CN102764125A - Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume - Google Patents
Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume Download PDFInfo
- Publication number
- CN102764125A CN102764125A CN2012102527778A CN201210252777A CN102764125A CN 102764125 A CN102764125 A CN 102764125A CN 2012102527778 A CN2012102527778 A CN 2012102527778A CN 201210252777 A CN201210252777 A CN 201210252777A CN 102764125 A CN102764125 A CN 102764125A
- Authority
- CN
- China
- Prior art keywords
- fat
- threshold value
- cut apart
- volume
- apart
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume, which comprises the steps of: marking a navel position in a human body abdomen magnetic resonance image to obtain a coordinate of a marked point of the navel position; setting a distance range of a fat quantitative region and establishing an elliptic equation, and determining a fat quantitative region on a coronal view; carrying out gray value counting for a pixel signal of each layer of magnetic resonance image, establishing a histogram, and setting a threshold of segmenting fat; and segmenting the abdomen fat through progressive scanning according to the threshold, and accumulating volumes of segmented fats of all layers of magnetic resonance images for realizing fat volume quantization. According to the method, the labor intensity is lowered, the working efficiency is greatly increased, the fat volume is quantized more accurately, the subjective deviation is small, and the repeatability is good.
Description
Technical field
The invention belongs to the mr imaging technique field, concrete is a kind of semi-automatic quantification method based on MRI human abdomen fat volume.
Background technology
Accurately quantitatively human abdomen's fat has great significance on clinical medicine.At first, health in current obesity sickness rate serious harm of sharply rising, and the stomach fat volume quantitative provides a very important foundation for the obesity symptomatic diagnosis; In addition, abdominal surgery is very popularized at present, and the stomach fat volume quantitative can provide information such as the fat thickness and the subcutaneous tissue degree of depth to the surgeon accurately, helps surgical operation to implement more accurately.
Nuclear magnetic resonance has characteristics such as not damaged, soft tissue contrast height, image resolution ratio height, has been widely used in clinical diagnose at present.Utilize magnetic resonance Single Shot Fast Spin,Echo sequence that the human abdomen is carried out the T2 weighted imaging; Fatty tissue signal wherein is very high; Form striking contrast with other tissue; The high motion artifacts of this signal noise ratio (snr) of image is few in addition, is observation human abdomen fatty tissue, and quantitatively the stomach fat volume provides effective imaging technique means.
Traditional human abdomen's fat volume quantitative methods is found in MRI by clinical radiologist needs the quantitatively abdominal part position of fat.Specific practice is to confirm earlier to do fatty quantitative image aspect, manually describes the fatty tissue in each tomographic image then, the volume of the fat region that last each layer of accumulation calculating described, thus reach the stomach fat volume quantitative.
Though the method for above-mentioned existing hand drawing can reach the quantitative purpose of human abdomen's fat volume, there are many weak points in this method.First shortcoming of this method is a length consuming time, and efficient is low and labor intensity is big, and this is because in the process of individual stomach fat volume quantitative to each, need describe tens of layers image, and the rendering results of each layer of accumulation calculating.Second shortcoming of this method is can't the required scope of doing the quantitative abdominal part of fatty volume of explication, for example clinically usually with reference to umbilicus, with the resulting elliptoid quantitative scope as roughly of setpoint distance up and down in crown position.In addition, manually describe also to exist subjective deviation big, shortcomings such as repeatable difference.
Summary of the invention
Weak points such as length consuming time, the workload to existing in above-mentioned prior art human abdomen's fat volume quantitative approach is big, accuracy is restricted, repeatability is poor, the present invention proposes a kind of semi-automatic quantification method based on MRI human abdomen fat volume.The inventive method is by the graphical interfaces software tool that can show MRI; Show human abdomen's magnetic resonance t2 weighted image; Precise and high efficiency is chosen the quantitative position and the scope of stomach fat, re-uses that image segmentation algorithm is cut apart the stomach fat in the selected scope and volume calculation quantitatively semi-automatic.
