Children bone age comprehensive evaluation method based on deep learning
By using a deep learning-based comprehensive bone age assessment method and matching analysis with skeletal assessment atlases, the problem of inaccurate bone age assessment results in children in existing technologies has been solved, and more comprehensive assessment results have been achieved.
Patent Information
- Application Number
- CN202510948198.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-10
- Publication Date
- 2025-10-17
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing methods for assessing children's bone age are unable to formulate corresponding assessment standards based on the children's actual age, and the analysis of bone imaging data is not comprehensive enough, resulting in a lack of comprehensiveness and accuracy in the assessment results.
Using a deep learning-based approach, the method acquires children's actual age and skeletal imaging data, and performs matching analysis using a skeletal assessment atlas, including assessment factor areas, abnormal factor areas, and assessment result areas, to generate a comprehensive assessment result of children's bone age.
This allows for the development of corresponding assessment standards based on children's actual age, improving the comprehensiveness and accuracy of children's bone age assessment.
Smart Images
Figure CN120809210A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of bone age assessment, and particularly relates to a child bone age comprehensive assessment method based on deep learning. BACKGROUND
[0002] Child growth and development monitoring is an important research direction in the field of pediatric medicine. Bone age, as a key indicator reflecting the maturity of child physiological development, plays an irreplaceable role in the diagnosis, treatment effect evaluation and adult height prediction of diseases such as growth hormone deficiency, precocious puberty and thyroid function abnormalities. Bone age is determined by characteristics such as epiphyseal development state, ossification center appearance sequence and epiphyseal line closure degree in skeletal X-ray images. The evaluation result directly affects the accuracy of clinical decision-making. With the rising incidence of child growth and development related diseases, higher requirements are put forward for the accuracy and efficiency of bone age assessment in clinical practice.
[0003] The child bone age comprehensive assessment method in the related art often cannot formulate corresponding bone age assessment standards according to the actual age of children, and the analysis of child skeletal image data is also not comprehensive enough, resulting in a lack of comprehensiveness and accuracy of child bone age assessment results, which needs to be improved. SUMMARY
[0004] The purpose of the embodiments of the present application is to provide a child bone age comprehensive assessment method based on deep learning to improve the problem that the child bone age comprehensive assessment method in the related art often cannot formulate corresponding bone age assessment standards according to the actual age of children, and the analysis of child skeletal image data is also not comprehensive enough, resulting in a lack of comprehensiveness and accuracy of child bone age assessment results.
[0005] The present application provides a child bone age comprehensive assessment method based on deep learning, comprising:
[0006] Step S1: obtaining the actual age of a target child, obtaining a skeletal evaluation atlas corresponding to the target child according to the actual age of the child, the skeletal evaluation atlas comprising an evaluation factor area, an abnormal factor area and an evaluation result area; wherein the evaluation factor area comprises skeletal standard factors and factor evaluation weights corresponding to the skeletal standard factors; the abnormal factor area comprises abnormal characteristic factors and abnormal diagnosis factors;
[0007] Step S2: obtaining skeletal image data of the target child, and obtaining skeletal monitoring factors of the target child according to the skeletal image data;
[0008] Step S3: inputting the skeletal monitoring factors of the target child as atlas input conditions into the skeletal evaluation atlas, and matching and analyzing the skeletal monitoring factors in the atlas input conditions with the skeletal standard factors in the evaluation factor area to obtain abnormal monitoring factors and abnormal factor parameters of the target child;
[0009] Step S4: obtaining an abnormal impact bone age of the target child according to the factor evaluation weight in the evaluation factor area and the abnormal factor parameter of the abnormal monitoring factor, and obtaining a child evaluation bone age of the target child according to the abnormal impact bone age and the actual age of the child;
[0010] Step S5: transmitting the abnormal monitoring factor to the abnormal factor area; performing matching analysis on the abnormal monitoring factor and the abnormal characteristic factor in the abnormal factor area to obtain a target characteristic factor and a target abnormal diagnosis factor corresponding to the target child; and obtaining a bone age evaluation symptom of the target child according to the target abnormal diagnosis factor;
[0011] Step S6: transmitting the child evaluation bone age and the bone age evaluation symptom to the evaluation result area, and obtaining a bone age comprehensive evaluation result of the target child according to the evaluation result area.
