CN105373986A - Realization method for hospital film-free cloud computing - Google Patents

Abstract

The invention provides a realization method for hospital film-free cloud computing, suitable for the field of medical Internet. All medical data in a medical system of a hospital is synchronized to a pre-positioned server, and further synchronized to a server and stored by the server, and the medical data is shared. A patient can inquire medical data of himself/herself via a user client, a doctor can inquire the medical data of the patient when a doctor client is authorized convenience is provided for referrer, the patients can store and manage medical data of themselves, and complete medical data of the patients is provided for the clinical doctors.

Description

The implementation method of hospital's film-free cloud computing

Technical field

The invention belongs to medical Internet technical field, particularly relate to the implementation method of a kind of hospital film-free cloud computing.

Background technology

Patient is when entering hospital and seeing a doctor inspection, usually need to do some image checks, such as CT, X-ray etc., image file after inspection is all kept in hospital medical system, and doctor only can to patient's more representational medical film after completing inspection, and this traditional medical film not only carries inconvenience but also very easily damages, and this is not the image data that patient is complete, when needs of patients is to other hospital admissions or when changing the place of examination, hospital requirements patient does image check again, time-intensive, expensive.

Summary of the invention

In view of the above problems, the object of the present invention is to provide the implementation method of a kind of hospital film-free cloud computing, be intended to solve the technical matters that the user's medical data in existing medical system cannot be shared with patient.

The implementation method of described hospital film-free cloud computing comprises the steps:

When there being newly-increased medical data in medical system, described newly-increased medical data being synchronized to front server, being then synchronized to Cloud Server;

Subscription client sends inquiry request to system server, and system server searches the front server belonging to medical data to be checked, and sends the address of an inquiry password and corresponding front server to subscription client;

According to address and the inquiry password of described front server, subscription client and described front server connect, and the medical data that download user phone number is corresponding with inquiry password.

The invention has the beneficial effects as follows: in the present invention, front server is arranged in hospital, and periodic refreshing view also obtains newly-increased medical data, and newly-increased medical data is synchronized to front server, and then is synchronized to Cloud Server; Reduce the load to hospital medical system itself; Such patient system's server obtains the address of inquiry password and corresponding front server; By oneself medical data just directly can be got by access front server, hospital also without the need to medical film to patient; The medical data with patient access also can be searched by authorizing by other hospitals, has saved time and charge costs.The present invention utilizes new technology of Internet, the medical data of all hospital medical system can be realized sharing with lower cost, for patient referral provides conveniently mode, user can inquire about medical data voluntarily, achieve the self-store and management of patient medical data, the diagnosis for clinician provides the complete medical data of patient.

Accompanying drawing explanation

The structural drawing of Tu1Shi hospital film-free system;

Fig. 2 is the process flow diagram of the implementation method of hospital's film-free cloud computing that the embodiment of the present invention provides;

Fig. 3 is the encryption flow figure of picture file;

Fig. 4 is the arrangement schematic diagram of the image of image file;

Fig. 5 is the fractionation schematic diagram of the image of image file;

Fig. 6 is phone number grouping schematic diagram;

Fig. 7 is a kind of particular flow sheet of step S203 in Fig. 2;

Fig. 8 is the process flow diagram generating login name and log in password;

Fig. 9 is the process flow diagram of subscription client authorized doctor client downloads patient medical data;

Figure 10 is the process flow diagram of Cloud Server medical data backup.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.

The inventive method realizes based on hospital's film-free system, as shown in Figure 1, this system comprises the front server 1 being positioned at each hospital, front server 1 is distributed in zones of different, each front server 1 network is connected to system server 2 and Cloud Server 3, described system server 2 is for information management and control, user's medical data that described Cloud Server 3 is uploaded for storing each front server 1, described system server 3 is also connected with subscription client 4 and doctor's client 5, can also online exchange between subscription client 4 and doctor's client 5.Subscription client 4 can download the medical data of oneself after obtaining the front server address of inquiry password and correspondence from system server from front server; Further, if after doctor's client obtains the mandate of subscription client, also can from front server the medical data of download user.In addition, described Cloud Server 3 is also connected with backup server 6, and the medical data in Cloud Server 3 is timed backup extremely described backup server 6 also.In order to technical solutions according to the invention are described, be described below by specific embodiment.

Fig. 2 shows the flow process of the implementation method of hospital's film-free cloud computing that the embodiment of the present invention provides, and illustrate only the part relevant to the embodiment of the present invention for convenience of explanation.

The implementation method of hospital's film-free cloud computing that the present embodiment provides comprises the steps:

Step S201, when there being newly-increased medical data in medical system, described newly-increased medical data being synchronized to front server, being then synchronized to Cloud Server;

Step S202, subscription client send inquiry request to system server, and system server searches the front server belonging to medical data to be checked, and send the address of an inquiry password and corresponding front server to subscription client;

Step S203, according to the address of described front server and inquiry password, subscription client and described front server connect, and download user phone number and inquire about medical data corresponding to password.