The image segmentation algorithm that adopts among the present invention is based on the algorithm of Threshold Segmentation, and decision threshold has utilized the method for traditional picture element signal block diagram statistics.Innovation has proposed to line by line scan and has cut apart and judge the concrete partitioning scheme of cutting apart termination among the present invention.
The present invention propose a kind of based on MRI software to the semi-automatic quantitative methods of human abdomen fat volume, may further comprise the steps:
1) labelling is carried out in umbilicus position in human abdomen's MRI, obtain the coordinate of umbilicus position mark point;
2) set fat quantitatively the zone distance range and set up elliptic equation, confirm the quantitative zone of fat on the crown position;
3) picture element signal to each layer MRI carries out the gray value statistics, sets up block diagram, sets the threshold value of cutting apart fat;
4) stomach fat is cut apart through lining by line scan according to said threshold value;
5) fat-body of the cutting apart accumulation with each tomographic image adds, and realizes fatty volume quantitative.
Wherein, in the said step 3) with between background pixel peak in the said block diagram and the fatty pixel peak apart from the intensity level of 0.7 position at background pixel peak as the threshold value of cutting apart fat.
In the said step 4) each tomographic image of cutting apart in the scope is cut apart fat with the mode of lining by line scan from left to right; Separate the beginning when detecting first pixel that is higher than threshold value; The right neighbor is carried out threshold decision; Cut apart to the right if be higher than threshold value then continue, cut apart otherwise finish this row; Then the image of cutting apart in the scope is cut apart fat with the mode of lining by line scan from right to left; Separate the beginning when detecting first pixel that is higher than threshold value; Left side neighbor is carried out threshold decision, cut apart left, cut apart otherwise finish this row if be higher than threshold value then continue.
This method is at first utilized umbilicus position in the graphical interfaces software tool hand labeled abdominal part MRI, and reads the imager coordinate of this gauge point automatically; Manual set distance range and calculate the oval quantitatively regional extent on the crown position through setting up elliptic equation; Through the picture element signal gray value of each tomographic image being done statistics and being set up block diagram, set the threshold value of cutting apart fat then; Realize stomach fat automatization is cut apart through the ground mode of lining by line scan again; The most every layer the volume of cutting apart adds up, realizes fatty volume quantitative.
Advantage of the present invention is: the present invention compares with the method for traditional quantitative fatty volume of being described manually by clinical radiologist of mode; Most of work is all accomplished by computer automatically; Only need a small amount of manual operation process; Reduce labor intensity largely, improved work efficiency greatly.Tradition doctor manual plotting method is done fatty volume quantitative to a piece of data needs the time more than 20 minutes, and this method is done fatty volume quantitative to a piece of data and can within half a minute, be accomplished; In addition, compare with traditional method, it is more accurate that this method is done fatty volume quantitative, and the subjectivity deviation is little, favorable repeatability.
Description of drawings
Fig. 1 is the sketch map that the umbilicus place makes marks in the abdominal part MRI.
Fig. 2 is the oval quantitatively sketch map of scope area on human abdomen's coronalplane.
Fig. 3 utilizes the block diagram of pixel grey scale statistics to set the sketch map of cutting apart fatty threshold value.
Fig. 4 is the final sketch map of accomplishing the fatty cross-section bit image of cutting apart of certain one deck.
The specific embodiment
In conjunction with following specific embodiment and accompanying drawing, the present invention is done further detailed description.The process of embodiment of the present invention, condition, experimental technique etc. except that the following content of mentioning specially, are the universal knowledege and the common practise of this area, and the present invention does not have special limiting content.
Practical implementation step based on the semi-automatic quantification method of MRI human abdomen fat volume comprises as follows:
1) utilize the umbilicus position of graphical interfaces software tool in the abdominal part MRI manually to choose and be labeled as reference point; Utilize computer program to read the imager coordinate of this gauge point; Promptly with respect to the left side (right side) of imaging center position, the distance of (descending) is gone up in preceding (back).
2) with respect to the gauge point in the step 1, the distance value on 6 directions of given front, back, left, right, up, down to be confirming quantitatively scope of fat, and calculates the oval quantitatively zone on the crown position through the elliptic equation formula.
