CN1877588A

CN1877588A - Computerized medical diagnostic system utilizing list-based processing

Info

Publication number: CN1877588A
Application number: CNA200510128901XA
Authority: CN
Inventors: 埃德温·C·伊利夫
Original assignee: FIRST ADVISORY Co
Current assignee: FIRST ADVISORY Co; First Opinion Corp
Priority date: 1996-07-12
Filing date: 1997-07-11
Publication date: 2006-12-13
Also published as: CN1783099A

Abstract

A system and method for providing computerized, knowledge-based medical diagnostic advice. The medical advice is provided to the general public over a network, such as a telephone network with the use of a telephone or the Internet with the use of an Internet access device. Alternatively, the medical advice can be provided to a patient in a stand-alone mode by use of a computer. The invention utilizes a list-based processing method of generating and executing diagnostic scripts. For the purpose of diagnosing a health problem of a patient, medical knowledge is organized into a list of the diseases to be considered. Each disease on the disease list includes a list of symptoms that is checked in a patient. Each symptom on the symptom list is then further described as a response to a list of one or more questions asked of the patient about the symptom. This triply-nested list structure is converted by suitable data structure transformations into a script that is stored. When a patient requires diagnosis, the script is played back as a sequence of questions. The responses of the patient are analyzed and converted into symptoms. The symptoms are accumulated into diseases. Finally the diseases are selected and reported as a diagnosis.

Description

Application is based on the computing machine medical diagnosis system of the processing of tabulation

The application is to be on July 11st, 1997 applying date, and application number is 97197872.7, the dividing an application of the Chinese invention patent application that denomination of invention is identical with the application.

The present invention relates to a kind of computing machine medical diagnosis system.More particularly, the objective of the invention is a kind of passing through to using the computer system of dynamic data structure in patient disease's the time-based diagnosis.

Current, the expense of health maintenance has spent national of the United States's total output value of suitable vast scale, and compares just in fast rise with other compositions of any other consumer price index.More seriously, because can not pay the medical services expense, many people can not get even the most basic medical treatment maintenance and information.

Many people are because reasons such as paying, time restriction or inconveniences, and incur loss through delay or fail to obtain medical services.If the public has comprehensive, the unrestricted and medical information that obtains easily so just can prevent numerous disease.Equally, just can prevent that for the early stage inspection of numerous disease and treatment numerous disease from developing into the stage of danger, for the health maintenance system of our country, these treatment costs have accounted for the pith of financial burden.Obviously, the U.S. is in the face of the significant problem relevant with health, and current solution is unsound.

Exploratory work at the health maintenance problem has in the past related to various forms of robotizations.The exploratory work of some of them has formed a kind of phone of answering medical care problem and has dialled in routine library.Other exploratory work has been directed to target when checking to patient and provides the area of computer aided support for the doctor.These methods relate to static process and algorithm.And desirable be a kind ofly can provide patient medical suggestion and diagnosis automated method fast, effectively and accurately.A kind of like this medical advice system must be the standard package that allows for novel medical care problem and inspection method expansion.

A kind of method of carrying out patient's inquiry comprises the medical diagnosis script.What need is a kind of form representative with script at the effective ways of each expert's aspect professional medical knowledge.Script must be used dynamic structure, is able to realize fast and effectively patient's diagnosis.

Processing based on tabulation is a kind of method that diagnoses the illness, and it forms one group of disease nested against one another, symptom and problem (DSQ) tabulation to disease, symptom and problem, handles these tabulations with this method, goes to produce the dialogue with patient.Produce one group of answer of determining for each problem of patient, and each answer the one group of problem determined that produces.Set up a dialogue like this, be able to draw symptom from patient.Symptom is processed goes to determine that with weighting disease is in boundary or outside the boundary.Disease in this class boundary limit is set up this diagnosis.A kind of disposal system based on tabulation is formed neat, structure list or matrix to medical knowledge, then a kind of specific algorithm of those list application is gone automatically to select next problem.Questions answer is drawn more problem, and finally draw a diagnosis.

In one embodiment of the present of invention, a kind of computer diagnosis method is arranged, it may further comprise the steps: one group of list of diseases is offered computer, and every kind of disease is relevant with one group of symptom tabulation, and every kind of symptom is relevant with basket tabulation; Repeatedly ask a question and draw answer, these are answered and constitute symptom, and the symptom of every kind of formation provides a weighted value to disease; And judge whether the weighted value for a kind of accumulation of disease meets or exceeds a threshold value, thereby draw a diagnosis.

This medical advice system also comprises the tabulation of the differential diagnosis among a kind of crowd of the patient place based on the geographic position, and when by based on the processor processing of tabulation the time, tabulation transfers a kind of differential diagnosis at concrete patient to.This system also comprises a tabulation about the disease occurrence frequency, utilizes the probability of disease among the on-site crowd of patient or incidence to come patient is assessed.The suggestion that this system also can provide special relevant with concrete condition selection that detects about the laboratory and image format to select at concrete patient is further made a definite diagnosis helping." reentry " function is exercised by this system, is accomplished laboratory test selection and image format and selects, and then the result is submitted to patient, patient's nurse and/or other mechanism that needs.This " reentry " function can be exercised by this system, is allowed patient to go to finish health check-up and operates (carry out in person or carry out through an assistant) and inquire about this system again, further to make a definite diagnosis.

Fig. 1 a is a block scheme that computing machine medical diagnosis of the present invention and the parts of the current most preferred embodiment of treatment suggestion (MDATA) system are shown;

Fig. 1 b is one the block scheme of parts of computer that explanation is presented at the user/patient of this MDATA system among Fig. 1 a is shown;

Fig. 2 is one the block scheme that batch processing, file and database in the system that is applied to Fig. 1 a are described is shown;

Fig. 3 a is the diagram of an off-line medical diagnosis script (MPS) generating routine, and this program produces a script file for the MDS database that has shown among Fig. 2;

Fig. 3 b is one and is used to illustrate the possible structural drawing for the DSQ tabulation of the script in two different time intervals;

Fig. 4 a is one and shows the process flow diagram of the definite disease part in " collect and organize medical knowledge " program in Fig. 3 a;

Fig. 4 b is one and shows the process flow diagram that obtains the disease knowledge part in " collect and organize medical knowledge " program in Fig. 3 a;

Fig. 5 is a process flow diagram that shows " script-editor " program in Fig. 3 a;

Fig. 6 a is the block scheme of the structure used of an operating process that is used for being presented at the diagnostic script engine;

Fig. 6 b is the block scheme of an output that is used for showing one group of structure and produces in the used input of the operating process of script engine with by the MDATA system;

Fig. 7 is a highest process flow diagram for the user program of the MDATA system service of Fig. 1;

Fig. 8 a is one and shows in Fig. 7, is used to finish the process flow diagram of " diagnostic script engine " program of on-line equiries operation;

Fig. 8 b is one and shows the process flow diagram of " classification suggestion " program in Fig. 8 a;

Fig. 9 be one in Fig. 8 a, shown be used for the part process flow diagram of " DSQ list script engine " program based on the processing of tabulation;

Figure 10 is a block scheme that is used for being presented at the part of the using tabulation that Fig. 8 a shown in the operating process of DSQ list script engine;

Figure 11 is another process flow diagram that shows " DSQ list script engine " program in Fig. 8 a;

Figure 12 is one and indicates the process flow diagram of " selection symptom " (the selection symptom that will be considered) program in Figure 11;

Figure 13 is one and indicates the process flow diagram of " handle and answer " (handling the answer that comes from the user) program in Figure 11;

Figure 14 is one and indicates the process flow diagram of " renewal list of diseases " (according to the symptom tabulation of having upgraded, upgrade the mark in the temporary transient tabulation of disease, and eliminate in the boundary or the outer disease of boundary) program in Figure 11; And

Figure 15 is a high level flow chart that is used to produce the another kind of embodiment of medical advice or diagnosis in the MDATA of Fig. 1 a system.

The specific descriptions of most preferred embodiment hereinafter are a specific concrete enforcement of the present invention.Yet the present invention is embodied in the various method in the scope that is defined by the claims and covers.In explanation, parts identical in are with reference to the accompanying drawings demarcated with identical Reference numeral from start to finish.

For convenience, the explanation of most preferred embodiment is made up of following major part: systematic review, medical diagnosis script, knowledge acquisition details, script produce details, script executing details and based on the advantage of the processing of tabulation.

One. systematic review

Medical diagnosis and treatment suggestion (MDATA) system are computer systems, and it carries out patient's inquiry of robotization for the purpose of setting up a medical diagnosis.In order to inquire about, MDATA uses a medical diagnosis script (MDS) database.Each MDS comprises a promising special medical condition service, inquires about a patient and output necessary data of diagnosis and order.Script supports that by other MDATA database this database comprises disease, symptom, treatment, medicament, expert ... promptly about medical diagnosis and needed all information of suggestion.Symptom can be taken as a history information, an information that obtains from physical examination, for example result who selects from the body signal, laboratory test result or the image format that get certainly usually.

Referring to Fig. 1 a, the block scheme of the current most preferred embodiment of MDATA system 100 will be described below.This MDATA system 100 comprises a network 102, and it can represent a Local Area Network, a wide area network (WAN), the Internet or other connection professional.

This MDATA program and database preferably are stored on one group of server 108, and it preferably interconnects with network 102 by a LAN (Local Area Network) 106 and a gateway 104.On the other hand, this MDATA program and database storing are on the individual server 110 of a use Network Interface Unit and software 112.This MDATA server 108/110 stores hereinafter tabulates the disease/symptom/problem (DSQ) of explanation.

Network 102 can be connected with user's computer 116, for example, and by using a modulator-demodular unit or passing through to use a network interface card connection.A user 114 on computer 116 can use a browser 120, uses keyboard and/or mouse device and a video display, and for example monitor 118, this MDATA program of remote access.On the other hand, when the MDATA program is carried out on computer 116 with the local pattern, just do not use browser 120.A video camera 122 can be connected with computer 116 selectively, to provide the video input, for example visible symptom.

Can use various miscellaneous equipments to be communicated with this MDATA server 108/110.If this server appliance has voice recognition or DTMF device, the user can be communicated with the MDATA program by using phone 124 so.A kind of telephong type embodiment has specific descriptions in No. the 08/176th, 041, applicant's's being entitled as of applying for simultaneously " Computerized medical diagnostic and treatment suggesting system for wearing " United States Patent (USP), the document is included in this patent as a reference.Other connection devices that are used to be communicated with MDATA server 108/110 comprise: one is equipped with the pocket pc of modulator-demodular unit or wave point, a cable interface equipment 128 that is connected with video display 130 or a satellite dish 132 that links to each other with televisor 136 with satellite receiver 134.Adopt additive method that user 114 is communicated with the MDATA server in addition.

Referring to Fig. 1 b, the structured flowchart of current optimal user/patient's computer has shown various possible and methods network interconnection.For " broadcast " script, use a specific program that is called script engine, this program is read the MDS file and is removed to finish query function with its code, and for example export a problem and give patient, and by answer of its input.Script also collect patient answer, diagnosis and new patient's medical records are more assessed, are made in its answer.Script engine preferably resides in user's the computer.Script engine can exist on hard disk drive or the CD-ROM, and in pack into primary memory or the cache memory so that carry out.

Currently preferably be presented among Fig. 1 b by the parts of the used computer 116 of the user 144 of the MDATA system 100 of robotization of the present invention.In addition, the miscellaneous equipment as the medical treatment inquiry as shown in Fig. 1 a, can be used to replace computer 116.

Computer 116 comprises the many parts that are contained in the casing 116.Telephone wire 156 is connected to computer 116 to public telephone network 158 via modulator-demodular unit 160.Telephone network 158 can link to each other with the network 102 that has been connected with MDATA system disk 108/110.In addition, the user can be connected with network 102 by using network interface unit 164.

In presents, " user " and " patient " these two speech can exchange use.Yet, should know that the user can be used as patient's procurator.If like this, the user then shows as patient's a assistant so.