In above-mentioned steps S201, the medical data of patient is all kept in hospital medical system, such as PACS system, LIS system, UIS system, RIS system etc., when having newly-increased medical data in the medical system of hospital, first described newly-increased medical data is synchronized to front server, then Timing Synchronization is preserved to Cloud Server.Therefore the medical data that each front server is preserved all is collected to Cloud Server and preserves.During this step specific implementation, first front server refreshes a view in interval at regular intervals, when finding there is newly-increased medical data in medical system, reading the patient code of newly-increased medical data, checking the store path of numbering and image file; Then read image file according to the store path of described image file, and be stored in local folders; Then newly-increased image file whether is had every Preset Time run-down, i.e. delta file; If there is delta file, then by the port that utilizes sockets socket to combine by the patient code of delta file and correspondence, check that numbering is synchronized to Cloud Server.

Hospital is when doing image check for patient, often do and once check and all can generate unique inspection numbering (ApllyID), and each patient has a patient code (PatientID), such as can comprise the information such as name, sex, cell-phone number, date of birth, simultaneously in medical system, image file all can have a store path.Suppose that inspection unique in Hospital PACS is numbered 7, need a byte; Name is the longest is 4 Chinese characters, need 8 bytes, cell-phone number is 13, need 2 bytes, add other information such as sex and birthdate in addition, article one, patient code's information probably has 16 bytes, and patient code's information of 60 patients is probably 1K byte, and transmitting the time delay that this patient code's information brings in network can ignore.In order to reduce the working pressure of front server device, front server is without the need to real time scan local file, only need every Preset Time run-down, such as 1 minute, due to patient code, check that the data volume of the store path of numbering and image file is less, within every 1 second, a view can be refreshed, image file be read out be kept at this locality in time.After Preset Time, when finding that there is newly-increased image file, utilizing sockets socket port, by the patient code of delta file and correspondence, checking that numbering is synchronized to Cloud Server.Described Cloud Server, as data storage center, can receive the data that each hospital's front server is uploaded.Each hospital arranges a front server, and each front server obtains the medical data in corresponding hospital medical system respectively like this, and then unification is uploaded to Cloud Server.When patient or hospital need to check image data, after obtaining the authorization, search from each front-end server and medical data needed for obtaining.

In the present embodiment, front-end server, after preservation image file, preferably, needs to carry out fractionation combined treatment to image file, i.e. encryption, and obtains key.And by process after delta file and key and correspondence patient code, check numbering be synchronized to Cloud Server.As shown in Figure 3, the encryption of image file specifically comprises the steps:

S301, images all in image file to be sorted according to time sequencing.

When an image check does in hospital, a large amount of check image can be obtained, for CT, generally can give birth to width image up to a hundred.Here each width image is arranged according to rise time order, as shown in Figure 4, hypotheses creation m width image.

S302, for every piece image, at Width stochastic generation w-1 point, at short transverse stochastic generation h-1 point, every width image splits into w × h grid block of pixels.

In every width image, the grid pixel block size split into is random, and unrelated, every width image has w × h block of pixels.

S303, take out a block of pixels and carry out the first round and choose at random from each width image in order, then at random from the remaining block of pixels of each width image take out a block of pixels and carry out second and take turns and choose, like this through w × h wheel choose after all block of pixels selected complete, obtain a storage block matrix B= $\left[\begin{array}{cccc}{T}_1{X}_{11}& T_2{X}_{12}& \mathrm{...}& T_m{X}_{1m}\\ T_1{X}_{21}& T_2{X}_{22}& \mathrm{...}& T_m{X}_{2m}\\ \mathrm{...}& \mathrm{...}& T_z{X}_{ij}& \mathrm{...}\\ T_1{X}_{wh1}& T_2{X}_{wh2}& \mathrm{...}& T_m{X}_{whm}\end{array}\right],$ Wherein m is amount of images, T _zx _ijbe i-th take turns when choosing, from the X of z width image random selecting _ijindividual block of pixels, here X _ij∈ [1, w × h].

With reference to Fig. 5, from piece image to m width image, random selecting block of pixels successively, then continues to continue random selecting in the remaining block of pixels of each image, until the block of pixels of all images is selected complete.Here mark m width image and be respectively T ₁, T ₂..., T _m, in the first round is chosen, image T ₁x ₁₁individual block of pixels is selected, image T ₂x ₁₂individual block of pixels is selected, image T _mx _1mindividual block of pixels is selected, is labeled as T respectively by the block of pixels selected ₁x ₁₁, T ₂x ₁₂, T _mx _1m.The like, so, i-th take turns choose time, from the X of z width image random selecting _ijindividual block of pixels, can be labeled as T _zx _ij.After the block of pixels of all images has been selected, obtain a storage block matrix B= $\left[\begin{array}{cccc}{T}_1{X}_{11}& T_2{X}_{12}& \mathrm{...}& T_m{X}_{1m}\\ T_1{X}_{21}& T_2{X}_{22}& \mathrm{...}& T_m{X}_{2m}\\ \mathrm{...}& \mathrm{...}& T_z{X}_{ij}& \mathrm{...}\\ T_1{X}_{wh1}& T_2{X}_{wh2}& \mathrm{...}& T_m{X}_{whm}\end{array}\right].$ This matrix is reconfigure the data obtained after image splits.

