical, similar patient's condition project assigns different fractional values.
In addition, the first default total score can refer to the numerical value for being more than 60 points and being less than in the range of 100 points, toolAccording to patient's condition project, the occupied proportion in user's case history is configured the numerical value of body, such as generally with 100 points for benchmark, instituteIt accounts for the numerical value depending on 100 points of proportion, but herein and is not construed as limiting, any numerical value that can be realized through the embodiment of the present invention,It all belongs to the scope of protection of the present invention.
The distribution of above-mentioned score proportion mainly has following some reasons, and user is filled out around suffered from tumor typeThe sundry item write, and the remedy measures of tumour are also to be distinguished according to tumor type, therefore are to need first to determineComparative information suffers from tumor type, and the similar users that user search each in this way comes out are just significant, therefore, suffer from tumour classScore proportion shared by type is larger.It is each swollen since tumour is substantially body cell malignant proliferation caused by a kind of gene mutationThe driving gene of knurl directly determines the selectable targeted drug of patient and prognosis situation, therefore for each tumor cells partingTo judge the uniqueness of individual institute's illness, and then be easier to determine user similar in database, therefore tumor cells pointIt is higher that the score value of type also occupies proportion.
During due to being matched to patient's condition project, be according to judge each patient's condition project particular content whether matching judgment, and it is not merely the project of patient's condition project one by one, therefore in the search method of the embodiment of the present invention, the patient's condition project packetInclude one or more sub- patient's condition projects, wherein the sub- patient's condition project for the diagnostic message of individual subscriber detailed description project withAnd the detailed description project of physiologic information, step 123 include：
Step 1231, judge corresponding in the patient's condition project of active user's case history and the case history of existing subscriber in the databasePatient's condition project between whether have matched sub- patient's condition project, if so, in active user's case history and the database haveThe patient's condition project of user matches.
In the embodiment of the present invention, the detailed description project of above-mentioned diagnostic message, for example be the presence or absence of family's medical history, symptom isFever or insomnia etc., the detailed description project of physiologic information, for example the man or female of gender, marital status are married, unmarried etc..
In the embodiment of the present invention, the flow of realization can be as follows：
First, now in the database by all one original total scores of patient's condition item setup；
Then, judge the number of active user's case history patient's condition option, be less than in the number of active user's case history patient's condition optionDuring total option number of the patient's condition project in database, need to carry out the total score of original setting to be reassigned to active user's diseaseThe each patient's condition option gone through；
Subsequently, the sub- patient's condition project of each patient's condition project of active user is matched, in sub- patient's condition projectAppearance is when matching, the affiliated patient's condition project of sub- patient's condition project that can be just matched, the score value redistributed；
Finally, the obtained score value redistributed is added, is exactly the case history of active user and the patient's condition project of databaseSimilar value.
Above-mentioned essence be in order to obtain similar value by matching after, and then find the user similar to active user's case history.Therefore it is redistributed for the comparison of multiple sub- patient's condition projects and by the score value in database, is provided to determine active userThe importance of case history, and then more accurate similar users are found, improve the experience effect of user.
The specific implementation of the specific embodiment of the present invention is exemplified below.
Assuming that share 16 projects (this certain project with what the patient's condition project of the case history in database was compared and appraisedIt is not construed as limiting, is likely to increase in follow-up expansion or application, the embodiment of the present invention is intended merely to facilitate what is understood and illustrateProject numerical value), and the total value of the patient's condition project is assigned as 70 points (this certain score value does not limit, but for protrusionThe importance of patient's condition project, this score value account for more than half of total score, and herein only for convenience of description, distribution score value is 70 points).
First check for whether two users all fill in and selected relevant options, only (one is currently to use there are two userFamily, one is the user in database) all selected relevant options, then the project is just compared, and is selected according to two usersWhether that selects is identical and judge score.It such as detects user and does not fill in any one project, then it cannot be with the user data of databaseIt is compared, also not score, during as in project that user fills in and selects not comprising project 5 (suffer from tumor type), is also carriedShow that active user must be filled in, otherwise cannot carry out subsequent retrieval and comparison task.