Form hardware and system software, following two key concepts should be arranged in the brain: for the use of the portability and the industrial standard element of other operating system.Like this, the adaptability of this system will be stronger, and can make that the free market competition promotes to produce continuously, and reduce cost simultaneously.In the time may relating to special hardware and software, should know and in native system, to adopt various parts.

Computer 116 is the PC that the special Pentium microprocessor 170 of English is housed preferably.Other computer for example also can use the Macintosh machine of Apple, Amiga multimedia computer, the VAX computing machine of Digital Equipment Corporation, or the large scale computer of IBM Corporation.Modulator-demodular unit 160 or network interface unit 164 link to each other with industrial standard architectures (ISA) or Peripheral Component Interconnect (PCI) bus 162.Bus 162 is passed through controller circuitry (chip or integrated circuit board) microprocessor 170 and many peripheral component interconnect.

Have a plurality of external units that are attached thereto by adapter or controller on the computer bus 162.Video adapter board 172, SVGA or have higher resolution preferably is with video display 118 interconnection.A serial communication circuit 176 docks with a pointing device, as Genius mouse 178.In a further embodiment, a parallel communications circuit can replace circuit 176.A keyboard control circuit 180 docks with keyboard 182.A 500Mb or bigger hard disk drive 184 preferably are connected with bus 162 with an optional CD-ROM drive 186.These hard disk 184 stored data base files are supported file as patient file, other MDATA file and scale-of-two.This CD-ROM drive 186 is the stored data base file also, as being file and the scale-of-two support file that uses the service for patients of computer 116.

A primary memory 190 links to each other with microprocessor 170.In current most preferred embodiment, computer 116 is preferably in 192 times work of Windows 95 operating systems.Storer 190 operational diagnostics script engines 194 and disease/symptom/problem (DSQ) list script engine 196.Script engine software Borland Delphi Pascal 2.0 language compilation.

Referring to Fig. 2, below explanation is used for batch processing, file and the database of MDATA system 100.Except that below with the explanation script engine program, MDS database, image format database and laboratory test database, these programs, file and database are described in the United States Patent (USP) 08/176,041 that is entitled as " computing machine medical diagnosis and treatment suggesting system for wearing " of applicant's co-applications.

This MDATA system 100 uses a plurality of master routines and relevant database.One group of patient registry program is confirmed by a kind of in the following three kinds of methods of system's 100 usefulness whether a patient registered in the past in this system: the 1) identification card number (PIN) by prompting input patient; 2), determine an assistant who in this system, has registered in the past by an assistant's of prompting input identification card number (AIN); 3) determine a patient that the assistant is arranged, register in this system by prompting input patient's identification card number before him.Batch processing 212 is used for registering patient or assistant.If the user is patient, a patient registry program is used for registering new or primary patient by system.If the user is not patient, assistant's accreditation process is used for registering new or primary assistant by system.Then, if also not registration of patient, an auxiliary patient registry program is used for registering patient by system so.

Whenever user's typing or registration, system just provides option program.The master routine that relates to present embodiment is a diagnostic routine 220 of carrying out patient's diagnosis.The selection and the image format of a laboratory test of these appraisal procedure 220 visits are selected database, to recommend suitable test mode at this moment in time to this patient, and visit a treatment table 250, to obtain current treatment information to this disease specific or diagnosis.In another embodiment, increase other selection and go to visit other medical information program.

Relevant with these programs an is patient and assistant's registration database 240, consulting history data storehouse 242, patient's answers databases 244, medical history result database 246, patient medical data storehouse 248, uncertain database 252, patient medical history database 254, medical diagnosis script (MDS) database 256, an image format database 258 and a laboratory test database 260.

Two. the medical diagnosis script

Medical diagnosis script (MDS) be one for the diagnosis that produces one or more patient condition that work out and program user session.Exploitation MDS has following a plurality of step: the language representation that obtain medical diagnosis knowledge, can understand with patient it, its layout become a kind of useful sequence, it be compiled into the script that can play, detect it, its formation particular communication media, make it can be used for patient packing in the compilation of its other script of embedding people and supporting database and to it.

The meaning is to obtain medical knowledge and it is processed into the logical problem string " to write a script ", goes finally to produce the early stage step of diagnostic result.Obviously, only just can finish these steps, and the MDATA system a plurality of automated process have been developed and have gone to support them for diagnosis special disease experienced doctor.

The present invention preferably adopts a kind of being called the specific process of " based on the processing of tabulation ", and it is from the tabulation of disease, symptom and problem.Use one then based on the list script developing instrument, these tabulations are processed into the script that to play.With this script developing instrument, the program designer can write and compile the source script, and it is compiled the script file that can broadcast, the loopback script, and various scripts selections are set, remove to carry out, assess and finely tune this script.

Script based on tabulation is made of the text of a special format, and wherein the deviser provides the fundamental element of script with multiple tabular form.Highest tabulation is that script is with the list of diseases of considering.To every kind of disease, script is listed symptom and their weighted value.To every kind of symptom, the deviser provides the one group of tabulation that will draw the problem and their weighted value of this symptom.Be each problem, the deviser provides a plurality of texts, comprises a preamble text that is used to the problem of introducing.In all tabulations is after script is ready to, and next step is " editor " script, for example it convert to can playback or be the script file of the specific coding of patient institute " RUN ".For at the development phase Run Script, the script developing instrument is selected a suitable next disease and is fit to the next symptom of the sort of disease.It shows question text and waits for patient's answer.According to patient's answer, the script developing instrument upgrades the disease mark and continues next symptom.When some condition of setting as the deviser reached, for example drawing first disease in the boundary as a diagnosis, or all diseases were considered, and script just finishes.

In the development phase, the deviser of this script may be provided with multiple " selection ", and it will change script and select the method for next disease and next symptom and the working time of script.This selectivity characteristic makes the program designer who does the script test find best the setting.

So three Main Stage about script are: 1) knowledge acquisition, 2) script generation and 3) script executing.All three phases of the deviser of script use under development and debug script.In the working time of using MDATA system 100, patient uses the script executing stage.

The stage of script

1. knowledge acquisition

The knowledge acquisition stage comprises and must draw about diagnosing the knowledge of given disease from the Medical Technologist, and those knowledge are reduced the staff certain useful form in producing script.This stage is typically represented script that is used to diagnose a kind of disease (as malaria) of needs from the leader of script exploitation.And then the scope of decision script, investigation medical literature, inquiry author and other expert, the structure of criterion problem and answer is set up the order of problem, and, if used the instrument of automatic knowledge acquisition, this problem flow process of operation in debugging is installed so.It may be the source file that generates automatically that this stage ends at one group, it (at least) comprises necessary all information of the script that can correctly diagnose the illness of writing out, for example, when have for patient/no malaria makes feedback test when answering, write out malaria or do not had malaria.At this moment about the final form of script, it is with the workbench of operation even be used for the natural language of communicating by letter with patient etc. and all do not know,

2. script produces

Produce the stage at script, producing script is a less relatively diagnosis algorithm that is recorded in the software.In this stage, target is to allow this script embody the method that the deviser solves diagnose the illness (as malaria) or other medical care problem automatically.Script includes data and program is gone to produce first problem, this answer is weighed, gone to produce another problem with this answer, can finally tell its user's yes or no malaria as a result and corresponding confidence level like this up to script.

But note script be not an independent utility, can be by the program of real patient operation.Script is preferably only known single main disease, as malaria, and does not diagnose other medical care problem, the disease as gout or asthma.

This script will become the script database of nearly 40000 scripts, and significantly different form and form are arranged.Now script must be translated into suitable human language (German, Spanish), augment suitable fault processing function, form suitable program language (C++, Java, HTML), be formatted into be applicable to suitable target media (PC, the mechanical, electrical words of Mac, LAN (Local Area Network), wide area network, the Internet) and with back-up system (database, intermediate function, the patient record, communications records) connect.

Next step, various patients' in groups answer is promptly sent in the expansion test of script experience on testing table to script, with the diagnosis of having known that can adopt, go to prove that this script really produces suitable output.

At last, script will be loaded into a generation system.It can be saved as a large-scale script database, or is packaged into one group of script and is written to that CD-ROM goes up or deliver to hospital via the Internet.No matter which type of form of script storehouse, script is all with the application scenarios management software, and numbered and register.Last in this stage, the real problem that the last commencement of commercial operation medical diagnosis system of script is able to can be used for to real patient is done real diagnosis.

3. script operation

In the operation phase of script, script will really be carried out sooner or later.Certainly, with patient's dialogue not from diagnostic script about malaria.One is the medical diagnosis system of public visit, before being used for any medical diagnosis, obviously has many management roles to finish.At first, we do not wish to have delayed patient's the state of an illness when password of application and social security number.So system enters the people of MDATA system probably for all, at first move first-aid room (ER) subsystem.This first-aid room subsystem is made up of some scripts, goes definite patient whether the compromised situation of life to be arranged and need immediately " first relief " to treat or suggestion." being medical first aid? ", " patient has and is breathing? ", " patient is bleeding? " it is the problem that several first-aid room system will put question to.

After this emergency situations was got rid of by system, system entered at leisure and determines patient and decision principal disease.System call script distance subsystem then, its work are decision patient's general considerations scopes.According to this information, next step selects a kind of order of highest script of the principal disease that is applicable to patient script distance system.For example, for fever, after significantly the script of appendicitis, enteric infection, food poisoning indicated " not having diagnosis ", this path script can be ordered for a trial at last and be malaria.Now, broadcast in this document the sample malaria script that we have developed at last.

Script is not the program that oneself can move.Script is a serial data, and it is by " script engine " operation, and removal search is used to ask the script of patient's next problem, and the problem format so that transmit (giving screen, phone or IP address).Patient's answer is also obtained by script engine, is used for script after formatted, and as from script, selecting next problem.This interaction of the script engine of script and it can be considered to patient's medical records, the information that up to the present provides at this session and even some determine next problem intermediate function.Last at script, this program is got back to control on the script of tropical disease path, and show " this patient's answer show the weight of malaria has 1350 exceed 1000 " or " this patient's answer only has 420 to exceed 1000, so I get rid of the possibility of malaria " effectively.This present primary decision of path script that is called the symptom script is to enter another diagnostic script, and perhaps some answers of user, for example " patient's answer shows only 275/1000 the tropical disease that has " are given in decision.

The characteristic of script

Time-based diagnostic script

The notion of time-based diagnostic script has expanded to time range to the DSQ diagnostic script.Except a diagnostic script is only arranged, the deviser of script can provide a plurality of scripts now, and for example per hour a script enters the disease program.Judge according to the best of deviser, produce a script through after a while from picking up counting of symptom.For example produce a miocardial infarction script with one hour or less time as a time period, and malaria is not like this.When operation, diagnostic system adopts the immediate diagnostic script of situation with patient.The symptom script has implicit symptom, and it adds extra weighted value for the disease that matches with the pattern of foretelling.

When symptom begins the system interrogation patient, partly according to this information, selects suitable script from time-based script group.In case script is chosen, just carry out this script.That is, each script of time-based script group can have some different symptoms and weighted value, makes the deviser set up time-based symptom, for those temporal modes and patient's the disease that matches is added extra weighted value.When script engine moved, these weighted values were by its automatic increase.Notice that these time-based symptoms will be implicit symptoms, will describe in detail below.

The deviser of each algorithm must (for example) each hour along with disease progression control, distribute or calculate these problems and suitable value.Then when patient inquires about this system, when first problem that will ask is " (or how long before) your symptom begins? ", the most close script of time that will play disappearance when that patient will begin with symptom then interrelates.

Implicit symptom

Notice that " symptom " is meant any data item about patient as can be known, comprises name, age, sex etc. directly or indirectly.Implicit symptom is one has or does not have a symptom that constitutes according to one or more other symptoms.The notion of implicit symptom allows the deviser of script to remove to tell script engine, and any symptom that provides (or symptom group) contains or do not contain one or more other symptoms.This allows the deviser that the relation of objective world is embodied in based on list script, and it allows the LB engine carry out reasoning from logic then, is able to remove unnecessary problem from tabulation, and makes script more accurate.