[0012] Preferably, the evaluation factor area includes a bone standard factor and a factor evaluation weight corresponding to the bone standard factor, and includes the following steps:
[0013] obtaining standard bone data of the target child according to the actual age of the child;
[0014] performing bone feature extraction according to the standard bone data to obtain a standard bone characteristic corresponding to the target child;
[0015] and generating a bone standard factor corresponding to the target child and a standard factor parameter interval corresponding to the bone standard factor according to the standard bone characteristic, the bone standard factor corresponding to the standard bone characteristic;
[0016] setting a factor evaluation weight of the bone standard factor corresponding to the target child according to the actual age of the child.
[0017] Preferably, the abnormal factor area includes an abnormal characteristic factor and an abnormal diagnosis factor, the abnormal characteristic factor and the abnormal diagnosis factor being connected with a factor diagnosis track, and including the following steps:
[0018] obtaining historical bone abnormal data corresponding to the actual age of the child, the historical bone abnormal data including bone abnormal monitoring data and a bone diagnosis result corresponding to the bone abnormal monitoring data;
[0019] performing bone feature extraction on the bone abnormal monitoring data to obtain a bone abnormal characteristic corresponding to the target child; and generating an abnormal characteristic factor and an abnormal factor parameter corresponding to the abnormal characteristic factor according to the bone abnormal characteristic, the abnormal characteristic factor corresponding to the bone abnormal characteristic;
[0020] The bone diagnosis result is subjected to bone feature extraction to obtain a bone diagnosis feature corresponding to the target child; and an abnormal diagnosis factor and a diagnosis factor parameter corresponding to the abnormal diagnosis factor are generated according to the bone diagnosis feature, the abnormal diagnosis factor corresponding to the bone diagnosis feature;
[0021] The abnormal feature factor is connected with the abnormal diagnosis factor corresponding to the abnormal feature factor to obtain a factor diagnosis track between the abnormal feature factor and the abnormal diagnosis factor.
[0022] Preferably, the bone image data of the target child is acquired, and a bone monitoring factor of the target child is obtained according to the bone image data, including:
[0023] The bone image data is subjected to bone feature extraction to obtain a bone monitoring feature of the target child;
[0024] And a bone monitoring factor of the target child and a monitoring factor parameter corresponding to the bone monitoring factor are generated according to the bone monitoring feature, the bone monitoring factor corresponding to the bone monitoring feature.
[0025] Preferably, step S3 includes:
[0026] The bone monitoring factor in the atlas input condition is subjected to factor matching with a bone standard factor in the evaluation factor area to obtain a bone standard factor in the evaluation factor area consistent with the bone monitoring factor, and the bone standard factor is recorded as a target standard factor;
[0027] The monitoring factor parameter of the bone monitoring factor is compared with a standard factor parameter interval of the target standard factor;
[0028] If the monitoring factor parameter of the bone monitoring factor is not in the standard factor parameter interval of the target standard factor, the bone monitoring factor is recorded as an abnormal monitoring factor, and an interval center parameter of the standard factor parameter interval is acquired;
[0029] An abnormal factor parameter of the abnormal monitoring factor is obtained according to the interval center parameter and the monitoring factor parameter.
[0030] Preferably, an abnormal influence bone age of the target child is obtained according to a factor evaluation weight in the evaluation factor area and the abnormal factor parameter of the abnormal monitoring factor, including:
[0031] The abnormal factor parameter of the abnormal monitoring factor is weighted and accumulated with the factor evaluation weight of the abnormal monitoring factor to obtain a bone age influence coefficient of the target child;
[0032] An abnormal influence bone age of the target child is obtained according to the bone age influence coefficient and an actual age of the child.