(1) such as active user is filled in and has been selected in 16 projects, then 16 projects is assigned and divided, accordinglyStep is compared, score, algorithm and formula are as follows：
1st, project 1, tumour medical history, the project have 3 options, are respectively：Primary tumo(u)r, has shifted many places primary tumo(u)rTumour, two users have selected mutually same item up to 2 points in 3 options.
2nd, project 2, recurrence, the project have 3 options, are respectively：Without recurrence, single recurrence, multiple relapse, such as twoA user has selected mutually same item up to 2 points in 3 options.
3rd, project 3, Family history of cancer, the project have 2 options, are respectively：There are family history, such as no family history, two useFamily has selected mutually same item up to 2 points in 2 options.
4th, project 4, other medical histories, the project have 9 options, are respectively：Nothing, hypertension, diabetes, chronic nephritis, firstHigh, rheumatoid arthritis, systemic loupus erythematosus, hepatic sclerosis, pulmonary tuberculosis, such as two users have selected identical in 9 optionsOne up to 2 points.
5th, tumor type is suffered from project 5, which has 14 options, is respectively：Liver cancer, lung cancer, cancer of pancreas, breast cancer,Cervical carcinoma, carcinoma of endometrium, colorectal cancer, gastric cancer, lung cancer, nasopharyngeal carcinoma, oophoroma, prostate cancer, the cancer of the brain, leukaemia, such as userTumor type be not belonging to more than 14 kinds, then drawings family can voluntarily add and be compared again, such as two users in these optionsMutually same item has been selected up to 20 points.(according to the diagnosis and treatment rule of current tumour, if tumor invasion position is identical, i.e. tumor typeIdentical, corresponding diagnose and treat measure is also identical, so the score value of suffered from tumor type occupies higher proportion)
6th, project 6, tumor cells parting, the project are more options project, and option is respectively：P-gP is positive, II sun of TOPOProperty, GST π are positive, Ki67 is positive, ER (estrogen receptor) is positive, PR (progesterone receptor) is positive, HER2 (ERBB2) is overexpressed,HER2 gene magnifications, HER2 genes Exon20 mutation, VEGFR1 is overexpressed, VEGFR2 is overexpressed, VEGFR3 is overexpressed, PTEN mistakesExpression, the amplification of PTEN gene delections, EGFR gene, EGFR gene Exon18 mutation, EGFR gene Exon19 mutation, EGFR geneExon20 mutation, EGFR gene Exon21 mutation, EGFR gene T790M mutation, KRAS gene Cs ode12 mutation, KRAS genesCode13 mutation, KRAS gene Cs ode61 mutation, KRAS gene Cs ode146 mutation, NRAS gene Cs ode12 mutation, NRAS genesCode13 mutation, NRAS gene Cs ode61 mutation, BRAF gene Exon11 mutation, BRAF gene Exon15 mutation, BRAF geneV600E site mutations, PIK3CA genes Exon9 mutation, PIK3CA genes Exon20 mutation, EML4-ALK fusions, ROS1Gene rearrangement, RET Gene Fusions, NTRK1 Gene Fusions, FGFR1 gene magnifications, FGFR2 gene magnifications, DDR2 gene mutations,P53 gene mutations, KIT (C-kit) genes Exon9 mutation, KIT (C-kit) genes Exon11 mutation, KIT (C-kit) geneExon13 mutation, KIT (C-kit) genes Exon17 mutation, PDGFRA (PDGFR α) gene Exon12 mutation, PDGFRA (PDGFRα) gene Exon14 mutation, PDGFRA (PDGFR α) gene Exon18 mutation, C-MET (MET) gene magnification, MEK1 genes are dashed forwardChange, JAK2 genes V617F mutation, RET gene Ms 918T mutation, PDL1 protein overexpressions, PD1 protein overexpressions, fusionBCR-ABL, fusion AML1-ETO, fusion PML-RARa, fusion TEL-AML1.As two users are a few at thisMutually same item has been selected in option up to 10 points.As 2 and more than 2 options it is identical if 20 points.It (is precisely treated according to tumourTheory, the number and type of the driving gene mutation of tumour, directly determine targeted drug that tumor patient can select in advanceSituation afterwards, so to occupy proportion higher for the score value of tumor cells parting project)
7th, project 7, initial diagnosis time are compared according to the time that user fills in, as identical such as the time filled in, and are sentencedObtain 2 points.