As a tangible example, need not ask a problem relevant to the patient male sex with female reproductive system.A human doctor knows this point certainly, but must tell script engine.The deviser of script has made the symptom tabulation of a following form simply:

If symptom A, symptom B so.

For example:

" patient is the male sex " implying " patient is not the women " and

" patient had appendectomy " implying " patient has not had appendix ".

The applied logic operational symbol as AND, OR or NOT, just can be set up with quite few problem and draw very complicated symptom relationship.Implicit symptom is listed in the script of source, resembles the table of " if A, B so " type statement.No matter when script engine receives a new symptom there from patient, and it also checks implicit symptom table, and whether go to find out has any other symptom by implicit.

Syndrome

Syndrome is meant such symptom: in the patient of any supposition, when certain group symptom when taking place simultaneously, they have a specific diagnostic significance.A DSQ based on tabulation source script in, each symptom has a special weighted value for the disease in the diagnosis, but certain group symptom can provide extra weighted value to diagnosis when taking place.For example, typically diagnose malaria be from be afraid of cold, (when night sweat shows effect owing to malaria pathogen enter blood circulation cause) that begin of fever and night sweat.Be afraid of cold, fever and night sweat independent to be not enough to diagnose patient be malaria, but all three symptoms can guide an inquiry about malaria when taking place surely.The notion of syndrome is supported the guiding of this inherence, and with such statement for example: " being afraid of cold " AND " has fever ", and AND " has night sweat " means " may be malaria ".Syndrome also plays an important role when determining a complication, and for example, because should often generation simultaneously of specific symptom set, for the public, they have its oneself title, as AIDS (AIDS).The deviser of script can go complication certainly with syndrome, and this also is very important for him.

Three. the knowledge acquisition details

For a script, the initiating task of knowledge acquisition is to determine to be drawn disease in the script, arrange order and distribute to the Medical Technologist and go various piece for their disease of having arranged exploitation script for every kind of disease.Each Medical Technologist produces the required suitable tabulation of various diseases then.This can be summarized as following content:

● determine the scope of the disease that relates to;

● list disease and their symptom;

● arrange grade, order and weighted value for disease and symptom;

● design is used proper words and the problem of weighting, it will draw symptom;

● disease, symptom and problem list format;

● predict these tabulations, use the testing tool of developing for this purpose specially; With

● tabulation is write as text, with any word processor that can handle ASCII character.

Should start from one group of interlock tabulation based on the disposal route of tabulation, it obtains the key element that is used to diagnose a special health problem.In this stage, the Medical Technologist notes their diagnostic techniques and skill with the form of a plurality of tabulations.In order to do this work, the expert preferably can use the Word that can buy on the market, and it can produce an ASCII character output file.

This ASCII character tabulation as script comprises three types tabulation, reduces as follows:

● a kind of list of diseases, it determines all diseases that script will be considered, and according to the order that is considered for diagnosis they is classified;

● a kind of symptom tabulation that is used for every kind of disease, it is determined symptom and gives each symptom fixed weighted value, goes to determine that it is to providing of diagnosing the illness;

● a kind of problem list that is used for every kind of symptom, it determines the problem of the weighting of the one or more symptoms that will draw patient.

Be the purpose of automatic medical diagnosis, medical diagnostic data is combined into range, and classification is to have symptom and symptom to draw from patient by puing question to according to disease, such universal and carrying out.

A kind of " disease " is meant a kind of health status that needs diagnosis and treatment or maintenance, for example: slight illness, worry, worries, situation, state, problem, obstacle, imbalance etc.For the patient of medical diagnosis on disease with a supposition, the MDATA system is since one group of possible list of diseases, shows disease according to patient's answer, and summarizes one group of diagnostic listing from this disease.

A kind of " symptom " is any information about patient that the MDATA system has.These comprise:

● patient's identity (as name, address, HMO, age, sex);

● patient's medical history (as former slight illness, father and mother's health status, the situation of abroad travelling recently);

● with the situation (as patient's disease and development history) that advances into MDATA;

● body illness (as the illness of major organs) and in person or the result of auxiliary physical examination exercise;

● chemical examination and test result;

● symptom, phenomenon, outward appearance, shape etc.

For every kind of disease is prepared one group of symptom tabulation.Be the fixed weighted value of every kind of symptom after given symptom, on behalf of patient, it the possible degree of this symptom is arranged.In order to simplify calculating, the MDATA system uses a threshold value 1000 to go the disease of judging that this is diagnosed, though also can use other threshold value.System also goes formally to announce not this disease of patient with a threshold value.In the boundary threshold value and out-of-bounds threshold value can change according to the sensitive factor group.This just allows the threshold level that customizes, for example is the individual patients customization.To go through this sensitive factor group below.

In fact, after given symptom, weighted value is the yardstick that the doctor in the diagnosis thinks the wish of this disease.After given symptom, weighted value also is used as the conditional probability that patient has disease.If talk about easily, this also can carry out the Bayesian probability analysis to this symptom.

A kind of symptom is drawn by one or more problems, often alternately occurs with information of how to answer a question and explanation.Be called one " flow process " because of drawing the required node that is provided with of symptom,, often be drawn on the little process flow diagram, illustrate with patient's dialogue how to carry out with it because it typically relates to the problem of a component distributary journey.

For MDATA system that the diagnostic data of being developed by the Medical Technologist is packed into, data must be organized and format.Be this purpose, use a kind of text, developed a kind of text file format.Though preferably use the ascii character sign indicating number, also can use any well-defined text character sign indicating number, as, the binary-decimal permutation code (EBCDIC) of expansion.

A script comprises a plurality of or data set as follows:

A. disease will be according to the symptom of weighting and is diagnosed;

B. symptom will be drawn by flow process or be comprised by other symptom;

C. implicit logicality ground connects symptom;

D. flow process is made of the path by node;

E. the node of path access problem;

F. patient is informed and/or advised to text;

G. problem is that patient is inquired in an answer;

H. button sends a concrete answer from patient.

These are parts of following paragraph about script " source " or text.

Title division

Comprise data in the title division, as the form of script and the symptom group of the principal disease that constitutes patient about whole script.

The disease part

Disease is partly listed the disease of being diagnosed by this script, their symptom and for the weighted value of the symptom of a diagnosis.When script moved in the script development phase, the script developing instrument selected in the disease to be used as next disease, selected one in the symptom of that disease to be used as next symptom then.Which disease and symptom are chosen to the next one, and this will select to decide according to the operation that the person of being designed selectes.Default order is the tab sequential of the symptom of disease and they in this paragraph.

Disease _ title

Disease name is to be the fixed unique symbol of each disease, as this disease of identification.It only is used for internal system, and patient will can't see forever.

The ICD-9_ sign indicating number

A kind of special coding is used to discern disease by the Medical Technologist.

Form _ title

The form title of disease.It is can increase at form in the future because be used for the common name and the abbreviation of disease here that this " form " title is used in.

Symptom _ title

The symptom title is the part of " finger mark " of diagnosis diagram or disease.The definition of symptom will describe in detail in the symptom part.In the DSQ list content, one " symptom " is a special and detailed fact that is assumed to be, admits, draws, implies.The deviser need not define any data item for symptom.If symptom is useful for the deviser, it can comprise the fact of non-medical treatment so, as sequence number of name, title and patient etc.This is intended that and allows the deviser freely go to represent his/her medical experience, by definition cardinal symptom and with any suitable method they is combined.

In order to design a symptom, the deviser can imagine one group of problem of weighting, and it will be assert uniquely or negate this symptom.If this is out of question, the deviser defines this symptom (in the symptom part) according to its problem and answer so.If symptom is too complicated, the deviser can be merotomized symptom, and each part is used as a symptom, and proposes the problem about this part.The deviser can allow patient set up each part respectively, uses the inference mechanism in the reasoning paragraph to go to make up cardinal symptom then.

Symptom _ weighted value

This symptom is added to the quantity on the gross score of disease.Technically, this quantity can be any value of from-10000 to+10000, and in fact it may be a little positive integer.After having write, script engine is used as a kind of method to disease " marking " to weighted value.When a symptom was present among the patient, script engine was added to the weighted value of this symptom on the gross score of this disease.When the mark of this disease reached 1000, script engine put this disease in the circle under.

The arithmetic addition of simple weighted value does not represent that this shows the concrete grammar that symptom exists disease.A kind of method of giving the deviser is to estimate one group of weighted value earlier, carries out this script, and how the mark of observing disease changes " balance again " these symptoms of going back then along with each problem and answer.

A kind of " syndrome " technology is of great use for the deviser in the exploitation weighted strategy.If two symptom A and B are arranged, if there are this two symptoms simultaneously in a patient, more weighted value when giving so than individualism, the 3rd an artificial symptom C can be defined as A and B self-evidently then, and gives the extra weighted value of this disease.Symptom C does not have relevant problem; This is an inside " parasitism " symptom, and it is only according to the existence of other symptom or do not exist, and is used for increasing or reducing weighted value.

The symptom part

All symptoms that the other parts of script were mentioned are partly listed and illustrated to symptom.Be each symptom, this paragraph is identified for drawing the problem flow process of symptom.

Symptom _ title

The symptom title is to be every kind of unique symbol that symptom is fixed, as this symptom in the other parts that are identified in this script.It only is used for internal system, and patient will can't see forever.

Flow process _ title

" flow process " this speech is the problem set that is used for describing special weighting, with a kind of problem that can be painted as the certain order of process flow diagram.Such flow process is represented single problem set.Because flow process can be drawn one in a plurality of symptoms, a plurality of symptoms will typically be determined same with the problem flow process that is used.Some symptom (as the principal disease symptom) does not have relevant problem flow process.

Implicit paragraph

Implicit paragraph is listed the logical deduction between symptom, makes script engine know those implicit other symptoms.Every row of this paragraph is determined one or more symptoms, their implicit together another symptoms.The logical formula that is every this form of behavior provides parameter:

If symptom A and symptom B and symptom C so then have D.

Symptom is implicit can form chain, makes the implicit symptom of a quilt individually or with other symptom to imply another symptom in combination.

A purposes of this paragraph is to set up " complication " symptom, makes patient's a specific symptom group will confirm an independent set symptom automatically.This combination symptom also can be used for increasing (or minimizing) extra weighted value, if a specific symptom group exists, for example, will allow the simultaneous a plurality of symptoms of patient " concurrent ".

The flow process part

Flow process is partly listed all flow processs in the script, and determines the problem that can draw by flow process and the order of symptom." flow process " is the abbreviation of " problem process flow diagram ".It can be imagined as a complicated problems, and it will set up in a plurality of symptoms one of them.Be familiar with reader based on branch's script and will think that this flow process can play a part to hold or introduces one wholely based on branch's script, is returned in a plurality of answer/codes.

Generalized case is need propose a plurality of problems to patient to go to draw a specific symptom from patient.For example, need some underlying issues (" you are do you do weed once? ") make place mat, be that (you have inhaled how many years altogether to very professional problem subsequently?) go accurately to determine patient's symptom.One whole flow process may accommodate 20 problems about smoking, and may draw in a plurality of symptoms one, for example: from non-smoking; Smoking once in a while; Inhaled 20 years and also in smoking; With inhaled 10 years cigarettes, do not inhale before 10 years.

According to flow process from a node to another node, each node in the process flow diagram has been compiled number.These paths are used for each node will carry out action.

The problem part

Problem is partly determined the details of problem cited below.Details comprise: preamble, practical problems, the button (on the phone panel) that can be pressed by patient and (being used for visual interface) are as the button sign of each answer.

Preamble _ text

Preamble be before problem proposes, read to or be shown to patient's text.It can problem in front after, continue to introduce a new theme, determine some projects, patient is provided the information about why asking a question and how to answer a question.Here only provide the title of text; This actual text partly provides at text fragment.Be not the preamble of a problem, this is denoted as 0 so.

Problem _ text

Question text is actual problem.Preamble may be 10 row or 100 multirows, and problem is typically very short, or even a point, and by point or by these keys, draws a very concrete problem.