3) all pixels of each tomographic image of needs fat volume quantitative are done the gray value statistics and set up gray value statistics block diagram; Detect background pixel peak and fatty pixel peak in the block diagram, with between two peaks apart from the intensity level of 0.7 position at background pixel peak as the threshold value of cutting apart fat.
4) each tomographic image of cutting apart in the scope is cut apart fat with the mode of lining by line scan from left to right; Separate the beginning when detecting first pixel that is higher than threshold value; The right neighbor is carried out threshold decision, cut apart to the right, cut apart otherwise finish this row if be higher than threshold value then continue; Adopt same quadrat method then; Image to cutting apart in the scope is cut apart fat with the mode of lining by line scan from right to left; Separate the beginning when detecting first pixel that is higher than threshold value; Left side neighbor is carried out threshold decision, cut apart left, cut apart otherwise finish this row if be higher than threshold value then continue.
5) calculate the segmentation volume and the accumulation calculating of each layer fat, obtain final fatty volume quantitative values.
Following step-by-step instructions utilization the present invention is based on the semi-automatic quantification method of MRI human abdomen fat volume and in human abdomen T2 weighting MRI, carries out the quantitative detailed process of fatty volume.The magnetic resonance image data that present embodiment is gathered is human abdomen's t2 weighted image (like Fig. 1, Fig. 2, shown in Figure 4), gathers through GE 1.5T magnetic resonance imaging system, and the imaging sequence of being selected for use is the Single Shot Fast Spin,Echo sequence; Sweep parameter is: image resolution ratio (Matrix size): 512 * 512; The visual field (FOV): 480 mm * 480 mm, repetition time (TR): 916 ms, echo time (TE): 93ms; Bed thickness: 8 mm, the number of plies: 44.
(1) be reference point to abdominal part umbilicus position hand labeled
Utilizing the graphical interfaces software tool in human abdomen T2 weighting MRI, manually to choose the umbilicus position and be reference point, is the legend that the umbilicus place makes marks in the abdominal part MRI like Fig. 1.The graphic interface software that adopts in the present embodiment is Osirix 4.0.Obtain the imager coordinate of this gauge point then, promptly this gauge point is with respect to the left side (or right) of imaging center position, and the distance of (or down) is gone up in preceding (or back).Wherein, the imaging center position is meant magnet and gradient coil isocenter point in the magnetic resonance imaging system.
In the present embodiment, the coordinate of umbilicus gauge point is a left side (L): 1.6mm, and preceding (A): 84.9mm goes up (S): 54.2 mm.Coordinate like the umbilicus gauge point among Fig. 1 is: a left side: 1.6mm, preceding: 84.9mm, on: 54.2 mm.
(2) the quantitative scope of setting stomach fat volume
With respect to the point of step (1) institute labelling, through the distance value on 6 directions setting the front, back, left, right, up, down to set quantitatively scope of fat.Be preceding 3 cm through the setpoint distance value in the present embodiment, back 5 cm, left 28 cm, right 28 cm, last 3 cm, 10 cm confirm to cut apart scope down, and calculate oval quantitatively regional on the crown position through the elliptic equation formula.Like Fig. 2 is the oval quantitatively legend of scope area through the human abdomen's coronalplane that calculates.
(3) set the threshold value of cutting apart fat
All pixels of each tomographic image of needs fat volume quantitative are done the gray value statistics and set up gray value statistics block diagram; Detect background pixel peak and fatty pixel peak in the block diagram, with between two peaks apart from the intensity level of 0.7 position at background pixel peak as the threshold value of cutting apart fat.In block diagram, background signal is low, approaches zero position; In the t2 weighted image data, fat signal shows with high signal.The present invention is according to contrast in tissue information in a large amount of clinical data images, sums up the intensity level that obtains with 0,7 position as the suitable threshold value of cutting apart fat.Like Fig. 3 is to utilize the block diagram of pixel grey scale statistics to set the legend of cutting apart fatty threshold value.
The threshold value of cutting apart fat in the present embodiment is 809.