[0033] Preferably, the abnormal monitoring factor is matched with the abnormal characteristic factor in the abnormal factor area to obtain a target characteristic factor corresponding to the target child and a target abnormal diagnosis factor, including:
[0034] The abnormal monitoring factor is matched with the abnormal characteristic factor in the abnormal factor area to obtain an abnormal characteristic factor in the abnormal factor area consistent with the abnormal monitoring factor and the abnormal characteristic factor, and the abnormal characteristic factor is recorded as a target characteristic factor;
[0035] A target factor connection track of the target characteristic factor is obtained, and a target abnormal diagnosis factor corresponding to the target characteristic factor is obtained according to the target factor connection track.
[0036] Preferably, a bone age evaluation symptom of the target child is obtained according to the target abnormal diagnosis factor, including:
[0037] A bone age evaluation symptom of the target child is obtained according to a bone diagnosis characteristic corresponding to the target abnormal diagnosis factor.
[0038] Preferably, a bone age comprehensive evaluation result of the target child is obtained according to the evaluation result area, including:
[0039] The bone age comprehensive evaluation result includes a child evaluation bone age and a bone age evaluation symptom of the target child.
[0040] In summary, the beneficial effects of the present application are: the present application obtains a target child corresponding bone evaluation graph according to the actual age of the target child, the bone evaluation graph includes an evaluation factor area, an abnormal factor area and an evaluation result area; wherein the evaluation factor area includes a bone standard factor and a factor evaluation weight corresponding to the bone standard factor; the abnormal factor area includes an abnormal characteristic factor and an abnormal diagnosis factor, the abnormal characteristic factor is connected with the abnormal diagnosis factor through a factor diagnosis track; in addition, the bone monitoring factor of the target child is obtained according to the bone image data of the target child; the bone monitoring factor of the target child is input into the bone evaluation graph as a graph input condition, the bone monitoring factor in the graph input condition is matched and analyzed with the bone standard factor in the evaluation factor area, the abnormal monitoring factor and the abnormal factor parameter of the target child are obtained; and the abnormal influence bone age of the target child is obtained according to the factor evaluation weight in the evaluation factor area and the abnormal factor parameter of the abnormal monitoring factor, and the child evaluation bone age of the target child is obtained according to the abnormal influence bone age and the actual age of the child; the abnormal monitoring factor is transmitted to the abnormal factor area; the abnormal monitoring factor is matched and analyzed with the abnormal characteristic factor, the target characteristic factor and the target abnormal diagnosis factor corresponding to the target child are obtained; and the bone age evaluation symptom of the target child is obtained according to the abnormal diagnosis factor; finally, the child evaluation bone age and the bone age evaluation symptom are transmitted to the evaluation result area, the bone age comprehensive evaluation result of the target child is obtained according to the evaluation result area, the corresponding evaluation standard, i.e. the bone evaluation graph, is formulated according to the actual age of the child, and the comprehensiveness and accuracy of the child bone age evaluation result are improved. BRIEF DESCRIPTION OF DRAWINGS
[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application, some of the drawings in the embodiments of the present application will be briefly described below. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be considered as limiting the scope of the present application.
[0042] Figure 1 A flowchart of a child bone age comprehensive evaluation method based on deep learning provided by the present application. DETAILED DESCRIPTION
[0043] The embodiments of the present application will be further described below, but the embodiments of the present application are not limited thereto. Figure 1 The embodiments of the present application will be further described below, but the embodiments of the present application are not limited thereto.
[0044] Referring to Figure 1 Fig. 1, a flowchart of a child bone age comprehensive evaluation method based on deep learning provided by the present application.
