CN111839570A - Device for predicting risk of femoral head collapse and using method and application thereof - Google Patents

Abstract

The invention provides a device for predicting risk of femoral head collapse and a using method and application thereof. The device comprises an image processing unit, an analysis unit, a separation unit, a calculation unit and an output unit; the device can obtain a plurality of data of the femoral head of a case according to the hip joint CT image of the case, wherein the data comprises hardened bone volume, soft tissue volume and total head bone density, the relation between each data and the disease course time of the case is obtained simultaneously, the femoral head collapse risk of the case is calculated according to the total head bone density and the hardened bone volume of the case, and reliable medical data are provided for clinical treatment of femoral head necrosis.

Description

Device for predicting risk of femoral head collapse and using method and application thereof

Technical Field

The invention relates to the technical field of medical image identification, in particular to a device for predicting risk of femoral head collapse and a using method and application thereof.

Background

Femoral head necrosis (ONFH), also known as Avascular necrosis (AVN), refers to a local blood circulation disorder in the femoral head caused by various reasons, resulting in hypoxia, shrinkage and death of bone cells. Femoral head necrosis is a pathological evolution process, which initially occurs in a weight bearing area of a femoral head, and a necrotic bone trabecular structure is damaged under stress, namely, a microfracture and a subsequent repair process aiming at damaged bone tissues. The reasons for osteonecrosis are not eliminated, the repair is incomplete, and the process of injury and repair is continued, which results in structural change of femoral head, collapse and deformation of femoral head, arthritis and dysfunction.

The collapse of the femoral head means that the pressure bearing capacity begins to decline after the femoral head necrosis reaches a certain degree, and the femoral head necrosis generates the sum of countless tiny compression fractures. Collapse of femoral head necrosis means failure of the mechanical properties of the subchondral bone plate, which ultimately leads to dysfunction of the affected hip joint. The most common symptom of femoral head necrosis is pain, and the pain part is hip joint and near thigh and can radiate to knee. Pain can be caused by necrotic tissue-repairing inflammatory lesions or high pressure within inflammatory lesions, and can be manifested as persistent pain, resting pain. The collapse and deformation of the osteochondral leads to secondary arthritis or chronic traumatic pain at the attachment site of the muscle ligament around the hip joint. Limited hip activity, particularly limited rotational activity, or painful and short-lived claudication.

However, femoral head necrosis has various symptoms and signs besides pain, and the pain occurs at different times and in different degrees, but all of them are based on pathological evolution. While various clinical manifestations are not specific to femoral head necrosis, many hip joint disorders can occur, in other words, it is difficult to make a diagnosis of femoral head necrosis by subjective symptoms and clinical examination of the patient.

At present, the femoral head necrosis is diagnosed clinically mainly by means of imaging, imaging manifestations of the femoral head necrosis are related to the severity of pathological changes and pathological processes, and imaging diversification is determined by pathological changes. The image segmentation plays an important role in quantitative and qualitative analysis of medical images, and directly influences subsequent analysis and processing work of a computer-aided diagnosis system. The method for segmenting the CT image of the femoral head mainly comprises expert manual segmentation, computer interactive segmentation and full-automatic segmentation.

In order to further treat femoral head necrosis and provide a better diagnosis and treatment scheme for a patient, researchers develop different diagnosis and treatment systems according to a computer. For example, CN110444293A discloses a femoral head necrosis diagnosis and treatment system and a cloud service system, the system includes: the detection data acquisition module is used for acquiring detection data of the patient from the detection equipment; the treatment information acquisition module is used for acquiring the treatment information of the patient; the diagnosis and treatment information module is used for determining and storing diagnosis and treatment information of the user according to the diagnosis and treatment information and the detection data; the query module is used for querying corresponding case information from the stored diagnosis and treatment information according to a query instruction of a user; the communication module is used for providing an online communication platform of the case information for target people, and the technical problem that an existing information system of a hospital cannot meet the diagnosis and communication requirements of users is solved. However, the invention mainly focuses on obtaining diagnosis and treatment information and provides a patient with long-term diagnosis and treatment, but the invention does not have obvious help on the femoral head necrosis degree and collapse risk prediction, and if the condition development of the patient can be predicted according to the CT image, the invention can provide better help for a doctor to provide a more appropriate treatment scheme and can also help the patient to recover earlier.

Therefore, it is of high clinical significance to provide a device capable of predicting risk of femoral head collapse.

Disclosure of Invention

In view of the problems in the prior art, the present invention provides a device for predicting risk of collapse of femoral head and methods of use and application thereof. The device can help a doctor to predict the condition development of a patient and give the probability of the risk of femoral head collapse of the patient.