Effective key

One group of effective key tell the diagnostic machine patient can by or put that key.

Key 1... key N

These are signs of key, only use in the visual version of display of script engine.They tell script engine to indicate how for each button, for example are, non-and uncertain.

Text fragment

Text fragment is listed by the actual text of all text items of the location of the title in other paragraph, for example preamble, key sign and question text.By the unique title of given each text, and list text in the text fragment, the deviser can be used in same text a plurality of places.

For the text of service for patients is placed on a position, also simplified the automatic processing of script all, for example record is used for the text of telephone network or the text that format is used for screen display.By its text fragment being replaced into the relevant text of other Languages, a script can be translated into foreign language.

Referring to Fig. 3 a, explanation is used to produce the off-line procedure 280 of a DSQ script below.Since an operation 284, medical knowledge is collected and is organized into listing file.Data as listing file are collected for one or more medical devisers 282.Program 284 has two parts.First is finished by the yardman or the management deviser of script typically, go to determine disease, and second portion is used for collecting the disease knowledge of script about each disease.The part that is used to collect disease knowledge is typically finished by many Medical Technologist in their field separately.The part of definite disease of program 482 will be described further in conjunction with Fig. 4 a, and the collection disease knowledge of program 284 part will be described further in conjunction with Fig. 4 b.The output of program 284 is e-texts, as an ASCII character file.This e-text exists with the form of DSQ tabulation, as disease, symptom and problem list 286.Comprise a demonstration script that is used for malaria in the appendix.This script is the representative of a DSQ tabulation.

Image demonstration about the time-based DSQ of a script shows in Fig. 3 b.The demonstration script 320 of corresponding time T 1 and the script 322 of corresponding time T 2 have been shown.In two scripts each all comprises a list of diseases 324,326 and problem lists 328 of a symptom tabulation.This figure wants to illustrate the level of disease, symptom and problem list, and only is exemplary.Attention: a disease can be pointed to, the symptom that limits by other symptom, and symptom can be pointed to the problem that is limited by other symptom.Like this, symptom can be reused by a plurality of medical devisers with the problem relevant with them.

Get back to Fig. 3 a now, program moves on to state 290, and it is tabulated DSQ and becomes the form of e-text, and by using the script data developing instrument to handle them.Script-editor 292 carefully uses the script data developing instrument to remove to produce a MDS file.Program 280 can use script data developing instrument and script-editor to remove to produce a final MDS file repeatedly.At state 294, this MDS file is write as a MDS database 300 by a MDS database management apparatus 298.This MDS file 296 is binary format preferably.An example of this MDS file be presented at Fig. 3 a 296 in, this MDS preferably includes the main list of diseases part of a preamble data division,, a main flow part and a main text tabulation part.In another embodiment, medical deviser can be with medical design language compile script, or node and branch, writes as showing at state 302.Other wscript.exe that can comprise editing machine shows in state 304, removes to produce a MDS 296.

Referring now to Fig. 4 a,, explanation collected and organizes definite disease program 350 in the medical knowledge program 284.Program 350 typically will be finished by the script yardman, though use other Medical Technologist of MDATA system also can finish these tasks.Program 350 preferably is not by computing machine but is finished by the script yardman that the yardman can use a computer to assist and finish following step.Since an initial state 352, program 350 moves on to state 354, and the principal disease relevant with current script determined in the there.When the subject matter that they will inquire about was described, principal disease comprised that patient offers the symptom of system at first.Proceed to state 356, the script yardman determine one group will be by the list of diseases of current script diagnosis.These diseases will provide a diagnosis about principal disease.Be included in disease IC (ICD-9) sign indicating number of the title of the disease in the tabulation, one section descriptor and this disease.Advance to state 358, according in common population, in certain certain the national or national area as the patient place, the possibility of generation is the disease classification.Move on to state 360, the yardman of script arranges priority according to the order of severity of emergency treatment and/or disease.According to fixed priority, script engine can directly go to check the first kind of disease that is arranged first aid and serious indication.Proceed to state 362, the script yardman divides the disease that is used for current script or distribute to one or more Medical Technologist and further develop then.Use a computer network, as the Internet and DSQ list database, a plurality of scripts can be developed simultaneously.The deviser of disease can be by making the problem and the explanation that can both obtain via database and network for all other devisers, and can work concurrently.This possibility allows exploitation script fast.Program 350 finishes at done state 364.

Referring now to Fig. 4 b,, explanation collected and organize and obtain disease knowledge part 380 in the medical knowledge program 284.Program 380 typically neither be by computing machine but is finished by the Medical Technologist or the expert who collects practically about a kind of special disease knowledge that uses a computer.Below step finish by the disease expert, resemble and arrange by the script yardman the state 362 in Fig. 4 a.

From initial state 382, program 380 moves on to decision state 384, and the there Medical Technologist determines whether script preferably collects time-based script of imaging.Many scripts in the promptly corresponding orderly time interval form a script series, and it is used to follow the tracks of disease over time.If be determined is a time-based script, and program 380 moves on to state 386 so, and the time interval that the script in script series is asked is determined in the there.For example, the script deviser can determine the script of generation in per two hours in 48 hours.When finishing the time interval of decision script series, if or script preferably be shown as single script, program 380 proceeds to state 388 so, the there Medical Technologist distributes to his or her disease for each and determines the mark of threshold value in the boundary and the mark of threshold value out-of-bounds.Move on to state 390, the Medical Technologist determines one group of relevant symptom for every kind of disease of distributing to them.The symptom tabulation comprises symptom title, one section descriptor and at least one weighted value, as below illustrating.Proceed to state 392, the Medical Technologist determines any relevant back response relation and the symptom of being determined by these relations.This back response relation can comprise simultaneously or concurrency relation, and taking place in the time of two or more symptoms of there to provide the more weighted value of summation than the weighted value of the symptom that takes place respectively for diagnosing a kind of disease.A kind of ordinal relation is meant that a symptom follows another symptom to take place, and this can provide the more weighted value of summation than the weighted value of the symptom that takes place respectively for diagnosing a kind of disease.A kind of variation of ordinal relation is meant the variation of the order of the beginning of disease or end, provides and occurs different weighted values separately.Implicit relation is meant that the appearance of a symptom means the appearance of another symptom.Symptom deviser can also and further make the symptom foundation relation in time that back response is handled for fixed symptom.This back response relation also relates to classification processing, PQRST matrix analysis or the severity of symptom analysis of symptom.This PQRST matrix is the matrix of N dimension, comes a minute dimension with the different attribute or the aspect of disease.For example this PQRST matrix can have 22 dimensions.

Proceed to state 394, the Medical Technologist is that every kind of disease symptoms is set a weighted value.Because symptom has a relevant scope, as the order of severity of the symptom of the order of severity of disease or other type, so the Medical Technologist can set the scope of the weighted value relevant with the order of severity of symptom.Weighted value both can be on the occasion of also can being negative value, and it determines the mark of a plus or minus.Move on to state 396, the Medical Technologist goes to draw or determines this symptom for each symptom is provided with basket node.Some symptoms can be determined by a problem, and most symptom needs many problems go to draw symptom.Because symptom needs many problems, so may answer the setting weighted value for problem at state 397.Like this, the symptom of the type has the relevant weighted value of certain limit.Advance to state 398, the Medical Technologist is that the problem node of each problem flow process is write the text at the problem node, must think that patient provides the problem of introduction or explanation, announcement, suggestion and reality.These introductions can determine requested content (an answer group) value scope or format other method of the answer of this expectation.These introductions and the explanation will help patient know problem be about and so on, propose this problem why and be set the stage for possible answer.

The deviser will form basket flow process for each symptom, and it is used to draw symptom.The flow process that this deviser uses also can be another doctor's a problem flow process.For example, the hypothesis symptom is a depression, and a doctor may ask: " your depression? "This can be called " depression _ problem _ 1 ".This deviser of hypothesis dislikes it, and it too succinctly can not really obtain the thing of wanting.So this deviser further seeks in issue database.The deviser may find and see " depression _ problem _ flow process _ 2 ".This problem flow process exquisiteness many.In this flow process, for answer a question " your depression? ", this doctor has designed one 10 problem list.These subproblems may even have other problem in database.In this problem flow process, patient is asked 10 problems.Each problem is weighted respectively, and after answering all problems, score is their summation, and if mark reached the threshold value that the deviser by problem determines, this doctor will think that this patient has depression so.

In another example, deviser of hypothesis wants to ask one about moving seasick problem.This deviser checks problem base.This deviser can find 50 about seasick different problem.Problem is asked " you are seasick? "This problem is not accepted by this migration deviser.Another deviser has a problem flow process, comprises 10 problems of weighting.If their mark has reached the predetermined threshold value of deviser, that doctor thinks that his patient is seasick so.Though this migration deviser preferr it that way almost wants to change the weighted value of one of them problem in the subproblem of weighting.In this situation, this migration deviser saves as " seasick _ problem _ n+1 " to this new problem of having revised weighted value.Use the seasick problem of this redaction or other version now as this migration deviser, it certainly will be by weighting respectively in determining different diseases.

If the different weights value of problem is unallowed in the problem flow process, so all problems will be confirmed as identical weighted value.But when a disease deviser wants to look at whether the hypothesis belly is injured, can ask that patient does a series of movements, for example " please cough.Your belly pain? "If patient replies "Yes", bitterly whether this disease deviser ask patient to push his belly and to ask then.This disease deviser puts question to usually or the request patient finishes " symptom " that many such actions remove to set up this abdominal injury.Yet these problems are not no less importants all in determine abdominal injury.If when patient in belly when pain sensation of pushing him, this action than cough is more meaningful.

Whether the problem node is that time-based script needs another time period in decision state 400 Medical Technologist decision after state 398 is finished.If if do not need another time period or current script not to be based on the script of time, program 380 finishes in return state 402 so.Yet if need another time period in a time-based script, program 380 gets back to 388 so, for another time period in the script series is reruned step 388 up to 400.

Four. script produces details

Inside in the MDATA system resembles based on the medical diagnostic data of tabulation and to store the script.These files are the diagnose interfaces between human doctor and the patient that just inquired about.In when operation, MDS file is " RUN " by driving script engine, and it is one and the MDS file is installed and according to the ordinary procedure of layout data hereof and order Run Script.The data of diagnosis store with the form of disease, symptom, problem and text node tabulation.

The content of the content reflection ASCII character listing file of a list-oriented MDS file.The key distinction between them is: the text data resemble and store the paragraph of line of text of feature string, and the MDS file data is to form with the tabulation of bigit.Second difference is: the MDS file data is stored as reference mutually, is supported these data of on-line access.

This MDS file preferably is formatted into the matrix of 32 very big bigits.This jumbo matrix is allocated to the data block of indefinite length then.Because the position of data block is its number in the file, ins succession and use the data item of another data block so it can resemble a data block.Say with regard to physical significance to how program language or operating system are independently to these data blocks, it can be sent in the hard disc of computer of any file that can store 32 figure places.Say with regard to logicality, these data blocks can with arbitrary method nested be connected, go to form the data structure tabulation, storehouse, arrangement, tree structure and the network that resembles articulating.The MDS file is formatted into a plurality of paragraph pieces that are called " master list ", and is as described below:

● title data,

● main list of diseases,

● the main flow tabulation,

● the primal problem tabulation,

● the tabulation of primary symptom shape,

● the main text tabulation.

In order to prepare a MDS file, the ASCII character listing file is read out and is converted into a MDS file by script-editor.This program comprises reads the ASCII character text line by line, and editor goes quick search about the suitable paragraph of MDS output file and the tabulation that produces mutual reference.Because some symbol just can use before being defined, dialogue program must pass through this file twice.By in the process, all provisional capitals are read into, and be converted into the MDS blocks of files, and their symbol are existed in the table for the first time.By in the process, symbol is replaced by the piece title of their reality for the first time.Certainly, can use other edit methods.

After script was edited, the deviser of this script detected this script, went to determine whether it works as plan.If not, so for example the deviser of script can adjust symptom/problem weighted value, go over word and the paragraph that is used for the problem node and revise any logic and the medical treatment mistake.The script deviser will update and move this script then, move as plan up to it.