8th, project 8, symptom of going to a doctor for the first time, the option which includes have：Pectoralgia, abdomen pain, hemoptysis, fever fever, blood urine,Have blood in stool, as the tumor type of user be not belonging to more than 6 kinds, then drawings family can voluntarily add and be compared again, such as two users existMutually same item has been selected in these options up to 2 points.
9th, which phase project 9 are currently in, which includes 4 options, is respectively：In early days, mid-term, middle and advanced stage, eveningPhase.As two users have selected mutually same item up to 4 points in 4 options.
10th, project 10, progression of disease situation, the project include 3 options, are respectively：TTP (tumour progression time), PFS(progression free survival phase), TTF (treatment failure time), such as two users have selected mutually same item up to 2 points in 3 options.
11st, project 11, symptom, side effect, the option which includes are respectively：Vomiting, diarrhea, insomnia, constipation, hairHot, weak, nervous, alopecia, hectic fever, ostalgia, canker sore, pigmentation, erythema, excoriation, skin ulceration, suffering limbOedema, suffering limb acid is swollen, praise difficulty, skin blister, suffering limb numbness, suffering limb shouting pain, colpoxerosis, colporrhagia, osteopyrexia and fever,Blood urine, such as two users have selected mutually same item up to 2 points in these options.
12nd, project 12, age, this project are compared according to the age that user fills in, if the age gap of two usersLittle Yu not be 5 years old, that is, judge identical, judge to obtain 2 points.
13rd, project 13, gender, this project are compared according to the gender that user fills in, and option is man, female, is selected identicalUp to 2 points.
14th, project 14, height carry out classification, if the height difference of two users herein according to the height that user fills inLess than 5cm, that is, judge to obtain 2 points, difference is less than 10cm, but is greater than 5cm and obtains 1 point.
15th, project 15, weight, weight is filled in user's selection herein, if the bodyweight difference of two users is obtained less than 10kgDivide 2 points, such as two user's weight difference are more than 10kg, but then score 1 is divided less than 20kg.
16th, project 16, obsterical history, there are three options herein, are respectively：It is unmarried, married, married to have educated, such as two usersIdentical one has been selected, then has judged to obtain 2 points.
Finally, if active user fills in all items, the case history project calculation formula can also be used：Case history projectTotal score SUMn=projects 1 score+project 2 score+project 3 score+... the score of project 16；
Certainly, as active user only carried out in n project filling in and select (n is integer, and 0 ＜ n ＜ 16), andProject 5 (suffering from tumor type) is contained, then only to the n project number with each user in the database of active user's selectionAccording to being compared, in database other users more than n or fill up 16, the n items that active user is also only selected to fill inIt is compared, the First Principles of comparison are to adjust the score of each single item, that is, give active user's n project of selection, one pointNumber adjustment weight, the number of options that the existing subscriber in this way in database fills in dynamically adjust the fractional value of each single item,To reach the ward mate for having maximum matching degree with active user.Corresponding retrieval logic and step are as follows：
1. it is first, similar as above, when first assuming that active user fills in complete 16 project, according to the score of each projectSituation calculates the total score of n project of its selection.Corresponding formula S UMn=projects x1Score value+project x2Score value+...Project xnScore value.
2. calculating n project filling in of active user, the score of each single item adjusts weight, will script mean allocation in 16The score (70 points) of a project, re-starts distribution in n project.Distributing formula is：The score value of project x=original project x'sScore value × (70 ÷ SUMn).
3. according to the n project that active user's selection is filled in, compared successively, the identical score of option, selection differsThen not score.
In addition, since active user is when inputting treatment project, the information of input is not necessarily all in database controlTreatment project, therefore in active user when input treatment project scans for, in order to guarantee the institute for inputting active userThere is treatment project that can be compared with the data in database, in the search method of the embodiment of the present invention, step 12 includes：
Step 124, it determines the sum of all treatment projects that active user's case history includes, and judges whether the sum is smallThe project sum of all treatment projects of existing subscriber, obtains the first result in the database；
Step 125, in first result for when being, all treatment projects of existing subscriber in the database are corresponded toThe second default total score, distribute to each treatment project of active user's case history, obtain each treatment of active user's case historyThe score value of project；
Step 126, the treatment item to match in the case history of existing subscriber in active user's case history and the database is determinedMesh, and calculate the sum of the score value for the treatment project that matches obtains active user's case history and is controlled with existing subscriber in the databaseTreatment project similar value.