S304, for each block of pixels in storage block matrix is multiplied by a luminance factor, obtain deformation matrix $C=\left[\begin{array}{cccc}{\mathrm{\α}}_{11}{T}_1{X}_{11}& {\mathrm{\α}}_{12}{T}_{12}{X}_{12}& \mathrm{...}& {\mathrm{\α}}_{1m}{T}_m{X}_{1m}\\ {\mathrm{\α}}_{21}{T}_1{X}_{21}& {\mathrm{\α}}_{22}{T}_2{X}_{22}& \mathrm{...}& {\mathrm{\α}}_{2m}{T}_m{X}_{2m}\\ \mathrm{...}& \mathrm{...}& {\mathrm{\α}}_{ij}{T}_z{X}_{ij}& \mathrm{...}\\ {\mathrm{\α}}_{wh1}{T}_1{X}_{wh1}& {\mathrm{\α}}_{wh2}{T}_2{X}_{wh2}& \mathrm{...}& {\mathrm{\α}}_{whm}{T}_m{X}_{whm}\end{array}\right],$ Wherein ${\mathrm{\α}}_{ij}=\frac{\underset{k=1}{\overset{m}{\Σ}}{X}_{ik}-X_{ij}}{w\×h}+\frac{\stackrel{\‾}{T_{i\left(j+1\right)}{X}_{i\left(j+1\right)}}+\stackrel{\‾}{T_{i\left(j-1\right)}{X}_{i\left(j-1\right)}}}{2D},$ Wherein function represent the average brightness value of block of pixels F, D is Benchmark brightness value, and the delta file after described process is described deformation matrix C.As j=m, $\stackrel{\‾}{T_{i\left(j+1\right)}{X}_{i\left(j+1\right)}}=\stackrel{\‾}{T_{i1}{X}_{i1}},$ As j=1, $\stackrel{\‾}{T_{i\left(j-1\right)}{X}_{i\left(j-1\right)}}=T_{im}{X}_{im}.$

In order to be encrypted the data in block of pixels further, in this step, a luminance factor is multiplied by each block of pixels, block of pixels T _zx _ijluminance factor be α _ij, take turns for i-th and choose, image T _zthe X of middle random selecting _ijthe luminance factor of individual block of pixels is relevant to the block of pixels that other all images are worked as selected by front-wheel, and the mean flow rate before and after also having selected by two sub-pictures is correlated with, and in expression formula, D is Benchmark brightness value, for giving tacit consent to known definite value.

S305, the key that generation current image file is corresponding, described key comprises w and h, also comprises cipher key matrix $E=\left[\begin{array}{cccc}{X}_{11}& X_{12}& \mathrm{...}& X_{1m}\\ X_{21}& X_{22}& \mathrm{...}& X_{2m}\\ \mathrm{...}& \mathrm{...}& X_{ij}& \mathrm{...}\\ X_{wh1}& X_{wh2}& \mathrm{...}& X_{whm}\end{array}\right].$

Image, after splitting combined treatment, needs could realize deciphering reduction by key at front server end.Here key comprises w and h selected when splitting in figure, certainly also comprises cipher key matrix $E=\left[\begin{array}{cccc}{X}_{11}& X_{12}& \mathrm{...}& X_{1m}\\ X_{21}& X_{22}& \mathrm{...}& X_{2m}\\ \mathrm{...}& \mathrm{...}& X_{ij}& \mathrm{...}\\ X_{wh1}& X_{wh2}& \mathrm{...}& X_{whm}\end{array}\right],$ Here cipher key matrix is during often wheel is selected, the numbering X of the block of pixels that every width image is specifically selected _ij.Certainly conveniently calculate fast, described key also comprises luminance factor matrix $\left[\begin{array}{cccc}{\mathrm{\α}}_{11}& {\mathrm{\α}}_{12}& \mathrm{...}& {\mathrm{\α}}_{1m}\\ {\mathrm{\α}}_{21}& {\mathrm{\α}}_{22}& \mathrm{...}& {\mathrm{\α}}_{2m}\\ \mathrm{...}& \mathrm{...}& {\mathrm{\α}}_{ij}& \mathrm{...}\\ {\mathrm{\α}}_{wh1}& {\mathrm{\α}}_{wh2}& \mathrm{...}& {\mathrm{\α}}_{whm}\end{array}\right].$

In above-mentioned cipher key matrix E, can be out of shape cipher key matrix further, make cipher key matrix safer.Here from the upper left corner, 2 × 2 region segmentation are carried out to described cipher key matrix of drawing together, there are 4 elements in each region, according to clockwise direction, a position is rotated to the element in each region, if the columns of cipher key matrix is odd number, then do not process last column element, if the line number of cipher key matrix is odd number, then do not process last column element.Such image file, through process, when not obtaining key, being to reduce the image file of patient, ensure that the privacy of patient, particularly for some Very Important Person, seem particularly important to image file encryption.