Referring to Fig. 5, the editing machine 292 of script will be described now.With the e-text form, as ASCII character, the DSQ of existence tabulation is collected by using the script data developing instrument, is handled by script-editor 292 then.From initial state 420, script-editor is that integrality, consistance and the unitarity of script handled this source script.In this state, discern grammar mistake.After having corrected all problems, editing machine proceeds to state 424, and this script is converted to the binary file layout that is stored from source format.Proceed to state 426, script-editor 292 these scripts of expansion are used for being accessed in various MDATA databases and MDATA fundamental mechanism or the back-up system that Fig. 2 has shown.Script-editor is finished in return state 428.

Five. script operation details

General introduction

When patient is diagnosis when entering MDATA system 100, the primary data of this system management and contact patients, discerns this patient, decision patient and need what service, the MDS file and the script engine that brings into operation that selection is correct.Pack into this MDS file and beginning of script engine is obeyed its a then order of one of having encoded.The result who obeys the order encoded is the one query with patient.Last in inquiry, script command script engine are carried out suitable terminal behavior (more new database, close file, write down this section dialogue) and finally the control of computing machine are returned MDATA system 100.

To illustrate that below using the MDS file to drive script engine goes to carry out on-line equiries.For access data library file, output information are given the fundamental operation system execution of this required support operation of patient, the answer of importing patient and printed report result by the script engine program run.

List-oriented MDS file of mode producing working time based on the disposal route of tabulating.This means in each step and must search for the action that disease, symptom and problem list remove to determine next problem and script.Script engine uses based on the method for tabulation will do more work than using based on the method for branch.

This MDS file must be a medical treatment encyclopedia about human diseases, the single problem of its certain aspect from senior list of diseases to a kind of symptom that is used to draw certain disease with the sequential storage from senior to rudimentary.In order to move a kind of like this data structure, need " conversion " this structure as a script, for example as the string of problem orderly to patient.In order to do this work in the pattern in working time, script engine is at first searched for the main list of diseases of MDS file and is gone to select next with the disease that is considered.The symptom of selecting the next one to put question to is gone in the tabulation of the symptom of the selected disease of script engine search then.The script engine search goes to select next problem of being putd question to for the problem of selected symptom setting then.Script engine is asked a question, is obtained to answer, upgrade the tabulation of various weightings and repeat this program up to obtaining a diagnosis or having looked into all diseases to patient.All effects are the diagnosis dialogues that produces between script and the patient, and conclusion is a diagnosis.

When script moved, the symptom group that script is kept patient resembled a temporary transient dynamic listing, was called " temporarily " tabulation.Each new symptom is recorded in this group, and as the tabulation of upgrading the disease of considering.The profile of a health status is set up in patient's answer, is used as next disease and symptom and the problem selected.This profile has many purposes:

● it goes to help to select next disease as upgrading all diseases that just are being considered;

● it can compare as the static state of situation;

● it allows the MDATA system dynamically to select the problem string according to the health status of especial patient;

● it allows script of MDATA system break, continues it later on again, reloads by storing this profile and time afterwards that it removes to continue this script.

When script engine began, it was supplied to an online patient and a script (for example MDS script).This script engine is opened the tabulation that the MDS file goes to set up disease, symptom and problem that access encoded.The record that it also opens patient goes to obtain patient's medical history and the result who talks with in the past, if any.From here on, the MDS file drives this and inquires next query steps under the order of script engine.Last what inquire about, the script command script engine goes to carry out suitable terminal behavior (more this section dialogue of new database, close file and record) and finally the control of computing machine is returned the MDATA system.

To the interested aspect of explanation based on the processing of tabulating is the algorithm that is used to put question to patient and sets up the symptom of a diagnosis of one group of correspondence.This algorithm by one be used to analyze and more the major cycle of new patient's symptom group constitute, just stop this circulation up to reaching every kind of condition.Major cycle comprises following general step:

● analyze patient's symptom group,

● select the next one with the disease that is considered,

● select the next one with the symptom that is considered,

● select next problem with proposition,

● offer patient's problem and handle answer,

● upgrade the symptom group according to answering,

● carry out about the back response of symptom group and handle,

● cycle analysis patient's symptom group.

Continue this major cycle up to because some tenth skill as forming a diagnosis, providing treatment and advise or patient is given to another script, just stops this script.

Referring now to Fig. 6 a,, explanation is used to operate the common formation of the MDATA system of diagnostic script engine 190.Diagnostic script engine 190 dock with MDATA back-up system 440 obtain with many databases 442 of MDATA system and have with medical profession in the ability that communicates of many entities.MDATA back-up system 440 is included in the program that shows among Fig. 2, comprises accreditation process 210, registration procedure 212 and diagnostic routine 220.Be included in also being useful in the MDATA back-up system 440 and carry out the program of input and output, give and from doctor 444, patient 114 and health agency 446 as a health maintenance mechanism (HMO) input and output.MDATA back-up system 440 uses communication network 102, and the front shows in Fig. 1 a and 1b.Data presented storehouse 442 comprises front data presented storehouse and other database in Fig. 2 in Fig. 6 a, for example about the database of the frequency of anatomical, a kind of admin table and the geographic distribution that disease occurs human diseases, medicament and medicament relation, human.This admin table is a kind of table about management and legal " rule ", and it allows system know has to show for patient for how much information.

Referring now to Fig. 6 b,, will structure and the input and output of using be described in the operating process of diagnostic script engine.According to from user 460 input, from record and other obtainable information from MDATA central database 442 of the database 254 of patient medical history, MDS 296 is selected from MDS database 300.On the other hand, if diagnostic script engine 190 moves on patient's PC, so local user's data storer 184 can be switched on, and replaces existing the MDATA database in the central database.Yet in fact most cases is that patient's medical history is existed in the central database, and reason is as follows: the security of record; Health maintenance supplier under the sun as needing analysis, comparative diagnoses, obtains any new methods of treatment, and make and can be apprised of very soon by the patient of this system, or the like.

MDS 296 is made into can be for diagnostic script engine 190 obtains, and it carries out patient's inquiry.Diagnostic script engine 190 can be write the information that receives to historical central database 254 of patient medical and local user's data storer 184 in patient's inquiry.At the conclusion part of current script, if or other script the operation, can produce medical diagnosis so or advise 462.This diagnosis or the suggestion preferably report to doctor 464, export to user 466 and have the MDATA central database or local user's data storer 184 in.Can produce other report 468 if desired.As below illustrating, such situation is arranged, diagnosis needn't directly report to the user, and in order further to report to the user later on, can send to the doctor earlier.

Referring to Fig. 7, below explanation is used for the user with the general five-star program 480 of the dialogue of MDATA system 100.Program 480 begins and moves on to state 482 to go to discern a kind of first aid situation from initial state 481.The problem of one group of initial (the hard disk coding) screen display is used to discern this first aid situation.If determined a kind of first aid situation, offer the suitable suggestion of user so, as call out 911.State 482 and state subsequently 484,486 and 488 explain in the U.S. Patent number 08/176041 that is entitled as " computing machine medical diagnosis and treatment suggesting system for wearing " of applicant's co-applications substantially.If program 480 determines not have the first aid situation, this program continues also to confirm safely this user at state 484 so.As what illustrate in the patent of applicant's co-applications, the user can be a patient or an assistant who serves this patient.The personal identification method of password, ID (identity number) card No., SoundRec or other type may be used.If patient's typing correctly this system, program 480 proceeds to the management role that state 486 removes to carry out any needs so.Proceed to state 488, program 480 access MDATA medical data base (Fig. 2) and system file and softwares.Proceed to state 490, carry out on-line equiries with the user.These on-line equiries are preferably carried out by diagnostic script engine program 490.Yet, can use other method of carrying out on-line equiries, as move a program or carry out a script.This user program 480 is finished at done state 492.

Referring now to 8a,, diagnostic script engine program 490 will be described.From initial state 492, diagnostic script engine program 490 proceeds to state 494 and removes to carry out the script routing function.According to such input parameter: the symptom of patient's principal disease, from symptom begin till now time, patient's past medical history, the result that comes from any other script or from current script series the result that comes of the script of time early, this script distance is selected a suitable DSQ script.The recognizer of patient's principal disease.Principal disease can be divided into following kind: relate to the cause of dissection system, patient's problem, as the tabulation of wound or infection, the principal disease that arranges in alphabetical order, corresponding they the ICD-9 number of disease or the MDATA catalog number (Cat.No.) of their principal disease.After selecting suitable DSQ script, program 490 proceeds to state 496 and goes to fetch selected script from script database 300 (Fig. 6 b).At this moment, diagnostic script engine program 490 requires DSQ list script engine 500 to go to use the DSQ tabulation in execution and patient's query script.To further specify this DSQ list script engine 500 in conjunction with Fig. 9 and 11.

Diagnostic script engine program 490 aftertreatment DSQ script engines are in the result of state 502.Various types of processing are carried out at state 502, as following exemplary illustration by state 506 to 526.May comprise in the action that state 502 is carried out and determine the list of diseases in play and the credibility of the disease in the list of diseases out-of-bounds.In play and out-of-bounds the credibility of some or all diagnosis in the list of diseases can report to patient and/or doctor.In the boundary with list of diseases out-of-bounds diagnosis compiled a kind of differential diagnosis tabulation with relevant credibility.The whole bag of tricks of determining the credibility of diagnosis comprises, for example checks the credibility of table or sensitive factor group.The patent of the former applicant of sensitive factor issue, U.S. Patent number 5594638, in be described, be entitled as " the computing machine medical diagnosis system comprises reentry function and sensitive factor ".Decide according to the result's who determines at decision state 504 type by the next one action that program 490 is carried out.The result type of various demonstrations will be described now.At state 506, diagnostic script engine program 490 is entrusted to another script to patient, and this script is selected at state 494 as the front has illustrated.At state 508, program 490 produces suitable medical diagnosis or suggestion.Move on to function 510, this suggestion is assigned to suitable part.To further specify function 510 in conjunction with Fig. 8.Being assigned with back program 490 in suggestion finishes at state 512.

At state 514, program 490 is carried out a specific intermediate analysis.It is how to change along with the time or generate that the diagnostic script engine goes to study in a kind of given disease a kind of specific symptom.At state 516, program 490 is stored in script and enters the result who collects in patient's the process of record.At state 518, program 490 makes patient enter a medical information storehouse, and it is the part in the MDATA system 100.At state 520, the later continuation that program 490 removes to be ranked the script that temporarily stops.Typically, this situation can take place when patient can not finish whole script in dialogue.Do not have in the situation that disease reaches threshold value in the boundary a kind of, the diagnostic script engine may offer patient to one group of list of diseases of weighting maximum in the possibility that constantly reduces.In this situation, at state 522, program 490 may be ranked one to reentry to talk with and go to allow to postpone a period of time and look at whether to have after a while diagnosis to produce.This reentry feature will explain in being entitled as in " computing machine medical diagnosis and treatment suggesting system for wearing " of applicant's co-applications.At state 524, program 490 request patients go to finish test and inquire about this system again.This test can comprise the test of self check behavior, image format (258, Fig. 2) or the laboratory test (260, Fig. 2).At state 526, program 490 provides any urgent result supplier's conduct action immediately to health maintenance.Program 490 finishes at state 512.

Referring to Fig. 8 b, this distribution diagnosis or suggestion function 510 will be described now.From initial state 511, function 510 proceeds to state 512, and the result of various tabulations is collected according to the situation that one or more diseases or diagnosis reach threshold value in the there.Proceed to state 515, function 510 is checked the suitable degree of treatment tables and the current methods of treatment of this diagnosis of being made by this system.Proceed to state 517, function 510 determines that who is the recipient of this diagnosis or suggestion.This part is realized by searching and managing profile of tooth table 519.According to the various factors, stay in any country as patient, the definite type that can tell patient's information of management profile of tooth table.As the result of searching and managing profile of tooth table 519, suggestion or diagnosis send formally access or be necessary to know the entity 521 of these medical informations of patient 114, doctor 444, management maintenance mechanism 446 or other to.There are a lot of information can to share with patient and must share with patient's doctor.For example, what in play with what out-of-bounds and what be patient's specific differential diagnosis? promptly after patient answers all problems, can be the mark graduation of all various disease.This is helpful for the doctor.This management profile of tooth table 519 uses obtainable information in patient's record, as patient's postcode or area code, goes to discern their address.