Traditional main user to similary cancer species distinguishes, more general due to distinguishing, and due to eachThere are many hypotypes again for kind cancer, and such classification is more inaccurate, and it is even more impossible to controlling between user is helped to be comparedTreatment mode, the drug taken, symptom, medical history etc. can not provide accurate numerical value to the similarity degree of each informationProperty judge.Therefore, the identical retrieval of data progress of patient's condition project of the embodiment of the present invention based on case history and treatment project, according to diseaseThe importance of condition project and treatment project carries out assignment, totalling etc. to information respectively and searches similar ward mate.
During due to being matched for treatment project, particular by judge each treatment project particular content whetherMatch, and be not merely the project for treating project one by one, therefore in the search method of the embodiment of the present invention, the treatment project packetOne or more sub- treatment projects are included, wherein the sub- treatment project is wrapped for a kind of drug or a kind of therapeutic modality, step 126It includes：
Step 1261, judge corresponding in the treatment project of active user's case history and the case history of existing subscriber in the databaseTreatment project between, if having the matched first sub- treatment project, if so, in active user's case history and the databaseThe treatment project of existing subscriber matches.
After step 124, step 12 further includes：Step 127, when first result is no, judge active userIn any treatment project of case history and the database in the case history of existing subscriber between corresponding treatment project, if havingAt least one sub- treatment project matched, obtains the second result；
Step 128, if second result is when being, by all treatment projects pair of existing subscriber in the databaseScore value of the second default total score answered as the treatment project of active user's case history.
In the embodiment of the present invention, the flow of realization can be as follows：
First, now in the database by all treatment one original total scores of item setup；
Then, judge the number of active user's case history treatment option, be less than in the number of active user's case history treatment optionDuring total option number of the treatment project in database, need to carry out the total score of original setting to be reassigned to active user's diseaseThe each treatment option gone through；
Subsequently, the sub- treatment project of each treatment project of active user is matched, in sub- treatment projectAppearance is when matching, treatment project, the score value redistributed belonging to the sub- treatment project that can be just matched；
Finally, the obtained score value redistributed is added, is exactly the case history of active user and the treatment project of databaseSimilar value.
Above-mentioned essence is in order to by matching to obtain similar value, and then finds the user similar to active user's case history.ThereforeComparison for multiple sub- treatment projects and the score value in database is redistributed, be provided to determine active user's case historyImportance, and then find more accurate similar users, improve the experience effect of user.
If in the number for judging active user's case history treatment option, it is less than in the number of active user's case history treatment optionDuring total option number of the treatment project in database, directly the sub- treatment project for the project for the treatment of is matched, is there is at least oneAfter item successful match, the total score of original setting is directly obtained.Also simplify matching rule in this way, when the project for the treatment of is more, noWith going to compare one by one, as long as there is a successful match, it is considered as matching, quickly finds similar user.
Specifically, in order to reduce the number of search, search time is saved, it is quick as possible and accurately find similar useFamily, in the search method of the embodiment of the present invention, step 12 includes：
Step 129, judge corresponding in the treatment project of active user's case history and the case history of existing subscriber in the databaseTreatment project between, if having except the first sub- treatment project of sub- matched residue for the treatment of project foreign minister, if so, current useThe treatment project of existing subscriber matches in family case history and the database.
In addition, in the search method of the embodiment of the present invention, step 12 includes：
Step 130, judge whether all treatment projects in active user's case history have the time being correspondingly arranged, if so, pressingThe treatment project in active user's case history is ranked up according to time order and function sequence, obtains the first ranking results；
Step 131, judge what first ranking results sorted with existing subscriber in the database according to time order and functionBetween second ranking results, if there is the treatment project that time sequencing matches, if so, active user's case history and the numberTreatment project according to existing subscriber in library matches.
