CN106558041A

CN106558041A - The band probe algorithm suppressed based on local in gel electrophoresiss digital picture

Info

Publication number: CN106558041A
Application number: CN201510627185.3A
Authority: CN
Inventors: 崔志刚
Original assignee: National Institute for Communicable Disease Control and Prevention of Chinese Center For Disease Control and Prevention
Current assignee: National Institute for Communicable Disease Control and Prevention of Chinese Center For Disease Control and Prevention
Priority date: 2015-09-29
Filing date: 2015-09-29
Publication date: 2017-04-05

Abstract

The present invention discloses a kind of method that swimming lane feature is automatically extracted in gel electrophoresiss digital picture.The characteristic curve of swimming lane column direction brightness flop is constructed first, then the average height and variance of statistic curve, according to the definition of extreme point, distance between extreme point relative altitude and extreme point calculates pre-maximum, these maximum points are filtered with average height and variance, the point of reservation is used as initialization bright wisp position, it is divided into some sections to characteristic curve further according to these positions, each section is suppressed respectively according to per section of average height and variance, subsequent recombinant, one new feature curve of construction, finally maximum is calculated on new feature curve, these maximum are used as final pillar location.The present invention can not only obtain the position of true band, and can filter out leave request band, moreover it is possible to improve the discrimination of low-light level band, greatly reduce the number of times of manpower intervention, improve work efficiency.

Description

The band probe algorithm suppressed based on local in gel electrophoresiss digital picture

Technical field

The invention belongs to technical field of image processing, is related to one kind and is visited from swimming lane figure automatically using zone suppression technique The method for surveying band, is processed to gel electrophoresiss digital picture.

Background technology

There is no in terms of the process of gel electrophoresiss image and find related document, but have part pertinent literature to occur in other Field.Chao Yang, Zengyou He and Weichuan Wu [Chao Yang, ZengyouHe, Weichuan Wu.Comparison of publ ic peak detection algorithms for MALDI mass Spectrometry data analysis, BMC Bioinformatics2009,10：4] by three extraction signal peaks, Signal smoothing process being carried out first, baseline then being calculated and is corrected, finally carry out peak detection, the method for peak detection is adopted Various methods, including threshold method, Peak Slope method, local extremum method, crestal line method etc..Hsein-Ping Kew with Do-Un Jeong [Hsein-Ping Kew, Do-Un Jeong.Variable Threshold Method for ECG R- Peak Detection, J MedSyst (2011) 35：1085-1094] using differential technique and Hilbert transform pairs electrocardiogram Curve carries out pretreatment, then constructs the peak value that multi thresholds method extracts signal.S.Satheeskumaran with M.Sabrigiriraj [S.Satheeskumaran, M.Sabrigiriraj.A New LMS Based Noise Removal and DWT Based R-peakDetection in ECG Signal for Biotelemetry Applications.Natl.Acad.Sci.Lett., 2014,37 (4)：341-349.] when electrocardiogram curve line is processed, first Curve is filtered with the sef-adapting filter based on least square, then Electrocardiographic peak is detected using wavelet analysis technology Value.

The content of the invention

The invention aims to automatic detection feature band in solving the problems, such as image, is that this present invention proposes one kind The method suppressed based on region detects the positional information of swimming lane band automatically, is applied to gel electrophoresiss Digital Image Processing On.The characteristics of this method has taken into full account gel electrophoresiss digital picture, that is, have regionality, and Regional Characteristics difference be big, because This takes the strategy for processing respectively, achieves good effect.

In order to complete the auto-Detection Technology of swimming lane band, it is proposed that based on the technical scheme that region suppresses, as follows：

Step one, structural features curve, and conscientious thick detection, and it is characterized curve zoning；

Step 2, the provincial characteristicss of calculated curve carry out local suppression；

Step 3, repressed characteristic curve is reconfigured, and is constructed new characteristic curve, is carried out accurately detecting.

The following detailed description of once realizing process：

Step S1, is swimming lane figure structural features curve, and the method for structure is the gray scale for counting every a line in swim-lane diagram picture Average, used as y values, using column direction position as x values, the two combines, and constructs a characteristic curve.

Step S2, using the extreme value point calculating method of discrete function, the extreme point position of detection feature curve.

Step S3, using different filterconditions, filters to initial extreme point, obtains first pillar location, such as According to the overall statistical nature (expecting and variance) of curve, remove the extreme point less than a certain height, according to adjacent extreme point Distance, removes the less extreme point of functional value.

Step S4, carries out region division to the characteristic curve of swimming lane according to first pillar location.

Step S5, the curve for counting each region are special, calculate which and expect and variance.

Step S6, suppresses to the region according to expectation and variance, i.e., all functional values deduct corresponding a certain height Value.

Step S7, the sectional curve after suppression is integrated, and obtains a new characteristic curve.

Step S8, execution step S2 and S3, obtain accurate pillar location.

Description of the drawings

Fig. 1 algorithm flow charts：Schematically illustrate the logical flow chart of the algorithm.

Fig. 2 algorithm process result figures：Illustrate the result of the algorithm.

Specific embodiment

Step one, swim-lane diagram picture isCharacteristic curve f (x, y)={ (M_i, i) }, i ∈ { 1,2 ..., n }

Step 2, the computational methods of discrete curve extreme point,Wherein g (i, j)=f ' (i, j)

Step 3, the expectation of curve and variance,

Filtercondition 1：F ＜ E+ α D, α ∈ [- 3,3], in calculating process, α=0.3 effect is preferable, it is proposed that be set as writing from memory Recognize value, while allowing user to change.

Filtercondition 2：Two neighboring point apart from d ＜ 5；

The interval of function, according to the maximum point set that step 3 is obtained, is divided into k sub-regions by step 4, be respectively [1, n₁], [n₁+ 1, n₂] ..., [n_k-1+ 1, n].

Step 5, calculates expectation and the variance of each section of function respectively according to the formula in step 3, is designated as E_k, D_k；

Step 6, the maximum suppressing method in the 1st region, h_l(i, j)=g_l(i, j)-(E_k+βD_k), β ∈ [- 3,3], Then by k h_l(i, j) is combined into new characteristic curve h (i, j), and in test, β=1 effect is preferable, it is proposed that be set to acquiescence Value, it is allowed to which user changes.

Step 7, calculates the position of band on h (i, j) according to step 2, three.

The above, the only specific implementation method of the method, but the protection domain invented is not limited to this, and it is any to be familiar with With understand the people of the technology in disclosed scope, it will be appreciated that or expect the conversion of some parameters, such as, α, β with And the distance between two maximum etc., should all cover within the scope of the including of the present invention.Therefore protection scope of the present invention should Should be defined by the protection domain of claims.

Claims

1. the method that the method for being suppressed using region detects the band in swimming lane automatically, its feature include following several respects：

First, quadratic probing technology is introduced, it is once thick to detect, it is once fine to detect；

Second, zone suppression technique is introduced, prominent real peak value suppresses false peaks；

3rd, zones of different is suppressed using different indexs.

2. the automatic detection method according to right 1, thick for the first time to detect the position for determining main band, and according to these positions Put and the characteristic curve of swimming lane is divided into into different regions.

3. the statistic of zones of different, according to right 1 and right 2, is calculated, and as the parameter that the segment signal suppresses.

4., according to right 1, zone suppression technique causes that the peak value in the region is more prominent, and low peak is less, this technology pair Low-light level band and background more effectively, can be made a distinction by the band of swimming lane end low-light level.

5. right 1 requires the automatic detection of the method suitable for swimming lane band gel electrophoresiss digital picture of described region suppression Process.