Referring to Fig. 9, DSQ script engine program 500 will be described now.From initial state 530, program 500 proceed to state 532 go access selected pass to its DSQ file by the diagnostic script engine.Proceed to state 534, the temporary transient tabulation that program 500 initialization are used by script engine.Temporarily referring to Figure 10, the temporary transient tabulation 522 of program 500 initialization symptoms is removed it, and the temporary transient tabulation 550 of initialization disease makes it that all diseases in the main list of diseases 324 be arranged.At this moment, one of them in the program 500 selection diseases is processed, and the identification of selection quilt is a symptom in this disease then.For the existence of determining this symptom among the patient whether, program 500 proceed to that state 536 goes to select symptom first to ask patient's problem.At state 538, program 500 asks a question to patient.Move on to state 540, program 500 receives patient's answer and checks the correctness of their answer according to the problem of this proposition.Being used to upgrade this DSQ in state 542 patients' answer then temporarily tabulates.

Proceed to decision state 544, program 500 has determined whether to reach the diagnosis or the end of script.If also do not have, it is not the next problem that is chosen in the current symptom that program 500 proceeds to state 546, goes to handle the next symptom that is used for current disease exactly, has been putd question to if be used for all problems of current symptom.Program 500 circulates between state 538 to 546 and reaches end, obtains diagnosis, a user and ask script to be terminated or script engine determines that this script must finish up to script.When reaching this diagnosis or finishing, program 500 is not to return this diagnosis, at state 543 patient is entrusted to different scripts, postpones current script at state 545 at state 541, is exactly to finish current script at state 547.Program 500 finishes in return state 548.

Referring now to Figure 10,, will the part of the tabulation of using be described in the operation process of DSQ list script engine 500.Begin till now time according to user's input 460 with from symptom, (Fig. 8 a) determines a script it to be sent to DSQ list script engine 500 the script distance 494 of this diagnostic script engine 490.Also used from the record of the current patient of patient's medical history 254 by script distance 494.The medical diagnosis script that use receives from this script distance, DSQ list script engine 500 access master list of diseases 324.Disease in the main list of diseases is copied to the temporary transient tabulation 550 of disease.Reasonable time in the operating process of DSQ list script engine 500 is copied to the temporary transient tabulation 552 of symptom selectively from the symptom of the primary symptom shape of current disease tabulation 326, will explain in conjunction with Figure 12.When symptom is identified in the query script patient, the weighted value of symptom and/or the weighted value that is used for the problem of symptom will be added to the mark in the current disease of the temporary transient tabulation 550 of disease.When a kind of mark of special disease reached in the boundary threshold value, this disease was moved in the boundary in the list of diseases 554.On the other hand, if the mark of current disease reaches out-of-bounds disease threshold value, this disease is moved to out-of-bounds in the list of diseases 556 so.The symptom that is identified, the disease in the boundary, disease out-of-bounds and neither also can not existed in patient's the medical history 254 in play in out-of-bounds disease.During during when a script is finished or at a terminal point or the checkpoint in scenario process, the disease that remains in the temporary transient tabulation 550 of disease also can be written in patient's the medical history 254.On the other hand, patient's symptom and disease information can be written in the local user's data storer 184 (Fig. 6 b), replace maincenter patient medical history 254.

Referring now to Figure 11,, DSQ list script engine 500 will be described below.This explanation will provide than the script engine program that provides in conjunction with Fig. 9 and summarize more detailed content.From initial state 580, script engine program 500 proceeds to state 582, and the there is from the temporary transient tabulation 550 of script master list of diseases 324 (Figure 10) initialization disease.Move on to state 584, this script engine program from current and/or before patient's dialogue access patient's data.Script engine program 500 is used MDATA back-up systems 440, and (Fig. 6 a) goes to obtain patient's the data and the data of any other needs with database 442.On the other hand, patient's data can be fetched from the storer 184 (Fig. 6 b) of local user data.

Proceed to state 586, script engine program 500 is selected the disease that is considered.Can make ins all sorts of ways selects the order of the disease that is considered.For example, the most urgent disease can at first be considered, and and then is serious disease, is general diseases then.On the other hand, perhaps in conjunction with should urgent/serious pattern, with the most general disease in first kind of disease being considered population that can be the patient place.This script engine program can be used other source of the certificate address information in telephone number, postal region number or patient's the historical record, goes to determine the address at this population cluster or patient place.In case this program begins, the another kind of method of disease order of selecting adopts this disease that the highest symptom weighted value is arranged, and for example this approaches the disease diagnosed most.The yardman of this script preferably is arranged to the order that they will be considered to the disease that is used for current script.Determined current with the disease that is considered after, script engine program 500 proceeds to " select with the symptom that is considered " program 588.Program 588 will be described further in conjunction with Figure 12 for definite this symptom that will be considered of current disease.

Script engine program 500 goes to check at decision state 590 whether sky indicates that it may be to be provided with to selected symptom in the operational process of program 588.If the selected symptom of this corresponding current disease is empty sign, program 500 advances to decision state 616 so, goes to determine whether other disease that will consider.Yet if this selected symptom is not empty sign, script engine program 500 proceeds to state 592 and goes to select the problem flow process so, and it is offered patient.Relevant with each symptom is a logic of questions flow process, and it draws symptom.A logic flow can be used as one " complicated problems ", and for example, a problem is made of one of them that also can produce in a plurality of answers a plurality of problems.Preferably select such problem flow process, for example, it can produce an answer, and include current have the highest may draw the symptom of going in the disease into just under consideration.Advance to state 594, script engine program 500 is carried out the node of current flow process then.Proceed to state 596, script engine program 500 partly offers the user to the problem of flow nodes.Each problem preferably includes one group of information text, command text and a problem.In order to ask a question, this script is at first exported to patient to information text, is command text then, is question text at last.At this moment question text indicates to patient expects an answer.

Proceed to and handle question-answering program 598, the answer that this script engine processes is come from the user.Further specify program 598 below in conjunction with Figure 13.Wherein a kind of in preferably following three types of this flow nodes: symptom, problem or program.Script engine program 500 is determined the type of flow nodes at decision state 600.If node type is problem or program, script engine program 500 moves on to state 594 (problem circulation Q) and goes to carry out next flow nodes so.Yet if the flow nodes type is a symptom type, program 500 proceeds to state 602 so, according to the answer that receives from patient, removes to upgrade the temporary transient tabulation of symptom 552 (Figure 10) there.According to this answer, for every kind of symptom is distributed a weighted value.On the other hand, if current symptom is used a plurality of problems, their some problem has relevant weighted value, and the weighted value (if there is) of so corresponding current problem is the accumulated value of corresponding current symptom.

When the DSQ script obtained a symptom, it upgraded all Symptomatic diseases.That is, the single answer that comes from patient can change the weighting of the symptom of all diseases that are being considered.These " promotion " one or more diseases are more near the threshold value of this diagnosis.

Proceed to function 604, script engine program 500 is carried out the back response and is handled, and goes further to upgrade the temporary transient tabulation 552 of symptom.The example that back response is handled comprises: if-relation of relation, simul relation, order relation and other similar type so.For example, if the value of serious symptom is 9, weighted value 75 can be added in the diagnosis of cholecystalgia so, and deducts weighted value 50 in the typhlitic diagnosis.Discussion has been done in conjunction with Fig. 4 b (obtaining disease knowledge) in other back response relation front.After response was handled after finishing, script engine program 500 proceeds to upgraded list of diseases program 606.In program 606, script engine upgrades the mark in the temporary transient tabulation of diseases according to the symptom of having upgraded tabulation temporarily 552 and cancellation circle is interior and disease out-of-bounds.To further specify in conjunction with Figure 14 and upgrade list of diseases program 606.

When program 606 was finished, some disease was drawn in the boundary or is marked out-of-bounds, reduces the length of the temporary transient tabulation of disease 550 (Figure 10) like this.Yet threshold value did not reach out-of-bounds threshold value yet in the boundary if a kind of disease had not both reached, and did not remove this disease from the temporary transient tabulation of disease so.Like this, be left to be the next iteration service of checking the symptom of corresponding disease in 608, one list of diseases of having upgraded of state and a symptom tabulation of having upgraded.Move on to decision state 610, script engine program 500 determines whether to also have the symptom of other corresponding current disease in the temporary transient tabulation 552 of symptom.If also have, script engine program 500 is according to the absolute value relevant with state 612 so, and selecting has this symptom of maximum weighted value, and proceeds to the problem flow process that state 592 (symptom circulation S) goes to select this new symptom of correspondence.Yet if do not have other symptom in the temporary transient tabulation 552 of symptom, as what determine at decision state 610, script engine program 500 proceeds to state 614 and delete current disease from the temporary transient tabulation 550 of disease so.

Proceed to decision state 616, whether script engine program 500 determines the temporary transient tabulation 550 of the disease of corresponding current script sky.If not empty, script engine program 500 moves on to state 586 (disease circulation D) and goes to consider next disease in the script so.If the temporary transient tabulation 550 of the disease of corresponding current script is empty, script engine program 500 proceeds to the type that decision state 618 goes the result of definite script so.At state 620, a kind of possible result is that one or more diseases are drawn in the boundary or marked out-of-bounds.At state 622, another kind of result's type is that script engine has determined to entrust another script or another business.Script engine program 500 finishes and gets back to diagnostic routine 490 (Fig. 8 a) in return state 624.

Referring to Figure 12, the selection symptom program 588 that this was mentioned will be described now in Figure 11.From initial state 640, select symptom program 588 to proceed to state 642 and remove to remove the temporary transient tabulation of symptom 552 (Figure 10).Proceed to state 644, the current disease of selecting symptom program 588 to be accessed in the script master list of diseases 324 (Figure 10).Advance to state 646, program 588 is determined the next symptom of current disease.Continue to decision state 648, program 588 determines whether the problem flow process of this symptom was moved for this patient in the past.For example, this symptom is in another disease or even determined in another script of this patient.If this problem flow process was not moved in the past, program 588 proceeds to state 650 and this symptom is added in the temporary transient tabulation of symptom so.After in this symptom being added to the temporary transient tabulation of symptom, if or moved before the problem flow process of this symptom, program 588 moves on to decision state 652 so.At decision state 652, program 588 determines whether the symptom of other corresponding current disease in addition.If also have, program 588 is moved back into the next symptom that state 646 goes to determine current disease so.

If no longer included the symptom of other corresponding current disease, and definite at state 652, and whether program 588 is proceeded so, remove to determine the temporary transient tabulation 552 of symptom sky at decision state 654.If empty, select symptom handling procedure 588 to move on to state 656 current disease of deletion from the temporary transient tabulation 550 of disease so.For example, if be considered before all symptoms of corresponding this disease, this situation so just takes place in this and that script.In this case, select symptom handling procedure 588 to return the symptom sign of a sky at state 658.Get back to decision state 654, if select symptom handling procedure 588 to determine that the temporary transient tabulation of symptom is not empty, continue so in state 660 operations, symptom temporary transient tabulation in there is classified according to the absolute value of weighted value.Proceed to state 662, program 588 is selected this symptom of the absolute value maximum of weighted value.Select symptom handling procedure 588 to get back to program 500 (Figure 11) with this selected symptom at state 664.

Referring to Figure 13, the processing question-answering program 598 that this was mentioned will be described now in Figure 11.From state 690, program 598 proceeds to state 692 and goes the validity of checking that the user answers.Proceed to decision state 694, program 598 determines whether this answer is effective.If it is invalid to answer, program 598 proceeds to state 696 question text is repeatedly exported to the user so, is moved back into the validity that state 692 goes to check this user's answer then.The inspection that a time uses up situation takes place in the operational process of handling question-answering program 598.It is to go to look at whether it means may lose consciousness or the change of intelligence state that evaluation time uses up.If so for example, enter an intelligence state subroutine or call out the emergency medical personnel.