In order to achieve the purpose, the invention adopts the following technical scheme:

in a first aspect, the present invention provides an apparatus for predicting risk of collapse of a femoral head, comprising:

an image processing unit: the method comprises the steps of processing hip joint images of a case to obtain a hardened bone mask, a cancellous bone mask and a soft tissue mask of the hip joint, reconstructing a femoral head three-dimensional model, and then performing surface meshing and body meshing;

an analysis unit: calculating Polyline of the femoral head three-dimensional model, and then reconstructing a femoral head mask based on Polyline;

a separation unit: based on the hardened bone mask, the cancellous bone mask and the soft tissue mask of the hip joint and the femoral head mask, separating the hardened bone mask, the cancellous bone mask and the soft tissue mask inside the femoral head by using an intersection algorithm of Boolean operation and reconstructing a three-dimensional model;

a calculation unit: calculating a hardened bone volume and a soft tissue volume by using the three-dimensional model, reading a grid gray value of the femoral head three-dimensional model to obtain a full skull density, and then calculating a relation between the hardened bone volume, the soft tissue volume and the full skull density and the course time of the case;

an output unit: and inputting the sclerotic bone volume and the total head bone density as input variables into a Logistic regression model, and outputting the probability of the risk of femoral head collapse.

The device for predicting the risk of femoral head collapse provided by the invention can reconstruct a three-dimensional model of the femoral head of a patient according to the hip joint image of the patient, and can also obtain a hardened bone mask, a cancellous bone mask and a soft tissue mask of the femoral head and reconstruct the three-dimensional model; the computing unit can also obtain the total head bone density of the femoral head, give out the relation between the disease course time of a case and each item of data, on one hand, help a doctor to know the state of an illness of a patient, on the other hand, can also help the doctor to estimate the development trend of the state of an illness of the patient, and give out a reasonable treatment scheme. The device can avoid the subjectivity of doctors in experience judgment, can reduce the workload of the doctors, and provides reliable medical data for clinical treatment of femoral head necrosis.

As a preferred technical solution of the present invention, the Logistic regression model is represented by the following equation:

logit (p) ═ 5.137+0.001 × sclerotic bone volume-9.674 × total head bone density.

Preferably, the probability of risk of collapse of the femoral head is calculated by the following equation:

preferably, the total skull density is calculated by the following equation:

the medical image processing technology and the three-dimensional reconstruction implementation mode can use the existing commercial software to carry out image segmentation and three-dimensional modeling, and can also carry out recognition and processing based on computer languages. As a preferable technical scheme of the invention, the image processing unit comprises Mimics medical image processing software and/or 3-matic medical image processing software.

The image processed by the image processing unit may be a CT (Computed Tomography, abbreviated as CT, i.e., Computed Tomography) image, an MRI (Magnetic Resonance Imaging, abbreviated as MRI), a DR apparatus (digital X-ray Imaging system) image, and other clinically common Imaging apparatuses. For example, after a CT image is acquired, the image processing unit processes the CT image as image data to obtain a three-dimensional femoral head model and masks for the respective portions.

Preferably, the separation unit separates a hardened bone mask, a cancellous bone mask and a soft tissue mask inside the femoral head based on an intersection algorithm of boolean operations.

Preferably, the method for acquiring the grid gray-scale value comprises the following steps: and (3) after carrying out surface meshing and volume meshing on the femoral head three-dimensional model, importing the femoral head three-dimensional model into Mimics medical image processing software, and reading a grid gray value.

Preferably, the femoral head three-dimensional model is obtained by reconstructing a proximal femur three-dimensional model, and the specific method is as follows: and separating a mask of the proximal femur from the hip joint image based on gray value segmentation, reconstructing the proximal femur three-dimensional model, and separating the femoral head three-dimensional model from the proximal femur three-dimensional model by taking a femoral head base part plane as a boundary.

The device provided by the invention can also predict the relationship between the course time of a case and the sclerotic bone volume, soft tissue density and total head bone density.

Wherein the relationship between the sclerotic bone volume and the duration of the disease can be expressed as:

sclerotic bone volume of 126.424 × time of course (month) + 206.061;

the relationship between soft tissue volume and time of course can be expressed as:

soft tissue volume 0.005 times duration +0.565

The relationship between total bone density and time of course can be expressed as:

full head bone density-144.674 times duration of disease (month) +11538.205

In other words, after obtaining the hip joint image of the patient, the doctor can calculate the disease duration of the patient according to the image, obtain the relation between the disease duration of the case and each parameter, and simultaneously predict the risk of femoral head collapse or necrosis of the case, and remind the patient to pay attention to protection.