After image file encryption, also need the patient code of the delta file after by process and key and correspondence, check that numbering is synchronized to Cloud Server, during specific implementation, first, send a handshake to Cloud Server; Then, after receiving feedback signal, by patient code with detect numbering and be sent to Cloud Server, and receive the memory address that Cloud Server returns; Finally, obtain last byte of patient code, when for odd number, the odd-line pixels block in deformation matrix C be synchronized to Cloud Server, then even rows block be synchronized to Cloud Server, finally by key synchronization to Cloud Server; When for even number, the even rows block in deformation matrix C is synchronized to Cloud Server, then odd-line pixels block is synchronized to Cloud Server, finally by key synchronization to Cloud Server.

In above-mentioned steps S202, subscription client sends inquiry request to system server, and system server searches the front server belonging to medical data to be checked.Concrete, patient information table is preserved in described front server, whenever completing user's medical data storage, generate login name and log in password and be sent to subscription client and system server, upgrade patient information table simultaneously, by described login name and log in password, subscription client and system server connect, such subscription client just can send inquiry request to described system server, and system server sends the address of an inquiry password and corresponding front server to subscription client.Described patient information table at least comprises patient code's information, checks the start-stop address of number information and medical data storage in addition, and described patient code's information at least comprises name, the phone number of user.Described front server is that the phone number in patient information table divides into groups, and is convenient to phone number and searches.Concrete packet mode is: from last position of phone number to antepenulatimate, the parity according to bits per inch word is divided into groups, and always has 512 groups.Fig. 6 shows for grouping schematic diagram, a phone number has 11, first is all 1, second is may be only generally 3,5 or 8, then 9 bit digital (i.e. A11-A3 in figure) distribute at random substantially, therefore the present embodiment only need divide into groups according to the parity of latter 9, in figure, 0 represents even number, 1 represents odd number, namely the coding of the first grouping is [000000000], and the coding of the second grouping is [000000001], and the coding of the 3rd grouping is [000000010] ... the coding of the 512nd grouping is [111111111], similar and binary coding.

In above-mentioned steps S203, subscription client, according to the address of described front server and inquiry password, connects with described front server, and the medical data that download user phone number is corresponding with inquiry password.During specific implementation, as shown in Figure 7, comprise the steps:

Step S701, according to the address of described front server and inquiry password, subscription client and described front server connect;

Step S702, read rear 9 bit digital of phone number from back to front successively, and find corresponding grouping according to the parity of bits per inch word;

Step S703, in current group, find corresponding phone number by the repeating query mode of searching;

The phone number that step S704, basis find obtains the start-stop address of medical data storage, and downloads the medical data between described start-stop address.

When user downloads medical data, first front server needs in the patient information table of this locality, search corresponding phone number, and for a hospital, its number of users, often hundreds of thousands rank, adopt repeating query mode to search if search, efficiency is very low.Consider the characteristic of phone number, front server divides into groups, when searching phone number according to each phone number of parity of latter 9 of phone number, never position starts, and reads phone number numeral successively, and according to shown in Fig. 6, root odd even is classified, and can find corresponding grouping fast; Then find corresponding phone number by the repeating query mode of searching in a packet, finally from patient information table, obtain the start-stop address of medical data storage according to the phone number found, and download the medical data between described start-stop address.

In the present embodiment, described patient code's information is 16 bytes, and described inspection is numbered 4 bytes, described whenever completing user's medical data storage, and generate login name and log in password, as shown in Figure 8, specific implementation step is as follows:

S801, patient code's information of 16 bytes and the inspection number information of 4 bytes are formed 20 byte sequences, and in units of position, random alignment is carried out to described 20 byte sequences;

S802, the byte of 20 after random alignment is divided into 5 byte section, obtains original group [r ₁, r ₂, r ₃, r ₄, r ₅], wherein r ₁~ r ₅represent described 5 byte section, each byte section has 4 bytes;

The valuation functions value of each byte section in S803, calculating original group wherein represent v byte of byte section, v=1,2,3 or 4;

S804, delete valuation functions value closest to 1 a byte section;

S805, then select in two byte section any position correspondence to exchange wherein, in two remaining byte section, select any position correspondence to exchange, obtain two generation group, described two generation group become by 4 byte section;

S806, finally optional two byte section composition login name from described two generation groups, two remaining byte section compositions log in password.