If be confirmed as effectively in 694 answers of decision state, program 598 proceeds to decision state 698 and goes to determine the current node types of being handled by DSQ script engine 500.If node type is the symptom node, program 598 proceeds to state 700 and goes to select the symptom value relevant with current flow nodes so.The symptom node resembles to the answer of challenge and returns this symptom.Be returned to symptom script engine program 500 (Figure 11) in state 702 symptom values then.If node type is a problem node, program 598 proceeds to state 704 answer is converted to path digital.Advance to state 706, program 598 numeral that on current flow nodes pathname, adds paths.State 704 and 706 is as the next problem node of determining to be performed.Get back to decision state 698, if determined that node types is a program node, program 598 proceeds to state 710 so.At state 710, program 598 is carried out by the program of present node appointment and is obtained one and returns numeral.Proceed to state 712, program 598 is added on current flow nodes pathname and is returned numeral.User mode 710 and 712 is gone definite next problem that will be performed.The program that is performed at state 710 can be a sub-script or other function or the subroutine that is used for drawing from patient other medical information.When state 706 and 712 was finished, program 598 turned back to DSQ script engine program 500 (Figure 11) in return state 708.

Referring to Figure 14, the renewal list of diseases program 606 that this was mentioned will be described now in Figure 11.From initial state 730, program 606 proceeds to state 732 and removes the temporary transient tabulation of access disease 550 (Figure 10).Proceed to decision state 734, program 606 determines whether also have other disease in the temporary transient tabulation 550 of disease.If no, program 606 turns back to DSQ script engine program 500 (Figure 11) in return state 736.Yet if in temporary transient tabulation other disease is being arranged, program 606 proceeds to state 738 and goes to be accessed in next disease in the temporary transient tabulation 550 of this disease so.Proceed to decision state 740, program 606 determines whether current diseases include just the symptom that the symptom answered by patient or any its back response are handled, as (Figure 11) that is determined in function 604.If have, program 606 moves on to state 742 and the weighted value that symptom is handled in the symptom of just having answered or back response is added on the mark of current disease so.If there is the response of a plurality of backs to handle symptom, this situation typically can take place.If like this, program 606 is moved back into state 742 weighted value of these other symptoms is added on this disease mark so.Need to handle if no longer include other symptom, as what determine at state 744, program 606 proceeds to decision state 746 so.

At decision state 746, program 606 determines whether the mark of disease has met or exceeded threshold value in the boundary.The value of threshold value preferably 1000 in this boundary, but can use the mark of threshold value in other boundary.If program 606 proceeds to state 748 current disease is added in the boundary in the list of diseases 554 (Figure 10).Move on to state 750, program 606 is eliminated this current disease from the temporary transient tabulation of disease 550 (Figure 10), is moved back into decision state 734 then and goes to determine whether also have other disease in this temporary transient tabulation 550.

Get back to decision state 746, if determined that mark does not meet or exceed threshold value in the boundary, program 606 proceeds to decision state 752 so.At decision state 752, program 606 determines whether the disease mark has met or exceeded out-of-bounds threshold value.If program 606 moves on to state 754 current disease is added to out-of-bounds in the list of diseases 556 (Figure 10) so.Advance to state 750, program 606 is eliminated this disease from the temporary transient tabulation of disease 550 (Figure 10), is moved back into decision state 734 then and goes to determine whether also have other disease in the temporary transient tabulation 550 of this disease.

Get back to decision state 752, if the mark of disease is not below or equal to out-of-bounds threshold value, program 606 is moved back into decision state 734 and goes to determine whether also have other disease in this temporary transient tabulation 550 so.Get back to decision state 740, if current disease does not include the symptom just answered or symptom is handled in any its back response, program 606 is moved back into decision state 734 and goes to determine whether also have other disease in this temporary transient tabulation 550 so.

In the boundary threshold value and out-of-bounds the use of threshold value following implication is arranged:

● the weighted value of symptom can be given as a positive number or negative;

● for every kind of disease has been preserved two operation marks: a positive number and a negative;

● positive weighted value is added on the positive mark, and negative weighted value is added on the negative mark;

● weighted value is not reduced;

● two threshold values are used, one in (as 1000 or 10000) draw disease into the boundary and a negative value (as-1000 or-10000) disease is marked out-of-bounds;

● when positive mark met or exceeded positive threshold value, disease was drawn in the boundary;

● when negative mark met or exceeded negative threshold value, disease was marked out-of-bounds;

● if disease two threshold values when script finishes all do not reach, and it is left in " uncertain " tabulation of a disease so, and it can be existed in patient's the medical history.

Referring to Figure 15, will illustrate that now a kind of application produces the replacement type embodiment of medical advice or diagnosis based on branch's script.From initial state 782, proceed to state 784 based on branch's shell script 780 and remove to open medical treatment script file based on branch.Proceed to state 786, program 780 from current and/or before with this patient's of session establishment of patient data.Shell script 780 proceeds to first problem in the state 788 beginning scripts.Advance to state 790, shell script 780 offers the user to current problem.Proceed to state 792, shell script 780 waits for that an effective user answers.Move on to state 794, the answer of shell script 780 recording users.At state 796, shell script 780 is to the node relevant with user's answer. proceed to decision state 798, shell script 780 determine next nodes whether one withdraw from node.If not, program 780 proceeds to state 790 and next problem is offered the user so.The order of the script node that shell script 780 bases are predetermined up to 798 cocycles, withdraws from node up to reaching one at state 790.When reaching when withdrawing from node, shell script 780 moves on to the type that decision state 800 goes to determine script results.Should return a diagnosis in return state 802 based on branch's script 780, return suggestion or return in return state 804 and relate to another script in return state 806.

Six. based on the advantage of the processing of tabulating

Should quick, accurate and complete advantage be arranged based on relative other medical diagnostic method of the disposal system of tabulation.Especially, should have based on the disposal route of tabulation:

-medical knowledge is organized into the also treatable tabulation of other method;

-providing can be to its accuracy and the diagnosis checked of integrality;

-produce than people and can use the better script that the script based on branch writes out;

-simplified more new script, when medical knowledge changes;

-allow to be tested by automatic mode;

-can be used as a kind of from call function based on branch's script;

-be that computer platform independence, media independence and language are independent;

-allow script is more easily translated into other human language;

The method of the actual diagnosis of-feedback doctor.

Appendix

DSQ is based on the example of list script

This tabulation has shown the example of an ASCII character file more up hill and dale, and it includes the tabulation that is used as based on the starting point of the processing of tabulating.This tabulation only is used for display format and relation.It may not express correct and complete medical information.The principal disease of this demonstration script is " uncomfortable ".

′MALARIA.TXT

DEF?H

h_format?5

h_complaint?s_malaise END?H DEF?D d_falc "084.0"?"Falciparum?Malaria"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder 100

s_2bouts_other 250

s_3bouts_other 250

s_pnotest 5

s_pnegative -700

s_pfalcip 700

s_pvivax -700

s_povale -700

s_pmalar -700

s_pmixed -700d_vivax "084.1"?"Vivax?Malaria"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder?100

s_2bouts_48 350

s_3bouts_48 450

s_pnotest 5

s_pnegative -700

s_pfalcip -700

s_pvivax 700

s_povale -700

s_pmalar -700

s_pmixed -700d_quartan "084.2"?"Quartan?Malaria"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder 100

s_2bouts_72 350

s_3bouts_72 450

s_pnotest 5

s_pnegative -700

s_pfalcip -700

s_pvivax -700

s_povale -700

s_pmalar 700

s_pmixed -700d_ovale "084.3"?"Ovale?Malaria"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder 100

s_2bouts_other 250

s_3bouts_other 350

s_pnotest 5

s_pnegative -700

s_pfalcip -700

s_pvivax -700

s_povale 700

s_pmalar -700

s_pmixed -700d_mixed "084.5"?"Mixed?Malaria"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder 100

s_1bout_23days 200

s_1bout_other 200

s_2bouts_other 200

s_3bouts_other 300

s_pnotest 5

s_pnegative -700

s_pfalcip -700

s_pvivax -700

s_povale -700

s_pmalar -700

s_pmixed 700d_unspec "084.6"?"Malaria，unspec"

s_tropics 200

s_lethargic 100

s_fever 200

s_chills 200

s_nochills -100

s_sweats 200

s_nosweats -100

s_cfsinorder 200

s_cfsnotinorder 100

s_pnotest 100d_notmal "-"?"Not?Malaria"

s_nofever 100

s_nochills 300

s_nosweats 300

s_nocfs 700

s_cfsnotinorder 300

s_pnegative 1000

s_pfalcip -600

s_pvivax -600

s_povale -600

s_pmalar -600

s_pmixed -600END?DDEF?S

s_malaise 0 "general?ill?feeling"

s_tropics f_tropics "recently?in?tropics"

s_nottropics f_tropics "notrecently?in?tropics"

s_lethargic f_lethargic "has?been?tired/lethargic"

s_notlethargic f_lethargic "not?tired/lethargic"

s_fever f_fever "has?fever"

s_nofever f_fever "has?no?fever"

s_chills f_chills "has?chills"

s_nochills f_chills "has?no?chills"

s_sweats f_sweats "has?sweating"

s_nosweats f_sweats "has?no?sweating"

s_nocfs f_cfs "did?not?have?CFS"

s_cfsinorder f_cfs "had?CFS"

s_cfsnotinorder f_cfs "had?CFS,but?not?in?order"

s_pnotest f_ptest "not?tested?for?Plasmodia"

s_pnegative f_ptest "plasm?test?negative"

s_pfalcip f_ptest "test+for?P.falciparum"

s_pvivax f_ptest "test+for?P.vivax"

s_povale f_ptest "test+for?P.ovale"

s_pmalar f_ptest "test+for?P.malariae"

s_pmixed f_ptest "test+for?mixed?Plasmodia"

s_1bout_1day f_cfs "1CFS?bout?lasting?1day"

s_1bout_23days f_cfs "1CFS?bout?lasting?2-3days"

s_1bout_other f_cfs "1cfs?bout?of?unk?duration"

s_2bouts_48 f_cfs "2CFS?bouts,48h?apart"

s_2bouts_72 f_cfs "2CFS?bouts,72h?apart"

s_2bouts_other f_cfs "2CFS?bouts;unknown?interval"

s_3bouts_48 f_cfs "3+CFS?bouts?every?48hours"

s_3bouts_72 f_cfs "3+CFS?bouts?every?72hours"