[0045] A child bone age comprehensive evaluation method based on deep learning, comprising:
[0046] Step S1: Obtain the actual age of the target child, and obtain the corresponding bone evaluation atlas of the target child according to the actual age of the child. The bone evaluation atlas includes an evaluation factor area, an abnormal factor area, and an evaluation result area. The evaluation factor area includes a bone standard factor and a factor evaluation weight corresponding to the bone standard factor. The abnormal factor area includes an abnormal characteristic factor and an abnormal diagnosis factor.
[0047] Step S2: Obtain the bone image data of the target child, and obtain the bone monitoring factor of the target child according to the bone image data.
[0048] Step S3: Input the bone monitoring factor of the target child into the bone evaluation atlas as a graph input condition, and perform matching analysis on the bone monitoring factor in the graph input condition and the bone standard factor in the evaluation factor area to obtain an abnormal monitoring factor and an abnormal factor parameter of the target child.
[0049] Step S4: Obtain the abnormal impact bone age of the target child according to the factor evaluation weight in the evaluation factor area and the abnormal factor parameter of the abnormal monitoring factor, and obtain the child evaluation bone age of the target child according to the abnormal impact bone age and the actual age of the child.
[0050] Step S5: Transmit the abnormal monitoring factor to the abnormal factor area, perform matching analysis on the abnormal monitoring factor and the abnormal characteristic factor in the abnormal factor area to obtain a target characteristic factor and a target abnormal diagnosis factor corresponding to the target child, and obtain a bone age evaluation symptom of the target child according to the target abnormal diagnosis factor.
[0051] Step S6: Transmit the child evaluation bone age and the bone age evaluation symptom to the evaluation result area, and obtain the bone age comprehensive evaluation result of the target child according to the evaluation result area.
[0052] The evaluation factor area includes a bone standard factor and a factor evaluation weight corresponding to the bone standard factor, including:
[0053] Obtain the standard bone data of the target child according to the actual age of the child.
[0054] Perform bone feature extraction according to the standard bone data to obtain a standard bone feature corresponding to the target child.
[0055] Generate a bone standard factor corresponding to the target child and a standard factor parameter interval corresponding to the bone standard factor according to the standard bone feature. The bone standard factor corresponds to the standard bone feature.
[0056] Set the factor evaluation weight of the bone standard factor corresponding to the target child according to the actual age of the child.
[0057] In some embodiments, the standard bone data of the target child can be obtained through historical bone evaluation data of the actual age of the child or a large database; when the standard bone data is subjected to bone feature extraction, the bone features consistent with the preset bone features can be extracted based on the preset bone features, wherein the preset bone features are preset according to the actual age of the child, and the bone age evaluation bone parameters corresponding to the actual age of the child are specifically preset, for example, the preset bone features can be "ulna and radius", "metacarpal bone", "phalangeal bone", etc., and when the standard bone data is subjected to bone feature extraction, the extracted standard bone features also include standard bone feature parameters.
[0058] In addition, each standard bone feature corresponds to a bone standard factor, and the standard factor parameter interval is set according to the standard bone feature parameter of the standard bone feature; the factor evaluation weight is the bone factor feature weight value for judging the bone age of the child according to the actual age of the child.
[0059] The abnormal factor area includes abnormal feature factors and abnormal diagnosis factors, and the abnormal feature factors and the abnormal diagnosis factors are connected with factor diagnosis tracks, including:
[0060] The historical bone abnormal data corresponding to the actual age of the child is obtained, and the historical bone abnormal data includes bone abnormal monitoring data and bone diagnosis results corresponding to the bone abnormal monitoring data;
[0061] The bone abnormal monitoring data is subjected to bone feature extraction to obtain bone abnormal features corresponding to the target child; and the abnormal feature factors and the abnormal feature factor corresponding abnormal diagnosis factor parameters are generated according to the bone abnormal features, and the abnormal feature factors correspond to the bone abnormal features;
[0062] The bone diagnosis results are subjected to bone feature extraction to obtain bone diagnosis features corresponding to the target child; and the abnormal diagnosis factors and the abnormal diagnosis factor corresponding diagnosis factor parameters are generated according to the bone diagnosis features, and the abnormal diagnosis factors correspond to the bone diagnosis features;
[0063] The abnormal feature factors are connected with the abnormal diagnosis factors corresponding to the abnormal feature factors to obtain the factor diagnosis tracks between the abnormal feature factors and the abnormal diagnosis factors.