In order to facilitate the acquisition of the patient information, the device for predicting the risk of femoral head collapse further comprises a terminal management module, wherein the terminal management module is used for acquiring request information of a terminal and outputting information content corresponding to the request information. The terminal can be a doctor workstation, so that doctors can conveniently acquire treatment information of patients in real time, and authorized users can acquire requested information.

In a second aspect, the present invention provides a method of using the apparatus of the first aspect, the method comprising the steps of:

(1) importing a hip joint image of a case into an image processing unit, separating a mask at the proximal end of a femur based on gray value segmentation, reconstructing a three-dimensional model of the proximal end of the femur, and simultaneously separating a hardened bone mask, a cancellous bone mask and a soft tissue mask in the hip joint image;

(2) taking a femoral head base part plane as a boundary, separating a femoral head three-dimensional model from a femoral proximal end three-dimensional model, and performing surface meshing and body meshing;

(3) importing the femoral head three-dimensional model obtained in the step (2) into an analysis unit, calculating Polyline of the femoral head three-dimensional model, and then reconstructing a femoral head mask based on the Polyline;

(4) guiding the hardened bone mask, the cancellous bone mask and the soft tissue mask obtained in the step (1) and the femoral head mask obtained in the step (2) into a separation unit, separating the hardened bone mask, the cancellous bone mask and the soft tissue mask inside the femoral head, and reconstructing a three-dimensional model;

(5) importing the three-dimensional model obtained in the step (4) into a calculation unit, calculating a hardened bone volume and a soft tissue volume, obtaining a full skull density according to a grid gray value of the femoral head three-dimensional model obtained in the step (1), and then obtaining a relation between the hardened bone volume, the soft tissue volume and the full skull density and the course time of the case;

(6) and (4) leading the hardened bone volume and the full head bone density in the step (5) into an output unit, and outputting the probability of the risk of femoral head collapse.

In a third aspect, the present invention provides the use of an apparatus according to the first aspect for predicting risk of femoral head necrosis or collapse.

It should be noted that the foregoing units are merely divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

Compared with the prior art, the invention has at least the following beneficial effects:

the device for predicting the risk of femoral head collapse provided by the invention mainly comprises an image processing unit, an analysis unit, a separation unit, a calculation unit and an output unit, can obtain a plurality of data of a case femoral head according to a hip joint CT image of the case, wherein the data comprises hardened bone volume, soft tissue volume and total head bone density, and simultaneously obtains the relation between each data and the case course time, and simultaneously calculates the risk of femoral head collapse of the case according to the total head bone density and the hardened bone volume of the case, thereby providing reliable medical data for clinical treatment of femoral head necrosis, helping a doctor to know the state of illness of the patient, helping the doctor to predict the development trend of the state of the illness of the patient and providing a reasonable treatment scheme.

Drawings

Fig. 1 is a schematic view of a reconstructed three-dimensional model in example 2, in which a is a proximal femur three-dimensional model, B is a proximal femur three-dimensional model bounded by a plane of a femoral head base, and C is a femoral head three-dimensional model.

FIG. 2 is a diagram of a hardened bone mask, a soft tissue mask and a cancellous bone mask obtained by separating the unit in example 2, wherein the diagram A is the hardened bone mask, the diagram B is the soft tissue mask and the diagram C is the cancellous bone mask.

Fig. 3 is a model diagram of the soft tissue volume and the hardened bone volume measured separately using the intersection boolean operation in example 2.

FIG. 4 is a graph showing the results of ROC curve analysis with different parameters.

Detailed Description

The technical solutions of the present invention are further described in the following embodiments with reference to the drawings, but the following examples are only simple examples of the present invention and do not represent or limit the scope of the present invention, which is defined by the claims.

Example 1

This example 1 provides a method of using a device for predicting risk of collapse of a femoral head, comprising the steps of:

(1) introducing a case hip joint CT file (di com format) into an image separation unit, processing by using Mimics medical image processing software, separating a mask at the proximal end of the femur based on gray value segmentation, and reconstructing a three-dimensional model of the proximal end of the femur;

(2) separating the femoral head three-dimensional model from the femoral head near-end three-dimensional model by taking the femoral head base part plane as a boundary;

(3) calculating Polyline of a femoral head three-dimensional model in an analysis unit, and then reconstructing a femoral head mask based on Polyline;

(4) separating a hardened bone mask, a cancellous bone mask and a soft tissue mask in an original hip joint CT file on the basis of gray value segmentation by using an image separation unit;

(5) in the separation unit, an intersection algorithm of Boolean operation is applied to separate a hardened bone mask, a cancellous bone mask and a soft tissue mask inside the femoral head, and a three-dimensional model is respectively reconstructed;

(6) and (3) importing 3-matic medical image processing software after obtaining the femoral head three-dimensional model, and performing surface meshing and body meshing.