The number information of patient comprises the information such as patient's name, phone number, sex, birthdate, age, city, place, always has 16 bytes.Patient hospital often do once check time, all can distribute one check number information, have 4 bytes.Therefore, in the present embodiment, user profile always has 20 bytes.Then random alignment is carried out in these 20 byte step-by-steps, obtain a collating sequence.Collating sequence implies user profile, and these 20 bytes are divided into 5 byte section, and each byte section has 4 bytes.Described valuation functions value is first and the 3rd byte and second and the otherness of the 4th byte for assessment of byte section, more close to 1, illustrates that the data variation amplitude of this byte section is little, is a fairly simple byte section, easily decodes.Therefore need to delete this byte section.Suppose there are two byte section, be respectively R1 and R2, in any position, these two byte section are exchanged, after such as selecting a position, R1 can be split as R11R12, and R2 can be split as R21R22, then, after exchanging, two new byte section R21R12 and R11R22 are just obtained.After optional position exchange is carried out to byte section, obtain two generation group, described two generation group become by 4 byte section equally.In the present embodiment, by selecting any position correspondence to exchange in first and second byte section, exchange selecting any position correspondence in the 3rd and the 4th byte section.Finally optional two byte section can be combined into login name, remaining two byte section are combined into and log in password.Such login name has 8 bytes, and logging in password also has 8 bytes.

What each above-mentioned step described is the flow process of user by the medical data of subscription client inquiry oneself, under actual conditions, doctor, by using doctor's client and also can the medical data of downloading patient after receiving the mandate of subscription client, is specially adapted to inline diagnosis.Patient uses user visitor end doctor to use doctor's client online chatting, patient can describe the direct state of an illness online, and can send mandate to online doctor's client, makes doctor's client be accessed front server, as shown in Figure 9, specifically comprise the steps:

S901, subscription client send authorization message Xiang doctor's client, and described authorization message comprises the name of user, phone number and authorization code.

User is by subscription client and doctor's online exchange, and described subscription client can be the mobile phone terminal or the computing machine that are provided with relative application software.First user uses during user terminal needs to carry out user's registration, fill in relevant information, username and password such as logging in, the information such as name, sex, phone number, ID (identity number) card No., home address of user, these information are all kept in system server.When needs of patients doctor is given in radiodiagnosis x suggestion, authorization message can be sent with subscription client to doctor's client, at least will comprise the name of user, phone number and authorization code.

S902, according to described authorization code, described doctor's client and system server connect.

Described authorization code unique identification patient user, verify authorization code in system server terminal, after being verified, doctor's client and system server connect.

The concrete front server that the medical data that S903, described system server search correspondence according to the name of user and phone number is positioned at, and send query statement to described doctor's client.

Each hospital is deployed with a front server, and for the medical data of store patient, in order to distinguish different front servers, each front server has a numbering.User message table is preserved, comprising the numbering of the name of user, phone number and concrete residing front server in described system server.If a patient user is once in different hospital admission mistakes, so in system server, the front server numbering corresponding to this user has multiple.By searching phone number and checking the name of user, the concrete front server that medical data can be found to be positioned in this locality, and send inquiry password Xiang doctor's client.

S904, described doctor's client connect with corresponding front server according to described query statement, download the thumbnail of the medical data corresponding with described user name and phone number and show.

Described inquiry password contains the numbering of the concrete front server needing inquiry.Doctor's client is connected with corresponding front server by inquiry password, and downloads the medical data corresponding with subscriber phone number from front server.Described medical data is made up of multiple images, and raw image data is comparatively large, for the ease of display, in this step, the thumbnail of medical data is sent to the display of doctor's client.

During this step specific implementation, first doctor's client connects according to the front server of the phone number of user and inquiry code and reference numeral; Then front server searches local patient information table according to phone number, finds the start-stop address of medical data storage; Medical data in described start-stop address is carried out packet fragmentation according to image number; To often organizing view data proportionate sampling, obtain a thumbnail data; All thumbnail datas are sent to doctor's client; Doctor's client demonstrates the thumbnail of medical data.When specifically sampling, for the odd row data of image, select sequence number be odd number as several point, for the even number of lines certificate of image, select sequence number be 2+4k as several point, wherein k is natural number; Then the pixel of selection is carried out arrangement and can obtain thumbnail.

S905, when thumbnail is wherein triggered, from corresponding front server download be subject to trigger the corresponding original complete view data of thumbnail and show.

If when doctor needs to check the original complete view data of wherein some thumbnail, the thumbnail checked needed for directly clicking at doctor's client end interface, then downloads the original complete view data corresponding to thumbnail and shows from corresponding front server.