[0064] In some embodiments, the bone abnormal monitoring data corresponds to the bone diagnosis results; when the bone abnormal monitoring data is subjected to bone feature extraction, the bone abnormal monitoring data is subjected to bone feature extraction based on the preset bone features, and the feature parameters corresponding to the bone abnormal features are extracted, one feature parameter of the bone abnormal features corresponds to one abnormal feature factor, that is, the bone abnormal features with different feature parameters correspond to different abnormal feature factors, and the abnormal factor parameters are obtained according to the feature parameters of the bone abnormal features;
[0065] In some embodiments, when the skeletal diagnosis result is subjected to skeletal feature extraction, the skeletal diagnosis result is subjected to skeletal feature extraction based on preset skeletal features, and a feature parameter corresponding to the skeletal diagnosis feature is extracted, one feature parameter of the skeletal diagnosis feature corresponds to one abnormal diagnosis factor, that is, different skeletal diagnosis features corresponding to different abnormal diagnosis factors, and the diagnosis factor parameter is obtained according to the feature parameter of the skeletal diagnosis feature;
[0066] It should be noted that the abnormal feature factor and the abnormal diagnosis factor correspond to each other, that is, the abnormal feature factor corresponds to the abnormal diagnosis factor, and the abnormal diagnosis factor also corresponds to the abnormal feature factor, and the factors in the abnormal factor area can be both the abnormal feature factor and the abnormal diagnosis factor.
[0067] In some embodiments, the factor diagnosis track is a one-way track, that is, the abnormal feature factor points to the abnormal diagnosis factor corresponding thereto, and the factors with a corresponding relationship in the abnormal factor area are connected through the factor diagnosis track.
[0068] Obtaining skeletal image data of a target child, obtaining skeletal monitoring factors of the target child according to the skeletal image data, comprising:
[0069] Performing skeletal feature extraction on the skeletal image data to obtain skeletal monitoring features of the target child;
[0070] And generating skeletal monitoring factors of the target child and monitoring factor parameters corresponding to the skeletal monitoring factors according to the skeletal monitoring features, the skeletal monitoring factors corresponding to the skeletal monitoring features.
[0071] In some embodiments, the skeletal image data corresponds to the skeletal monitoring features; when the skeletal image data is subjected to skeletal feature extraction, the skeletal image data is subjected to skeletal feature extraction based on preset skeletal features, and a feature parameter corresponding to the skeletal monitoring features is extracted, one feature parameter of the skeletal monitoring features corresponds to one skeletal monitoring factor, that is, different skeletal abnormal features corresponding to different skeletal monitoring factors, and the monitoring factor parameter is obtained according to the feature parameter of the skeletal monitoring features.
[0072] Step S3, comprising:
[0073] Matching the skeletal monitoring factors in the atlas input condition with the skeletal standard factors in the evaluation factor area to obtain the skeletal standard factors in the evaluation factor area consistent with the skeletal monitoring factors, and recording the skeletal standard factors as target standard factors;
[0074] Comparing the monitoring factor parameters of the skeletal monitoring factors with the standard factor parameter interval of the target standard factors;
[0075] If the monitoring factor parameter of the skeletal monitoring factor is not within the standard factor parameter interval of the target standard factor, the skeletal monitoring factor is recorded as an abnormal monitoring factor, and an interval center parameter of the standard factor parameter interval is obtained;
[0076] An abnormal factor parameter of the abnormal monitoring factor is obtained according to the interval center parameter and the monitoring factor parameter.