(7) And in a computing unit, importing the obtained femoral head mesh into Mimics medical image processing software to carry out material assignment. Automatically reading a grid gray value (Hounsfiled unit), calculating the total skull density according to the following formula, and obtaining the relation between the total skull density and the disease course time;

(8) in the calculation unit, measuring the volume of each part according to the reconstructed three-dimensional model of the sclerotic bone, the three-dimensional model of the cancellous bone and the three-dimensional model of the soft tissue, and obtaining the relationship between the volume of the sclerotic bone and the soft tissue and the duration of the disease;

(9) and inputting a Logistic regression model by taking the sclerotic bone volume and the total head bone density as input variables through an output unit, and outputting the probability of the risk of femoral head collapse.

Example 2

In this example, original CT images of 50 (28 males and 22 females, average 38.12 + -10.14 years old, 50 hips total) cases were reviewed according to the method provided in example 1.

Wherein the 50 cases are divided into 25 hips each in the collapsed group and the non-collapsed group; in addition, 5 (5 hip) healthy volunteers were recruited as healthy controls.

(1) Importing the original CT image of each case into Mimics medical image processing software for morphological analysis, separating out a mask at the proximal end of the femur based on gray value segmentation, and reconstructing a three-dimensional model of the proximal end of the femur; separating the femoral head three-dimensional model from the femoral head near-end three-dimensional model by taking the femoral head base part plane as a boundary;

the reconstructed three-dimensional model of a case is given in fig. 1, where a is the proximal femur three-dimensional model and in B the plane of the femoral head base is shown and C is the femoral head three-dimensional model.

(2) Segmenting a femoral head model according to the gray value, separating masks of the above 3 structures in the femoral head according to gray value intervals of the hardened bone, the cancellous bone and the soft tissue, and measuring the volume after reconstructing a three-dimensional model;

as shown in FIG. 2, a hardened bone mask (A), a soft tissue mask (B) and a cancellous bone mask (C) are shown in a case;

as shown in fig. 3, where the intersection algorithm of boolean operations is applied, the soft tissue volume and the hardened bone volume are measured separately.

(3) Carrying out mesh division and material assignment on the whole skull to measure and calculate the whole skull density, wherein the whole skull density is calculated by the following equation:

(4) analyzing the linear correlation relationship between the different bone structure volumes and the total bone density and the disease course time and the correlation between the different bone structure volumes and the total bone density and the collapse fate;

wherein the relationship between the sclerotic bone volume and the duration of the disease can be expressed as:

sclerotic bone volume of 126.424 × time of course (month) + 206.061;

the relationship between soft tissue volume and time of course can be expressed as:

soft tissue volume 0.005 × disease duration + 0.565;

the relationship between total bone density and time of course can be expressed as:

total skull density-144.674 × duration of disease (month) + 11538.205.

Measuring and calculating the risk of femoral head collapse according to a Logistic regression model, wherein the Logistic regression model is represented by the following equation:

logit (p) ═ 5.137+0.001 × sclerotic bone volume-9.674 × total head bone density;

the probability of risk of femoral head collapse is calculated by the following equation:

the accuracy of the above index in predicting collapse was analyzed using ROC curves, as shown in fig. 4, where AUC of the whole skull density curve is 0.323; AUC of the sclerotic bone volume curve is 0.600; AUC for soft tissue volume is 0.573; AUC for the sclerotic bone and soft tissue volume curve is 0.654; the AUC of the Logistic regression model is 0.765; from the AUC values, the Logistic regression model is most accurate.

And (4) analyzing results:

through the analysis of the femoral head morphology of 50 cases and 5 healthy controls, the femoral head with femoral head necrosis is found to have discontinuous sclerotic bone boundaries, and the soft tissue volume is reduced;

at the same time, sclerotic bone volume and total head bone density increase with the course of the disease, while soft tissue volume decreases (P < 0.05).