(4) fat in each tomographic image is cut apart and quantitatively
Each tomographic image in the area-of-interest is cut apart fat with the mode of lining by line scan from left to right; Separate the beginning when detecting first pixel that is higher than threshold value; The right neighbor is carried out threshold decision, cut apart to the right, cut apart otherwise finish this row if be higher than threshold value then continue; Adopt same quadrat method then; Image to cutting apart in the scope is cut apart fat with the mode of lining by line scan from right to left; Separate the beginning when detecting first pixel that is higher than threshold value; Left side neighbor is carried out threshold decision, cut apart left, cut apart otherwise finish this row if be higher than threshold value then continue.Wherein, area-of-interest is defined according to practical situation by the clinician.In the present embodiment, require to be fixed as 3cm in the upwards scope with respect to reference point according to the radiologist, fixed-site is 10cm downwards, and all the other 4 directions are freely debugged according to fat or thin the doing suitably of patient's physique by the doctor, with reference to figure 2, and Fig. 4 and Fig. 5.Among the present invention, cut apart fat and be meant the extracting section that belongs to fat in the image is come out.
Like Fig. 4 is the final cross-section bit image instance of certain one deck that fat is cut apart of accomplishing.
(5) calculate fatty volume total amount
Structure at all levels is cut apart the fat mass addition that obtains, can obtain fatty volume total amount,
Present embodiment fat volume quantitative results is 2045.1 milliliters.
The inventive method is by MRI software display image; In image, choose the position and the scope of required quantitative stomach fat; Utilize image segmentation algorithm that the stomach fat in the selected scope is cut apart and volume calculated again, realize the semi-automation of stomach fat volume quantitative.Compare with the method for the quantitative fatty volume of conventional manual, greatly reduce hand labor intensity, improved quantitative time efficiency and accuracy.
Protection content of the present invention is not limited to above embodiment.Under spirit that does not deviate from inventive concept and scope, variation and advantage that those skilled in the art can expect all are included among the present invention, and are protection domain with the appending claims.
Claims (3)
1. the semi-automatic quantification method based on MRI human abdomen fat volume is characterized in that, may further comprise the steps:
1) labelling is carried out in umbilicus position in human abdomen's MRI, obtain the coordinate of umbilicus position mark point;
2) set fat quantitatively the zone distance range and set up elliptic equation, confirm the quantitative zone of fat on the crown position;
3) picture element signal to each layer MRI carries out the gray value statistics, sets up block diagram, sets the threshold value of cutting apart fat;
4) stomach fat is cut apart through lining by line scan according to said threshold value;
5) fat-body of the cutting apart accumulation with each tomographic image adds, and realizes fatty volume quantitative.
2. method according to claim 1 is characterized in that, in the said step 3) with between background pixel peak in the said block diagram and the fatty pixel peak apart from the intensity level of 0.7 position at background pixel peak as the threshold value of cutting apart fat.
3. method according to claim 1; It is characterized in that; In the said step 4) each tomographic image of cutting apart in the scope is cut apart fat with the mode of lining by line scan from left to right, separate the beginning when detecting first pixel that is higher than threshold value, the right neighbor is carried out threshold decision; Cut apart to the right if be higher than threshold value then continue, cut apart otherwise finish this row; Then the image of cutting apart in the scope is cut apart fat with the mode of lining by line scan from right to left; Separate the beginning when detecting first pixel that is higher than threshold value; Left side neighbor is carried out threshold decision, cut apart left, cut apart otherwise finish this row if be higher than threshold value then continue.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012102527778A CN102764125A (en) | 2012-07-20 | 2012-07-20 | Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2012102527778A CN102764125A (en) | 2012-07-20 | 2012-07-20 | Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102764125A true CN102764125A (en) | 2012-11-07 |