[0077] In some embodiments, if the monitoring factor parameter of the skeletal monitoring factor is within the standard factor parameter interval of the target standard factor, the skeletal monitoring factor is a normal monitoring factor; the abnormal factor parameter of the abnormal monitoring factor can be calculated by a calculation function: abnormal factor parameter = monitoring factor parameter - interval center parameter, and the abnormal factor parameter reflects the abnormality degree of the abnormal monitoring factor.
[0078] An abnormal impact bone age of the target child is obtained according to the factor evaluation weight in the evaluation factor interval and the abnormal factor parameter of the abnormal monitoring factor, including:
[0079] The abnormal factor parameter of the abnormal monitoring factor is weighted and accumulated with the factor evaluation weight of the abnormal monitoring factor to obtain a bone age impact coefficient of the target child;
[0080] An abnormal impact bone age of the target child is obtained according to the bone age impact coefficient and the actual age of the child.
[0081] In some embodiments, the factor evaluation weight is preset according to the bone age impact weight of the bone factor characteristics on the actual age of the child, the bone age impact coefficient of the target child is obtained by weighting the abnormal factor parameter with the factor evaluation weight corresponding to the abnormal factor parameter, and then accumulated and calculated; the abnormal impact bone age of the target child is calculated by a calculation function: abnormal impact bone age = actual age of child * bone age impact coefficient, wherein the abnormal impact bone age is an age impact value caused by the abnormal monitoring factor on the actual age of the child; the child evaluation bone age of the target child is calculated by a calculation function: child evaluation bone age = actual age of child - abnormal impact bone age; wherein the abnormal factor parameter, the bone age impact coefficient and the abnormal impact bone age can be positive or negative.
[0082] The abnormal monitoring factor is matched and analyzed with the abnormal characteristic factor to obtain a target characteristic factor and a target abnormal diagnosis factor corresponding to the target child, including:
[0083] The abnormal monitoring factor is matched with the abnormal characteristic factor in the abnormal factor interval to obtain the abnormal characteristic factor in the abnormal factor interval which is consistent with the abnormal monitoring factor and the abnormal characteristic factor, and the abnormal characteristic factor is recorded as the target characteristic factor;
[0084] The target factor connection track of the target characteristic factor is obtained, and the target abnormal diagnosis factor corresponding to the target characteristic factor is obtained according to the target factor connection track.
[0085] In some embodiments, the abnormal factor area includes at least one target characteristic factor consistent with the abnormal monitoring factor, each target characteristic factor has a factor diagnosis track, and the factor diagnosis track of the target characteristic factor is recorded as a target factor connection track. It should be noted that if the abnormal diagnosis factor corresponding to the target characteristic factor is also an abnormal characteristic factor, the factor diagnosis track of the target characteristic factor is extended to the factor diagnosis track of the abnormal diagnosis factor, until the abnormal diagnosis factor is only an abnormal diagnosis factor, and the final target factor connection track is obtained. All abnormal diagnosis factors on the target factor connection track are recorded as target abnormal diagnosis factors corresponding to the target characteristic factor.
[0086] The bone age assessment symptoms of the target child are obtained according to the target abnormal diagnosis factor, including:
[0087] The bone age assessment symptoms of the target child are obtained according to the target abnormal diagnosis factor, including:
[0088] In some embodiments, after the target abnormal diagnosis factor is obtained, the bone diagnosis characteristic corresponding to the target abnormal diagnosis factor can be obtained, wherein the bone diagnosis characteristic further includes a characteristic parameter corresponding to the bone diagnosis characteristic, and the bone age assessment symptoms of the target child are constituted according to the bone diagnosis characteristic. It should be noted that the target child can include at least one target abnormal diagnosis factor, and the bone age assessment symptoms of the target child include the bone diagnosis characteristics corresponding to all target abnormal diagnosis factors of the target child.