In summary, the device for predicting the risk of femoral head collapse provided by the invention can obtain multiple data of a case femoral head according to a hip joint CT image of the case, including the hardened bone volume, the soft tissue volume and the total head bone density, obtain the relationship between each data and the case course time, calculate the risk of femoral head collapse of the case according to the total head bone density and the hardened bone volume of the case, and provide reliable medical data for the clinical treatment of femoral head necrosis.

The applicant declares that the above description is only a specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and it should be understood by those skilled in the art that any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are within the scope and disclosure of the present invention.

Claims (10)

1. An apparatus for predicting risk of collapse of a femoral head, the apparatus comprising:

an image processing unit: the method comprises the steps of processing hip joint images of a case to obtain a hardened bone mask, a cancellous bone mask and a soft tissue mask of the hip joint, reconstructing a femoral head three-dimensional model, and then performing surface meshing and body meshing;

an analysis unit: calculating Polyline of the femoral head three-dimensional model, and then reconstructing a femoral head mask based on Polyline;

a separation unit: based on the hardened bone mask, the cancellous bone mask and the soft tissue mask of the hip joint and the femoral head mask, separating the hardened bone mask, the cancellous bone mask and the soft tissue mask inside the femoral head by using an intersection algorithm of Boolean operation and reconstructing a three-dimensional model;

a calculation unit: calculating a hardened bone volume and a soft tissue volume by using the three-dimensional model, reading a grid gray value of the femoral head three-dimensional model to obtain a full skull density, and then calculating a relation between the hardened bone volume, the soft tissue volume and the full skull density and the course time of the case;

an output unit: and inputting the sclerotic bone volume and the total head bone density as input variables into a Logistic regression model, and outputting the probability of the risk of femoral head collapse.

2. The apparatus of claim 1, wherein the Logistic regression model is expressed by the following equation:

logit (p) ═ 5.137+0.001 × sclerotic bone volume-9.674 × total head bone density.

3. The apparatus of claim 1 or 2, wherein the probability of risk of femoral head collapse is calculated by the equation:

4. the apparatus of claim 1, wherein the total skull density is calculated by the equation:

5. the device according to claim 1 or 4, wherein the grid gray scale value is obtained by:

and (3) after carrying out surface meshing and volume meshing on the femoral head three-dimensional model, importing the femoral head three-dimensional model into Mimics medical image processing software, and reading a grid gray value.

6. The apparatus according to claim 1, wherein the image processing unit comprises a mics medical image processing software and/or a 3-matic medical image processing software.

7. The apparatus of claim 1, wherein the separation unit separates a hardened bone mask, a cancellous bone mask, and a soft tissue mask inside the femoral head based on an intersection algorithm of boolean operations.

8. The apparatus of claim 1, wherein the three-dimensional model of the femoral head is obtained by reconstructing a three-dimensional model of a proximal femur by:

and separating a mask of the proximal femur from the hip joint image based on gray value segmentation, reconstructing the proximal femur three-dimensional model, and separating the femoral head three-dimensional model from the proximal femur three-dimensional model by taking a femoral head base part plane as a boundary.

9. Use of a device according to any of claims 1 to 8, characterized in that it comprises the following steps:

(1) importing a hip joint image of a case into an image processing unit, separating a mask at the proximal end of a femur based on gray value segmentation, reconstructing a three-dimensional model of the proximal end of the femur, and simultaneously separating a hardened bone mask, a cancellous bone mask and a soft tissue mask in the hip joint image;

(2) taking a femoral head base part plane as a boundary, separating a femoral head three-dimensional model from a femoral proximal end three-dimensional model, and performing surface meshing and body meshing;

(3) importing the femoral head three-dimensional model obtained in the step (2) into an analysis unit, calculating Polyline of the femoral head three-dimensional model, and then reconstructing a femoral head mask based on the Polyline;

(4) guiding the hardened bone mask, the cancellous bone mask and the soft tissue mask obtained in the step (1) and the femoral head mask obtained in the step (2) into a separation unit, separating the hardened bone mask, the cancellous bone mask and the soft tissue mask inside the femoral head, and reconstructing a three-dimensional model;

(5) importing the three-dimensional model obtained in the step (4) into a calculation unit, calculating a hardened bone volume and a soft tissue volume, obtaining a full skull density according to a grid gray value of the femoral head three-dimensional model obtained in the step (1), and then obtaining a relation between the hardened bone volume, the soft tissue volume and the full skull density and the course time of the case;

(6) and (4) leading the hardened bone volume and the full head bone density in the step (5) into an output unit, and outputting the probability of the risk of femoral head collapse.

10. Use of a device according to any one of claims 1 to 8 for predicting the risk of femoral head necrosis or collapse.

Images