Family
ID=47091870
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2012102527778A Pending CN102764125A (en) | 2012-07-20 | 2012-07-20 | Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102764125A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104143035A (en) * | 2013-05-10 | 2014-11-12 | 上海联影医疗科技有限公司 | Method for partitioning breast lesion |
| CN106846264A (en) * | 2016-12-29 | 2017-06-13 | 广西南宁灵康赛诺科生物科技有限公司 | A kind of quantitative analysis method for being suitable to primate laboratory animal abdominal subcutaneous fat |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030120147A1 (en) * | 2001-12-20 | 2003-06-26 | Siemens Aktiengesellschaft | Method for determining the total body fat content of a subject by analysis of MR images |
| US20100111390A1 (en) * | 2008-11-03 | 2010-05-06 | Seimans Corporate Research, Inc. | Robust Classification of Fat and Water Images from 1-point-Dixon Reconstructions |
| WO2011139232A1 (en) * | 2010-05-03 | 2011-11-10 | Nanyang Technological University | Automated identification of adipose tissue, and segmentation of subcutaneous and visceral abdominal adipose tissue |
-
2012
- 2012-07-20 CN CN2012102527778A patent/CN102764125A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030120147A1 (en) * | 2001-12-20 | 2003-06-26 | Siemens Aktiengesellschaft | Method for determining the total body fat content of a subject by analysis of MR images |
| US20100111390A1 (en) * | 2008-11-03 | 2010-05-06 | Seimans Corporate Research, Inc. | Robust Classification of Fat and Water Images from 1-point-Dixon Reconstructions |
| WO2011139232A1 (en) * | 2010-05-03 | 2011-11-10 | Nanyang Technological University | Automated identification of adipose tissue, and segmentation of subcutaneous and visceral abdominal adipose tissue |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104143035A (en) * | 2013-05-10 | 2014-11-12 | 上海联影医疗科技有限公司 | Method for partitioning breast lesion |
| CN104143035B (en) * | 2013-05-10 | 2016-01-20 | 上海联影医疗科技有限公司 | A kind of method splitting breast lesion |
| CN106846264A (en) * | 2016-12-29 | 2017-06-13 | 广西南宁灵康赛诺科生物科技有限公司 | A kind of quantitative analysis method for being suitable to primate laboratory animal abdominal subcutaneous fat |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP5562598B2 (en) | Image display apparatus, image display method, and magnetic resonance imaging apparatus | |
| US10925582B2 (en) | Method and device for selecting detection area, and elasticity detection system | |
| CN106600609B (en) | Spine segmentation method and system in medical image | |
| CN100561518C (en) | Self-adaptation medical image sequence interpolation method based on area-of-interest | |
| US8659603B2 (en) | System and method for center point trajectory mapping | |
| CN101796544A (en) | Voxel data visual | |
| CN106580368A (en) | Full-automatic ultrasonic diagnosis method | |
| CN101066211A (en) | User interface and method for displaying information in an ultrasound system | |
| CN101496723A (en) | Method for acquiring nerve navigation system imaging data | |
| CN104112265B (en) | Colon image dividing method and device | |
| CN1805711A (en) | Reference image display method for ultrasonography and ultrasonograph | |
| JP5233374B2 (en) | Medical image processing system | |
| JP4648709B2 (en) | Medical diagnostic imaging equipment | |
| CN106646301B (en) | A method of eliminating magnetic resonance imaging gradin-echo odd even phase difference | |
| CN102764125A (en) | Magnetic resonance image-based semi-automatic quantization method of human body abdominal fat volume | |
| CN107204045A (en) | Virtual endoscope system based on CT images | |
| CN104915989A (en) | CT image-based blood vessel three-dimensional segmentation method | |
| CN104523287A (en) | Stereoscopic orientation frame for PET, CT and MR image fusion of head and neck range | |
| Ogier et al. | 3D dynamic MRI for pelvis observation-a first step | |
| Chang et al. | Automatic selection of representative slice from cine-loops of real-time sonoelastography for classifying solid breast masses | |
| US20170199261A1 (en) | Method for assessing and improving data quality in fine structure analysis data | |
| CN204394548U (en) | For the stereotactic frame of incidence PET, CT and MR image co-registration | |
| Gevers et al. | Reproducibility of flow territories defined by plannings-free vessel encoded pseudo-continuous arterial spin labeling | |
| Negahdar et al. | Cine phase-contrast MRI measurement of CSF flow in the cervical spine: a pilot study in patients with spinal cord injury | |
| EP4339879A1 (en) | Anatomy masking for mri |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20121107 |