s_3bouts_other f_cfs "3+CFS?bouts?of?unk?interval" END?S DEF?I

s_nochills s_nocfs

c_ncfever s_nocfs

s_nosweats s_nocfs

s_chills s_fever s_sweats s_cfs END?I

DEF?Ff_tropics

"1" q_tropics

"11"?s_tropics

"12"?s_nottropicsf_lethargic

"1" q_lethargic

"11"?s_lethargic

"12"?s_notlethargicf_fever

"1" q_fever

"11"?s_fever

"12"?s_ncfeverf_chills

"1" q_chills

"11"?s_chills

"12"?s_nochillsf_sweats

"1" q_sweats

"11"?s_sweats

"12"?s_nosweatsf_cfs

"1" q_cfs

"12"?s_nocfs

"11"?q_cfsorder

"112"s_cfsnotinorder

"111" q_cfsbouts

"1110" s_nocfs

"1111" q_d1bout

"11111" s_1bout_1day

"11112" s_1bout_23days

"11113" s_1bout_other

"1112" q_d2bouts

"11121" s_2bouts_48

"11122" s_2bouts_72

"11123" s_2bouts_other

"1113" q_d3bouts

"11131" s_3bouts_48

"11132" s_3bouts_72

"11133" s_3bouts_otherf_ptest

"1" q_ptest

"12" s_pnotest

"11" q_pfound

"110" s_pnegative

"111" s_pfalcip

"112" s_pvivax

"113" s_povale

"114" s_pmalar

"115" s_pmixedEND?F

DEF?Q

q_tropics 0?t_qtropics 12?t_kyes t_kno

q_lethargic 0?t_qlethargic 12?t_kyes t_kno

q_fever 0?t_qfever 12?t_kyes t_kno

q_chills 0?t_qchills 12?t_kyes t_kno

q_sweats 0?t_qsweats 12?t_kyes t_kno

q_cfs 0?t_qcfs 12?t_kyes t_kno

q_cfsorder 0?t_qcfsorder 12?t_kyes t_kno

q_ptest 0?t_qptest 12?t_kyes t_kno

q_pfound 0?t_qpfound?012345?t_knone?t_falc t_viv t_ov t_mal t_mix

q_cfsbouts 0?t_qcfsbouts 0123?t_knone t_kone t_ktwo t_k3plus

q_d1bout 0?t_qd1bout 123 t_kup1day t_k23days t_kother

q_d2bouts 0?t_qd2bouts 123 t_k48hours t_k72hours t_kother

q_d3bouts 0?t_qd3bouts 123 t_k48hours t_k72hours t_kotherEND?QDEF?T

t_falc FALCIPARUM

t_mal MALARIAE

t_mix MIXED

t_ov OVALE

t_viv VIVAX

t_k23days 2-3?DAYS

t_k3plus THREE+

t_k48hours 48?HOURS

t_k72hours 72?HOURS

t_kno NO

t_knone NONE

t_kone ONE

t_kother OTHER

t_ktwo TWO

t_kup1day UP?TO?ONE?DAY

t_kyes YES

t_qcfs Did?you?have?Chills,Fever,and?Sweating

t_qcfsbouts How?many?bouts?of?C-F-S?did?you?have

t_qcfsorder Did?you?have?C-F-S?in?that?order

t_qchills Do?you?have?chills

t_qd1bout How?long?did?that?1?bout?last

t_qd2bouts What?was?the?time?between?those?2?bouts

t_qd3bouts How?far?apart?were?these?bouts

t_qfever Do?you?have?fever

t_qlethargic Have?you?been?tired?or?lethargic

t_qpfound What?Plasmodia?were?found?in?blood

t_qptest Did?you?have?a?blood?test?for?Plasmodia

t_qsweats Do?you?have?sweating

t_qtropics Have?you?been?in?the?tropics?recently?END?T

Claims

1. computerized diagnostic system comprises:

Be used for providing the device of at least one group of list of diseases to computing machine, every kind of disease all relevant with one group of symptom tabulation and

Be used for the patient data is imported the device of described computing machine;

For the current symptom that presents, be used for total weighted value of the accumulation of the disease relevant with described symptom is provided the device of a weighted value;

Be used to produce one or more devices that may diagnose of described disease; With

Be used for from the described device that one or more may be diagnosed of computing machine output;

Each described symptom is all relevant with at least one problem in one group of table;

The described device that is used to import the patient data comprises and is used for order and proposes a plurality of described problems to draw the device of answer to the user, and wherein each problem is all carried out Dynamic Selection according to the patient data that before may import in the described computing machine; With

Describedly be used to produce the device that may diagnose and comprise:

At with determine to exist the device of one or more described symptoms on the basis of answer of the relevant problem of fast symptom; Also comprise:

By being compared, total weighted value of described accumulation and threshold value from described possible diagnosing the illness, select a kind of device that is used for described computing machine that is diagnosed the illness.

2. according to the system of claim 1, it is characterized in that this symptom is that existence or non-existent situation according to one or more other symptoms constitutes.

3. according to the system of claim 1, it is characterized in that total weighted value of the accumulation that exists at least a disease of one group of selected symptom increases extra weighted value.

4. according to the system of claim 1, it is characterized in that in the weighted value of the symptom of first select time of diagnostic procedure different with the weighted value of the symptom of second select time.

5. according to the system of claim 1, the weighted value of the symptom in it is characterized in that the weighted value of the symptom in the first time, the order of severity was reported and the order of severity being reported for the second time is different.

6. according to the system of claim 1, the difference accumulated weights value that it is characterized in that offering this diagnosis with the selected symptom group that a kind of specific chronological order takes place be not with a kind of certain order generation one group selected that symptom was provided was different.

7. according to the system of claim 1, it is characterized in that with the beginning of a selected symptom group or finish accumulated weights value that order offers total weighted value and independent selected different that these symptoms were provided.

8. according to the system of claim 1, it is characterized in that total weighted value, this disease is put under be used for further diagnosis inquiry in the boundary according to this accumulation.

9. according to the system of claim 1, it is characterized in that total weighted value, this disease is marked out-of-bounds be used for further diagnosis inquiry according to this accumulation.

10. according to the system of claim 1, it is characterized in that before asking a question, puing question to problem earlier about the emergency treatment disease for non-emergency treatment disease.

11. according to the system of claim 1, the total weighted value that also comprises the accumulation that is used for determining corresponding a kind of disease whether less than threshold value out-of-bounds to get rid of a kind of device of possible diagnosis.

12. system according to Claim 8 is characterized in that providing this disease is put under definite degree in the boundary.

13., it is characterized in that providing this disease marked out-of-bounds definite degree according to the system of claim 9.

14., it is characterized in that each diagnosis respectively has many diagnosis of definite degree to be accumulated a patient's specific difference diagnostic listing according to the system of claim 1.

15., it is characterized in that each problem that patient is asked about all constitutes a relevant issues weighted value of the total weighted value of accumulation of at least a disease according to the system of claim 1.

16. according to the system of claim 1, thus it is characterized in that the described device that is used to import the patient data also comprise at certain time intervals after order propose further problem to the user and draw further answer.

17. according to the system of claim 16, the symptom that it is characterized in that is according to the existence of one or more other symptoms or does not exist and constitute.

18., it is characterized in that the total weighted value of the accumulation that exists at least a disease of one group of selected symptom increases extra weighted value according to the system of claim 16.

19., it is characterized in that selecting the weighted value of the symptom when the user asks a question for the first time and selecting the weighted value of the symptom when the user asks a question different for the second time according to the system of claim 16.

20. according to the system of claim 16, the weighted value of the symptom in it is characterized in that the weighted value of the symptom in the first time, the order of severity was reported and the order of severity being reported for the second time is different.

21. according to the system of claim 16, the accumulated weights value that it is characterized in that offering total weighted value with one group of selected symptom that a kind of specific chronological order takes place be not with a kind of specific order generation one group selected that symptom was provided was different.

22., it is characterized in that with the beginning of one group of selected symptom or finish accumulated weights value that order offers total weighted value and independent selected different that symptom was provided according to the system of claim 16.

23. according to the system of claim 16, it is characterized in that total weighted value, this disease is put under be used for further diagnosis inquiry in the boundary according to this accumulation.

24. according to the system of claim 16, it is characterized in that total weighted value, this disease is marked out-of-bounds be used for further diagnosis inquiry according to this accumulation.

25., it is characterized in that before asking a question, puing question to problem earlier about the emergency treatment disease for non-emergency treatment disease according to the system of claim 16.

26. according to the system of claim 16, whether the total weighted value that also comprises the accumulation that is used for determining corresponding a kind of disease less than threshold value out-of-bounds, to get rid of a kind of device of possible diagnosis.

27., it is characterized in that providing this disease is put under definite degree in the boundary according to the system of claim 23.

28., it is characterized in that providing this disease marked out-of-bounds definite degree according to the system of claim 24.

29., it is characterized in that each diagnosis respectively has many diagnosis of definite degree to be accumulated a differential diagnosis tabulation according to the system of claim 16.

30., it is characterized in that all providing the weighted value of relevant issues for the total weighted value of the accumulation of at least a disease to each problem that patient proposes according to the method for claim 16.

31., it is characterized in that determining by a doctor at interval during this period of time according to the system of claim 16.

32. according to the system of claim 31, it is characterized in that this doctor determines one group of time interval for a kind of special disease, and can be and the different at interval At All Other Times time periods any a period of time at interval.

33. according to the system of claim 32, it is characterized in that being every period definite time interval, Computer Storage symptom tabulation and be used for the problem list of every kind of symptom.

34. according to the system of claim 32, it is characterized in that designing these problems is in order to follow the tracks of the time dependent process of a kind of disease.

35., it is characterized in that also being included in receiving and set up the device that has described symptom after the questions answer relevant with described symptom according to the system of claim 1.

36. system according to claim 35 is characterized in that also being included in and receives described device of answering total weighted value of weighted value increase of the accumulation of back for the every kind disease relevant with described symptom.

37. system according to claim 36 is characterized in that also being included in receiving after the described answer threshold value in total weighted value of each described accumulative total and the boundary or is out-of-bounds compared sharp device.

38. system according to claim 35, it is characterized in that also comprising, after each described device determines to exist described symptom, also determine with described disease association and at its other symptoms that also do not ask a question, and after each described device determines to exist described symptom, dynamically reconfigure described residue symptom.

39., it is characterized in that described dynamically reconfiguring depends on the described weighted value relevant with each described residue symptom according to the described system of claim 38.

40. system according to claim 16 is characterized in that also being included in and receives the device there is described symptom at setting up after the answer of the problem relevant with described symptom.

41., it is characterized in that also being included in and receive the described back provides a weighted value for total weighted value of the accumulative total of every kind of disease relevant with described symptom the device of answering according to the described system of claim 40.

42., it is characterized in that also being included in and receive after the described answer device that the value of cutting off from total weighted value of each described accumulation and the boundary or is out-of-bounds compared according to the described system of claim 41.

43. according to the described system of claim 40, after it is characterized in that also being included in each described device and determining to exist described symptom, determine the device with the residue symptom that does not also ask a question of described disease association, and after each described device determines to exist described symptom, reconfigure described remaining symptom dynamically.

44., it is characterized in that described dynamically reconfiguring depends on the described weighted value relevant with each described remaining symptom according to the described system of claim 43.

45. a computerized diagnostic system comprises:

One or more described diseases of described computer diagnosis; With

Be used for exporting the device of described diagnosis from described computing machine;

The described device that is used to import the patient data comprises and is used for order and proposes a plurality of described problems drawing the device of answer to the user, and proposes further problem and draw further answer to user's order after certain time interval;

The computing machine of described one or more described diseases of diagnosis comprises:

On at the basis of the questions answer relevant, set up the device that has one or more described symptoms with the symptom of described foundation; With

By being compared, the total weighted value of described accumulation and the value of cutting off from diagnose the device that whether has described disease.

46. a computerized diagnostic system comprises:

One or more described diseases of described computer diagnosis; With

Be used for exporting the device of described diagnosis from described computing machine; Wherein

Each described symptom is all closed with one group of epiphase of at least one problem;

The described device that is used to import the patient data comprises and is used for order and proposes a plurality of described problems drawing the device of answer to the user, and selects each problem dynamically according to the patient data that may be input in advance in the described computing machine; And

Order proposes further problem and draws further answer to the user after certain time interval;

The computing machine of described one or more diseases of diagnosis comprises:

Be used at the basis of the questions answer relevant, setting up the device that has one or more described symptoms with the symptom of described foundation; With

According to described symptom at the device that whether has described disease.

47. a computerized diagnostic system comprises:

One has the computing machine of storer;

Be stored in the tabulation of the disease data in this storer, each disease data is relevant with the tabulation of symptom data, and each symptom data is relevant with the tabulation of a problem; And

A software program of on this computing machine, carrying out, this program is used for constantly asking a question to draw answer, these answer the symptom time that constitutes, the symptom of every kind of formation provides a weighted value for the selected disease data in this tabulation, and this program judgement judges whether to meet or exceed a threshold value to draw a diagnostic result to each a kind of disease.

48., it is characterized in that this storer comprises a random access memory according to the system of claim 47.

49., it is characterized in that this storer comprises a nonvolatile memory according to the system of claim 47.

50., it is characterized in that problem list and the symptom tabulation after this time interval is different from first problem list and the tabulation of first symptom according to the system of claim 16.