In the embodiment of the present invention, treatment project sequencing is matched, tax score value during each single item information matchesWhen (score value can be constantly adjusted and optimize on backstage) and active user only fill in partial information and are retrieved, informationThe fractional value of comparison is adjusted into Mobile state, and different fractional values are assigned, and then obtain similar to active user's case history according to importanceDatabase in user.
It should be noted that adjustment has also been made in the case of active user only selects to fill in partial information, i.e., only to work asInformation that preceding user fills in carries out single keyword (i.e. a sub- treatment project), multi-key word (i.e. multiple sub- treatment projects), suitableSequence retrieval (treats sequence of the project according to the time), and corresponding extension set system of assigning also is optimized and adjusts, to reach retrieval knotThe optimization of fruit helps active user to find the people most like with his state of an illness.
The specific implementation of the specific embodiment of the present invention is exemplified below.
Comprising six projects in " the treatment record " of the case history of each user, each project includes one or more select, wherein each option can be regarded as a keyword, it is respectively：
Project 1：Operative treatment (option includes two options, has the or no)：
Project 2：Chemotherapy (option includes multiple options, and each option is a kind of drug)：
Project 3：Endocrine therapy (option includes multiple options, and each option is a kind of drug)：
Project 4：Targeted therapy (option includes multiple options, and each option is a kind of drug)：
Project 5：Radiotherapy (option includes, have or no)：
Project 6：Replacement therapy (option includes multiple options, and each option is a kind of therapeutic modality)：
Has the step of case history treatment information search comparison, score mode of patient data for active user and databaseIt is divided into three steps with rule, respectively single keyword search, multi-key word search, sequential search：
1. single keyword search：Using 6 projects above as range, the project and database filled in active user's selection areThe data for having patient carry out comparison scoring.This search total score is 10 points, and (this certain score value does not limit, and is intended merely to facilitateIt is bright, but if in order to emphasize the list key search, then can separately make score value distribution).
, there are one keywords wherein in chemotherapy, endocrine therapy, targeted therapy, replacement therapy this 4 projects (i.e.Have a drug or therapeutic modality) it is identical, then judge that the project is identical.Assuming that active user selects the item number filled in as n (nFor integer, and 0≤n≤6), active user it is identical in m project with user a certain in database (m is integer, and 0≤m≤N), score principle is adjusted according to the size of number of items n that active user's selection is filled in into Mobile state, specific as follows：
1) if active user's selection fills in the project of 3 or more.If optional 3 project in 3 projects, has anyProject is identical then to obtain 10 points.
If 2) active user's selection fills in 2 projects, comparison and tax point principle are to divide if first project is assigned4 points, identical score 4 is divided, and second project, which is assigned, divides 6 points, and identical score 6 is divided, the identical not score of 0 project.
It, must not if differing if 1 project is identical up to 10 points if 3) active user's selection fills in 1 projectPoint.
4) if active user does not select to fill in any project, not score.
In single keyword search by rear, multi-key word search is just carried out.
2. multi-key word is searched for, carry out analysis on the basis of the first step, i.e., two users are in such as chemotherapy, endocrineTreatment, targeted therapy, replacement therapy four have in the project of multiple options, analyse whether that the option of 2 or more is identical and (include2), the multi-key word search result score of the project is then judged in this way.This search total score is 10 points, and (this certain score value is not doneIt limits, merely for convenience of description).Assuming that active user selects the item number filled in as n (n is integer, and 0≤n≤4), whenPreceding user is identical in m project with user a certain in database (m is integer, and 0≤m≤n), then multi-key word score principleIt is as follows with formula：
1) if active user's selection fills in the project of 3 or more.Optional 3 project, if in 3 projects, in addition to listOutside keyword, also at least one mesh is identical then obtains 10 points.
If 2) active user's selection fills in 2 projects, comparison and tax point principle, which are that first project is assigned, divides 4 points,Identical score 4 is divided, and second project, which is assigned, divides 6 points, and identical score 6 is divided, and the identical not score of 0 project (it is identical to refer to, in addition to single passOutside keyword, also at least one mesh is identical).
It, must not if differing if 1 project is identical up to 10 points if 3) active user's selection fills in 1 projectPoint.