Finally, the medical data in Cloud Server also needs to back up to backup server.As shown in Figure 10, specifically comprise the steps:

Step S101, statistics Cloud Server in one day free time section probability distribution, determine several time points of the idle maximum probability of Cloud Server.Cloud Server generally can not a whole day non-stop run, and just can work when receiving upstream server uploading data, therefore Cloud Server exists certain free time section, and each free time section probability distribution different.In this step, with one day for time shaft, need add up Cloud Server in one day free time section probability distribution, then several free time section of maximum probability is found, and the time point chosen in space time section, these time points are several time points of the idle maximum probability of Cloud Server.

Step S102, at each described time point, to Cloud Server send query statement;

Step S103, to confirm after instruction when the free time receiving Cloud Server feedback, obtained the newly-increased medical data of current point in time apart from a upper Backup end moment;

Step S104, described newly-increased medical data is sent to backup server and preserves;

During every described time point, send query statement to cloud service, if Cloud Server is busy, then not feedback information, if Cloud Server is idle, then shows that now Cloud Server can complete backup operation, now Cloud Server feedback idle confirmation instruction.Then current point in time is sent to backup server preservation apart from the newly-increased medical data in last Backup end moment by Cloud Server.

In the present embodiment, first add up Cloud Server free time section probability distribution, when putting at one's leisure, the very large probability of Cloud Server is in idle condition, if Cloud Server does, then wait until the state of next free time point inquiry Cloud Server, until to certain some free time, Cloud Server is idle, now can complete backup operation.Compared to timed backup, this method can not affect the work of Cloud Server, utilizes the free time of Cloud Server to carry out data backup, improves backup efficiency.

Step S105, free time according to Cloud Server on the same day, upgrade Cloud Server free time section probability distribution.

This step is preferred steps, often crosses after one day, all needs the free time section obtaining the same day, then upgrade Cloud Server free time section probability distribution.Because every day Cloud Server free time section may be different, by adding up the free time section of every day, can improve further Cloud Server free time section probability distribution, make Cloud Server free time section closer to theoretical value.

Step S106, in the process sending newly-increased medical data, if the interrupt instruction of receiving, then interrupt sending and returning;

Wherein said step S101, specifically comprises:

Obtain the history operational data of Cloud Server, obtain the free time section of Cloud Server every day;

In units of sky, be divided into multiple chronomere by one day, add up the idle number of times of Cloud Server in each chronomere on a timeline;

Several chronomeres that the idle number of times selecting statistics to obtain is maximum, and several time points using the center time point of chronomere as the idle maximum probability of described Cloud Server.

In step S101, when adding up the probability distribution of free time section, first the history operational data according to Cloud Server is needed, obtain the free time section of the every day of Cloud Server, the time limit of described history operational data is not limit, and such as can adopt the data of a year and a day, then with the unit in sky, be divided into multiple chronomere by one day, such as 10 minutes is a chronomere, adds up the idle number of times of Cloud Server in each chronomere on a timeline, with reference to the Cloud Server shown in Figure 11 free time section probability distribution statistical schematic diagram, for history data statistics in 1 year in figure, always have 365 days, day1-day365 respectively in corresponding diagram, on a timeline, it within every 10 minutes, is a chronomere, always have 144 chronomeres, in figure, the time interval of every bar dotted line is a chronomere, and thick line segment table is shown in the free time section of Cloud Server in figure, then the idle number of times of Cloud Server in each chronomere is added up, only have Cloud Server in this chronomere once to add up for idle condition just counts always.After having added up, select several chronomeres (such as 10) that statistics number is maximum, then choose center time point several time points as the idle maximum probability of Cloud Server of each chronomere.

Wherein step S106 specifically comprises:

Increase newly in medical data process in transmission, after often sending a certain size data, send a sign position, and delete the upper data indicated between position and this transmission sign position of transmission;

After receiving interrupt instruction, send interrupt request to backup server;

Described backup server deletes the data after the sign position finally received.

In this step, carrying out, in the process backed up, likely there will be the situation that Cloud Server needs work, if receive the interrupt instruction of Cloud Server transmission in this process, then interrupting backup operation and return.During specific implementation, first increase newly in medical data process in transmission, after often sending a certain size data, send a sign position, and delete the upper data indicated between position and this transmission sign position of transmission; After receiving interrupt instruction, send interrupt request to backup server; Described backup server deletes the data after the sign position finally received, to ensure the data syn-chronization between Cloud Server and backup server.And in existing redundancy technique scheme, when generally occurring interrupting in backup procedure, need whole packet to resend once, compare and lose time and resource; Or the data between Cloud Server and backup server are carried out comparison repeatedly, determines the position sending data before interrupting, then continue to send, implementation procedure is also more loaded down with trivial details, also easily occurs identical data segment, judges that interruption position may be made mistakes.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.

Claims (9)

1. an implementation method for hospital's film-free cloud computing, is characterized in that, described method comprises:

When there being newly-increased medical data in medical system, described newly-increased medical data being synchronized to front server, being then synchronized to Cloud Server;

Subscription client sends inquiry request to system server, and system server searches the front server belonging to medical data to be checked, and sends the address of an inquiry password and corresponding front server to subscription client;

According to address and the inquiry password of described front server, subscription client and described front server connect, and the medical data that download user phone number is corresponding with inquiry password.