[0089] The bone age comprehensive assessment result of the target child is obtained according to the assessment result area, including:
[0090] The bone age comprehensive assessment result includes the child assessment bone age of the target child and the bone age assessment symptoms.
[0091] In some embodiments, after the bone monitoring factor of the target child is input into the evaluation factor area of the bone age evaluation graph as a graph input condition, the bone age comprehensive assessment result of the target child can be obtained through the evaluation result area of the bone age evaluation graph, wherein the bone age comprehensive assessment result includes the child assessment bone age of the target child and the bone age assessment symptoms of the target child.
[0092] The above are only preferred embodiments of the present application, and the protection scope of the present application is not limited to the above-mentioned embodiments. Any technical solutions falling within the concept of the present application shall fall within the protection scope of the present application. It should be noted that, for ordinary skilled persons in the art, some improvements and refinements without departing from the principles of the present application shall also be considered as falling within the protection scope of the present application.
Claims
1. A comprehensive assessment method for children's bone age based on deep learning, characterized by: include: Step S1: Obtain the actual age of the target child, and obtain a corresponding skeletal assessment map of the target child based on the actual age of the child, wherein the skeletal assessment map includes an assessment factor area, an abnormal factor area, and an assessment result area; wherein the assessment factor area includes skeletal standard factors and factor assessment weights corresponding to the skeletal standard factors; and the abnormal factor area includes abnormal characteristic factors and abnormal diagnostic factors. Step S2: Obtaining bone image data of the target child, and obtaining a bone monitoring factor of the target child based on the bone image data; Step S3: Inputting the target child's bone monitoring factors as atlas input conditions into the bone assessment atlas, matching and analyzing the bone monitoring factors in the atlas input conditions with the bone standard factors in the assessment factor area, and obtaining the target child's abnormal monitoring factors and abnormal factor parameters; Step S4: obtaining the abnormal impact bone age of the target child based on the factor evaluation weights in the evaluation factor area and the abnormal factor parameters of the abnormal monitoring factors, and obtaining the child assessed bone age of the target child based on the abnormal impact bone age and the child's actual age; Step S5: transmitting the abnormal monitoring factor to the abnormal factor area; matching and analyzing the abnormal monitoring factor with the abnormal characteristic factor in the abnormal factor area to obtain the target characteristic factor and target abnormal diagnosis factor corresponding to the target child; and obtaining the bone age assessment symptoms of the target child based on the target abnormal diagnostic factor; Step S6: The child's assessed bone age and bone age assessment symptoms are transmitted to the assessment result area, and a comprehensive bone age assessment result of the target child is obtained based on the assessment result area.
2. A method for comprehensive assessment of children's bone age based on deep learning according to claim 1, characterized in that: The evaluation factor area includes the skeletal standard factor and the factor evaluation weight corresponding to the skeletal standard factor, including: obtaining standard skeletal data of the target child according to the actual age of the child; Extracting skeleton features based on the standard skeleton data to obtain standard skeleton features corresponding to the target child; and generating a skeletal standard factor corresponding to the target child and a standard factor parameter interval corresponding to the skeletal standard factor based on the standard skeletal feature, wherein the skeletal standard factor corresponds to the standard skeletal feature; The factor assessment weight of the target child's corresponding bone standard factor is set according to the child's actual age.