4) as n=0, multi-key word search score is drawn a blank.
When multi-key word is searched for by rear, sequential search is just carried out.
3. sequential search：On the basis of the first step, i.e., the project selected according to active user is ranked up, it is assumed that currentUser selects to fill in n project (n is integer, and 0≤n≤6).It need to call what active user inputted when filling in classification during sequenceTime (without input time project then without sequence, all items do not have the time then without this compare).According to whenBetween the precedence that occurs, the project that user's selection is filled in active user, database is ranked up, is followed successively by respectivelyM1、M2…Mn.When active user is identical with the 1st project of user in database, then M1=1 is enabled, such as the two is in the 1st itemMesh is different, then enables M1=0；Then the comparison of the 2nd project, n-th project is carried out, it is identical, Mn=1 is enabled, is differed, enables Mn=0；Assign point principle and the formula of sequential search are adjusted according to the number of options that active user fills in into Mobile state, due to selecting firstThe remedy measures selected assign higher score weight for patient effect's maximum, therefore for project in the top.SpecificallyIt is as follows：
If 1) active user selection fill in more than 3 project, optional 3 project, if in 3 projects, sequence firstAt least one of project mesh it is identical, then 10 points (this certain score value does not limit, merely for convenience of description).
If 3) active user's selection fills in 2 projects, first item is assigned score value 6 and is divided, and 6 are obtained when clooating sequence is identicalPoint, second item is assigned score value 4 and is divided, and 4 points are obtained when clooating sequence is identical.
If 4) active user selection fill in 1 project, sort first project assign score value 10 divide, then must not differPoint.
5) if active user's selection fills in 0 project, not score.
By the way that the similar value of searchers Yu database each user is carried out descending arrangement, matched degree is retrievedHighest ward mate.
Finally, after whole realization flow as shown in Figure 2, active user's matching degree of the embodiment of the present invention calculates public affairsFormula：Matching degree=(patient's condition project score+mono- keyword search score+multi-key word search score (in the presence of)+sequential searchScore (in the presence of)) divided by 100.
Each active user after the search is completed, the information of active user is stored, therefore in the embodiment of the present inventionIt includes remedy measures in case history, drug, symptom etc. and has more information, it is convenient subsequently more accurately to find similar users.
Correspondingly, as shown in figure 3, in a kind of retrieval device of the embodiment of the present invention, including：
Acquisition module 31, for obtaining at least one of active user's case history patient's condition project and at least one treatment itemMesh, wherein the patient's condition project is used to describe the diagnostic message and physiologic information of individual subscriber, the treatment project is used to retouchState the medical information of individual subscriber；
Matching module 32, for by active user's case history, being matched with the case history of existing subscriber each in database, reallyDetermine the similar of active user's case history and the similar value of patient's condition project in existing subscriber's case history in the database and treatment projectValue, and the similar value of the similar value based on the patient's condition project and the treatment project, obtain active user and the dataThe matching degree of one or more existing subscribers in library；
For the matching degree according to active user and each existing subscriber, it is big to select matching degree for selecting module 33In the existing subscriber of predetermined threshold value.
By the way that the information categorization in active user's case history is divided into disparity items, then search progress in the databaseMatch, can more accurate similar users be obtained by more careful project search in this way, facilitate user and find and itself case historyThe user's case history to match, improves user experience effect.
In the retrieval device of another embodiment of the present invention, the matching module 32 includes：
First determination sub-module, for determining all patient's condition projects that active user's case history includes；
First distribution sub module, for being closed according to score value ratio preset between patient's condition project each in the databaseSystem by the corresponding first default total score of all patient's condition projects of existing subscriber in the database, distributes to active user's diseaseThe each patient's condition project gone through obtains the score value of each patient's condition project of active user's case history；
Second determination sub-module, for determining in existing subscriber's case history to match in active user's case history and the databasePatient's condition project, and calculate the sum of the score value for the patient's condition project that matches, obtain active user's case history with having in the databaseThe patient's condition project similar value of user.