2. method as claimed in claim 1, it is characterized in that, described method also comprises: the medical data in Cloud Server is synchronized to backup server.

3. method as claimed in claim 1, it is characterized in that, described method also comprises:

Subscription client sends authorization message Xiang doctor's client, and described authorization message comprises the name of user, phone number and authorization code;

According to described authorization code, described doctor's client and system server connect;

The concrete front server that the medical data that described system server searches correspondence according to the name of user and phone number is positioned at, and send query statement to described doctor's client;

Described doctor's client connects with corresponding front server according to described query statement, downloads the thumbnail of the medical data corresponding with described user name and phone number and shows;

When thumbnail is wherein triggered, download from corresponding front server and be subject to trigger the corresponding original complete view data of thumbnail and show.

4. method as claimed in claim 1, is characterized in that, described when there being newly-increased medical data in medical system, described newly-increased medical data is synchronized to front server, is then synchronized to Cloud Server step, specifically comprises:

Front server at regular intervals interval refreshes a view, when finding there is newly-increased medical data in medical system, reading the patient code of newly-increased medical data, checking the store path of numbering and image file;

Store path according to described image file reads image file, and is stored in file;

Newly-increased image file whether is had, i.e. delta file every Preset Time run-down;

If there is delta file, then by the port that utilizes sockets socket to combine by the patient code of delta file and correspondence, check that numbering is synchronized to Cloud Server.

5. method as claimed in claim 4, is characterized in that, reads image file, and after being stored into the step in file, also comprise at the store path of described image file:

Image is carried out to the image file stored and splits combined treatment, and generate key;

And described by the patient code of delta file and correspondence, check that numbering is synchronized in Cloud Server step, by the patient code of the delta file after process and key and correspondence, check that numbering is synchronized to Cloud Server;

The wherein said image file to storing carries out image and splits combined treatment, and generates key step, specifically comprises:

Images all in image file are sorted according to time sequencing;

For every piece image, at Width stochastic generation w-1 point, at short transverse stochastic generation h-1 point, every width image splits into w × h grid block of pixels;

Take out a block of pixels and carry out the first round and choose at random from each width image in order, then at random from the remaining block of pixels of each width image take out a block of pixels and carry out second and take turns and choose, after w × h wheel is chosen, all block of pixels are selected complete like this, obtain a storage block matrix $B=\left[\begin{array}{cccc}{T}_1{X}_{11}& T_2{X}_{12}& ...& T_m{X}_{1m}\\ T_1{X}_{21}& T_2{X}_{22}& ...& T_m{X}_{2m}\\ ...& ...& T_z{X}_{ij}& ...\\ T_1{X}_{wh1}& T_2{X}_{wh2}& ...& T_m{X}_{whm}\end{array}\right],$ Wherein m is amount of images, T _zx _ijbe i-th take turns when choosing, from the X of z width image random selecting _ijindividual block of pixels, here X _ij∈ [1, w × h];

For each block of pixels in storage block matrix is multiplied by a luminance factor, obtain deformation matrix $C=\left[\begin{array}{cccc}{\mathrm{\α}}_{\mathrm{11}}{T}_1{X}_{11}& {\mathrm{\α}}_{\mathrm{12}}{T}_2{X}_{12}& ...& {\mathrm{\α}}_{\mathrm{1m}}{T}_m{X}_{1m}\\ {\mathrm{\α}}_{\mathrm{21}}{T}_1{X}_{21}& {\mathrm{\α}}_{22}{T}_2{X}_{22}& ...& {\mathrm{\α}}_{\mathrm{2m}}{T}_m{X}_{2m}\\ ...& ...& {\mathrm{\α}}_{ij}{T}_z{X}_{ij}& ...\\ {\mathrm{\α}}_{\mathrm{wh1}}{T}_1{X}_{wh1}& {\mathrm{\α}}_{wh2}{T}_2{X}_{wh2}& ...& {\mathrm{\α}}_{\mathrm{whm}}{T}_m{X}_{whm}\end{array}\right],$ Wherein ${\mathrm{\α}}_{ij}=\frac{\underset{k=1}{\overset{m}{\mathrm{\Σ}}}{X}_{ik}-X_{ij}}{w\×h}+\frac{\stackrel{\‾}{T_{i(j+1)}{X}_{i(j+1)}}+\stackrel{\‾}{T_{i(j-1)}{X}_{i(j-1)}}}{2D},$ Wherein function represent the average brightness value of block of pixels F, D is Benchmark brightness value, and the delta file after described process is described deformation matrix C;

Generate the key that current image file is corresponding, described key comprises w and h, also comprises cipher key matrix $E=\left[\begin{array}{cccc}{X}_{11}& X_{12}& ...& X_{1m}\\ X_{21}& X_{22}& ...& X_{2m}\\ ...& ...& X_{ij}& ...\\ X_{wh1}& X_{wh2}& ...& X_{whm}\end{array}\right].$