3. A method for comprehensive assessment of children's bone age based on deep learning according to claim 2, characterized in that: The abnormal factor area includes abnormal characteristic factors and abnormal diagnostic factors, including: Obtaining historical bone abnormality data corresponding to the child's actual age, wherein the historical bone abnormality data includes bone abnormality monitoring data and bone diagnosis results corresponding to the bone abnormality monitoring data; Extracting bone features from the bone abnormality monitoring data to obtain bone abnormality features corresponding to the target child; and generating abnormal feature factors and abnormal factor parameters corresponding to the abnormal feature factors based on the bone abnormality features, wherein the abnormal feature factors correspond to the bone abnormality features; Extracting bone features from the bone diagnosis results to obtain bone diagnostic features corresponding to the target child; and generating abnormal diagnostic factors and diagnostic factor parameters corresponding to the abnormal diagnostic factors based on the bone diagnostic features, wherein the abnormal diagnostic factors correspond to the bone diagnostic features; The abnormal characteristic factor is connected with the abnormal diagnosis factor corresponding to the abnormal characteristic factor to obtain a factor diagnosis trajectory between the abnormal characteristic factor and the abnormal diagnosis factor.
4. A method for comprehensive assessment of children's bone age based on deep learning according to claim 3, characterized in that: Obtaining skeletal image data of the target child, and obtaining a skeletal monitoring factor of the target child based on the skeletal image data, including: Extracting bone features from the bone image data to obtain bone monitoring features of the target child; A bone monitoring factor of the target child and a monitoring factor parameter corresponding to the bone monitoring factor are generated based on the bone monitoring feature, and the bone monitoring factor corresponds to the bone monitoring feature.
5. A method for comprehensive assessment of children's bone age based on deep learning according to claim 4, characterized in that: Step S3 includes: Factor matching is performed on the bone monitoring factor in the atlas input condition and the bone standard factor in the evaluation factor area to obtain the bone standard factor in the evaluation factor area that is consistent with the bone monitoring factor, and the bone standard factor is recorded as the target standard factor; Comparing the monitoring factor parameters of the bone monitoring factor with the standard factor parameter intervals of the target standard factor; If the monitoring factor parameter of the bone monitoring factor is not within the standard factor parameter interval of the target standard factor, the bone monitoring factor is recorded as an abnormal monitoring factor, and the interval center parameter of the standard factor parameter interval is obtained; The abnormal factor parameter of the abnormal monitoring factor is obtained according to the interval center parameter and the monitoring factor parameter.
6. A method for comprehensive assessment of children's bone age based on deep learning according to claim 5, characterized in that: The abnormal impact bone age of the target child is obtained based on the factor evaluation weights in the evaluation factor area and the abnormal factor parameters of the abnormal monitoring factors, including: Performing weighted accumulation of the abnormal factor parameter of the abnormal monitoring factor and the factor evaluation weight of the abnormal monitoring factor to obtain the bone age influence coefficient of the target child; The abnormally affected bone age of the target child is obtained based on the bone age influence coefficient and the child's actual age.
7. A method for comprehensive assessment of children's bone age based on deep learning according to claim 6, characterized in that: Matching and analyzing the abnormal monitoring factors with the abnormal characteristic factors in the abnormal factor area to obtain target characteristic factors and target abnormal diagnostic factors corresponding to the target child, including: Matching the abnormal monitoring factor with the abnormal characteristic factor in the abnormal factor area to obtain an abnormal characteristic factor in the abnormal factor area that is consistent with the abnormal monitoring factor and the abnormal characteristic factor, and recording the abnormal characteristic factor as a target characteristic factor; A target factor connection trajectory of the target characteristic factor is obtained, and a target abnormality diagnosis factor corresponding to the target characteristic factor is obtained according to the target factor connection trajectory.
8. A method for comprehensive assessment of children's bone age based on deep learning according to claim 7, characterized in that: And based on the target abnormal diagnostic factors, the bone age assessment symptoms of the target child are obtained, including: Obtain the bone diagnostic features corresponding to the target abnormal diagnostic factors, and obtain the bone age assessment symptoms of the target child based on the bone diagnostic features corresponding to the target abnormal diagnostic factors.
9. A method for comprehensive assessment of children's bone age based on deep learning according to claim 8, characterized in that: Based on the assessment results, a comprehensive bone age assessment result of the target child is obtained, including: The comprehensive bone age assessment results include the child-assessed bone age and bone age assessment symptoms of the target child.