In the retrieval device of another embodiment of the present invention, the patient's condition project includes one or more sub- patient's condition projects,Wherein described sub- patient's condition project is the detailed description project of diagnostic message of individual subscriber and the detailed description item of physiologic informationMesh,
Second determination sub-module includes：
First judges matching unit, for judging existing subscriber in the patient's condition project of active user's case history and the databaseCase history in whether there is matched sub- patient's condition project between corresponding patient's condition project, if so, active user's case history and describedThe patient's condition project of existing subscriber matches in database.
In the retrieval device of another embodiment of the present invention, the matching module 32 includes：
It determines judging submodule, for determining the sum of all treatment projects that active user's case history includes, and judges instituteThe project sum whether sum is less than all treatment projects of existing subscriber in the database is stated, obtains the first result；
Second distribution sub module, in first result for when being, by the database existing subscriber it is allThe corresponding second default total score for the treatment of project distributes to each treatment project of active user's case history, obtains active user's diseaseThe score value of each treatment project gone through；
Third determination sub-module, for determining in active user's case history and the database phase in the case history of existing subscriberThe sum of the treatment project matched, and calculate the score value for the treatment project that matches, it obtains in active user's case history and the databaseThere is the treatment project similar value of user.
In the retrieval device of another embodiment of the present invention, the matching module 32 further includes：
First judging submodule, for when first result is no, judging any treatment item of active user's case historyIn mesh and the database in the case history of existing subscriber between corresponding treatment project, if there is matched at least one son to controlTreatment project obtains the second result；
Third distribution sub module, if being when being, by the institute of existing subscriber in the database for second resultThere is score value of the corresponding second default total score for the treatment of project as the treatment project of active user's case history.
In the retrieval device of another embodiment of the present invention, the treatment project includes one or more sub- treatment projects,Wherein described sub- treatment project be a kind of drug or a kind of therapeutic modality,
The third determination sub-module includes：
Second judges matching unit, for judging existing subscriber in the treatment project of active user's case history and the databaseCase history between corresponding treatment project, if there is the matched first sub- treatment project, if so, active user's case history andThe treatment project of existing subscriber matches in the database.
In the retrieval device of another embodiment of the present invention, the third determination sub-module further includes：
Third judges matching unit, for judging existing subscriber in the treatment project of active user's case history and the databaseCase history between corresponding treatment project, if having except the first sub- treatment project of the sub- matched residue for the treatment of project foreign minister,If so, the treatment project of existing subscriber matches in active user's case history and the database.
In the retrieval device of another embodiment of the present invention, the third determination sub-module further includes：
Judge sequencing unit, all treatment projects for judging in active user's case history whether have be correspondingly arranged whenBetween, if so, the treatment project in active user's case history is ranked up according to time order and function sequence, obtain the first sequence knotFruit；
4th judges matching unit, for judge in first ranking results and the database existing subscriber according to whenBetween between the second ranking results for successively sorting, if there is the treatment project that time sequencing matches, if so, active userThe treatment project of existing subscriber matches in case history and the database.
It should be noted that：The information number that case history is included in case history comparison process, when each information is identical, closeFractional value be all adjustable, will necessarily also combine user and be continued to optimize using the feedback of, clinical expert and perfect.Because appointingThe process of what change case history comparison, the information number that case history includes, the forms such as change of fractional value belong to the embodiment of the present inventionProtection domain.
It should be noted that device provided by the invention is the device using above-mentioned search method, then above-mentioned search methodAll embodiments be suitable for the device, and can reach the same or similar advantageous effect.
Correspondingly, the embodiment of the present invention also provides a kind of terminal, including retrieving device as described above.
The retrieval device of the embodiment of the present invention is applied at intelligent mobile phone terminal or PC ends, with application program of mobile phone or websiteForm is presented, and after the medical record information of oneself is carried report up by active user, passes through the disease with the existing patient of background data baseIt counts one by one according to the same or similar inquiry of progress.
Accordingly due to the retrieval device of the embodiment of the present invention, applied to terminal, therefore, the embodiment of the present invention additionally providesA kind of terminal, wherein, the realization embodiment of above-mentioned retrieval device can also reach phase suitable for the embodiment of the terminalSame technique effect.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the artFor, without departing from the principles of the present invention, several improvements and modifications can also be made, these improvements and modificationsIt should be regarded as protection scope of the present invention.