6. method as claimed in claim 5, it is characterized in that, from the upper left corner, 2 × 2 region segmentation are carried out to described cipher key matrix of drawing together, there are 4 elements in each region, according to clockwise direction, a position is rotated to the element in each region, if the columns of cipher key matrix is odd number, does not then process last column element, if the line number of cipher key matrix is odd number, then do not process last column element;

Described by process after delta file and key and correspondence patient code, check numbering be synchronized to Cloud Server step, specifically comprise:

A handshake is sent to Cloud Server;

After receiving feedback signal, by patient code with detect numbering and be sent to Cloud Server, and receive the memory address that Cloud Server returns;

Obtain last byte of patient code, when for odd number, the odd-line pixels block in deformation matrix C be synchronized to Cloud Server, then even rows block be synchronized to Cloud Server, finally by key synchronization to Cloud Server; When for even number, the even rows block in deformation matrix C is synchronized to Cloud Server, then odd-line pixels block is synchronized to Cloud Server, finally by key synchronization to Cloud Server.

7. method as claimed in claim 2, is characterized in that, described medical data in Cloud Server is synchronized to backup server step, specifically comprises:

Statistics Cloud Server in one day free time section probability distribution, determine several time points of the idle maximum probability of Cloud Server;

At each described time point, send query statement to Cloud Server;

After the free time receiving Cloud Server feedback confirms instruction, obtained the newly-increased medical data of current point in time apart from a upper Backup end moment;

Described newly-increased medical data is sent to backup server to preserve;

According to the free time of Cloud Server on the same day, upgrade Cloud Server free time section probability distribution;

In the process sending newly-increased medical data, if the interrupt instruction of receiving, then interrupt sending and returning;

Wherein said statistics Cloud Server in one day free time section probability distribution, determine several time point steps of the idle maximum probability of Cloud Server, specifically comprise:

Obtain the history operational data of Cloud Server, obtain the free time section of Cloud Server every day;

In units of sky, be divided into multiple chronomere by one day, add up the idle number of times of Cloud Server in each chronomere on a timeline;

Several chronomeres that the idle number of times selecting statistics to obtain is maximum, and several time points using the center time point of chronomere as the idle maximum probability of described Cloud Server;

In the wherein said process sending newly-increased medical data, if the interrupt instruction of receiving, then interrupt sending and returning step, specifically comprise:

Increase newly in medical data process in transmission, after often sending a certain size data, send a sign position, and delete the upper data indicated between position and this transmission sign position of transmission;

After receiving interrupt instruction, send interrupt request to backup server;

Described backup server deletes the data after the sign position finally received.

8. method as claimed in claim 1, it is characterized in that, patient information table is preserved in described front server, whenever completing user's medical data storage, generate login name and log in password and be sent to subscription client and system server, upgrade patient information table simultaneously, by described login name and log in password, subscription client and system server connect, subscription client just can send inquiry request to described system server, described patient information table at least comprises patient code's information, check the start-stop address of number information and medical data storage, described patient code's information at least comprises the name of user, phone number, described front server is that the phone number in patient information table divides into groups, and packet mode is: from last position of phone number to antepenulatimate, the parity according to bits per inch word is divided into groups, and always has 512 groups,

According to address and the inquiry password of described front server, subscription client and described front server connect, and the medical data step that download user phone number is corresponding with inquiry password, specifically comprise:

According to address and the inquiry password of described front server, subscription client and described front server connect;

Read rear 9 bit digital of phone number from back to front successively, and find corresponding grouping according to the parity of bits per inch word;

In current group, find corresponding phone number by the repeating query mode of searching;

Obtain the start-stop address of medical data storage according to the phone number found, and download the medical data between described start-stop address.

9. method as claimed in claim 8, it is characterized in that, described patient code's information is 16 bytes, and described inspection is numbered 4 bytes, described whenever completing user's medical data storage, and generate login name and log in password, specific implementation step is as follows:

Patient code's information of 16 bytes and the inspection number information of 4 bytes are formed 20 byte sequences, and in units of position, random alignment is carried out to described 20 byte sequences;

After random alignment 20 byte is divided into 5 byte section, obtains original group [r ₁, r ₂, r ₃, r ₄, r ₅], wherein r ₁~ r ₅represent described 5 byte section, each byte section has 4 bytes;

Calculate the valuation functions value of each byte section in original group wherein represent v byte of byte section, v=1,2,3 or 4;

Delete valuation functions value closest to 1 a byte section;

Then select in two byte section any position correspondence to exchange wherein, in two remaining byte section, select any position correspondence to exchange, obtain two generation group, described two generation group become by 4 byte section;

Finally optional two byte section composition login name from described two generation groups, two remaining byte section compositions log in password.