A clinical decision method, device and electronic equipment for osteomyelitis
By evaluating patients' medical records, images, and wound data, automated staging and classification of osteomyelitis were achieved, solving the problem of high misdiagnosis rates, simplifying the diagnosis and treatment process, and improving the accuracy of diagnosis and treatment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- RENMIN HOSPITAL OF WUHAN UNIVERSITY (HUBEI GENERAL HOSPITAL)
- Filing Date
- 2026-01-16
- Publication Date
- 2026-05-29
Smart Images
Figure CN122117316A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of clinical support technology, and more particularly to a clinical decision-making method, apparatus, and electronic device for osteomyelitis. Background Technology
[0002] Osteomyelitis is an inflammatory disease caused by pyogenic bacterial infection of the bone marrow, cortical bone, and periosteum. The most common causes include hematogenous infection, traumatic infection, and surgical infection. In adults, it commonly affects long bones; in diabetic patients, it frequently occurs in the foot or at sites of penetrating bone injury caused by trauma; and in children, it commonly affects the well-vascularized metaphysis. Acute osteomyelitis presents with high fever and localized pain. If treatment is delayed or improper, it can easily develop into chronic osteomyelitis, characterized by ulceration, purulent discharge, sequestrum, or cavity formation. Severe cases are often life-threatening, sometimes necessitating amputation and resulting in lifelong disability. Osteomyelitis is frequently misdiagnosed; acute cases are easily confused with soft tissue infections, while chronic cases often present with insidious pain, easily overlooked and delaying medical attention. Currently, approximately 2 million osteomyelitis surgeries are performed globally each year. The classification and treatment of osteomyelitis are quite complex, requiring a comprehensive evaluation of imaging examinations, laboratory tests, wound assessment, and the selection of appropriate clinical treatment plans, including surgical plans and antibiotic use plans, based on the doctor's experience. The process is tedious and complex and easily affected by subjective factors, making it difficult to guarantee the certainty of the treatment effect. Summary of the Invention
[0003] This invention provides a clinical decision-making method, device, and electronic device for osteomyelitis, which addresses the problem of high misdiagnosis rates in human osteomyelitis diagnosis.
[0004] According to one aspect of the present invention, a clinical decision-making method for osteomyelitis is provided, comprising: Acquire patient medical record text data, wound image data, and medical imaging data; Medical imaging data is input into the corresponding image recognition model to obtain the patient's bone condition; The wound image data is input into the corresponding wound recognition model to obtain the patient's wound status; The determination of whether a patient has osteomyelitis is based at least on the medical record text data and the bone condition assessment of the patient. If the patient has osteomyelitis, the patient's osteomyelitis is staged based on the medical record text data, the bone condition, and the wound condition. Based on the patient's medical record text data, medical imaging data, and wound condition, the patient's osteomyelitis was classified. Treatment recommendations are generated based on the patient's osteomyelitis diagnosis, staging, and subtype results. Optionally, the classification of osteomyelitis based on the patient's medical record text data, medical imaging data, and wound condition includes: The patient's X-ray images were input into the osteomyelitis classification model to identify the patient's anatomical type; Based on the patient's medical record text data, the wound status is used to perform host typing of the patient's osteomyelitis.
[0005] Optionally, prior to inputting the patient's X-ray images into the osteomyelitis classification model to identify the patient's anatomical classification, the method further includes: Historical X-ray image data of different patients were obtained and annotated to construct the first dataset; the annotation content was the anatomical classification type. The osteomyelitis subtyping model is trained based on the first dataset to obtain the trained osteomyelitis subtyping model; the osteomyelitis subtyping model is used to identify the anatomical subtype of the patient.
[0006] Optionally, if the medical imaging data includes X-ray images, CT images, and MRI images, then the step of inputting the medical imaging data into the corresponding image recognition model to obtain the patient's bone condition includes: The patient's X-ray image is input into the first image recognition model to identify whether the patient has bone destruction, sequestrum, or periosteal reaction. The patient's CT images are input into a second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume; The patient's MRI images are input into a third image recognition model to identify whether the patient has bone marrow edema.
[0007] Optionally, before inputting the patient's X-ray image into the first image recognition model to identify whether the patient has bone destruction, sequestrum, or periosteal reaction, the method further includes: Historical X-ray image data of different patients were obtained and annotated to construct a second dataset; the annotation content included bone destruction, sequestrum, and periosteal reaction. The first image recognition model is trained based on the second dataset to obtain the trained first image recognition model; the first image recognition model is used to identify the patient's bone destruction, necrotic bone, and periosteal reaction.
[0008] Optionally, before inputting the patient's CT image into the second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume, the method further includes: Historical CT image data of different patients were acquired and annotated to construct a third dataset; the annotation content was the segmentation contour of bone defects. The second image recognition model is trained based on the third dataset to obtain the trained second image recognition model; the second image recognition model is used to identify the segmentation contour and volume of the patient's bone defect.
[0009] Optionally, before inputting the patient's MRI image into a third image recognition model to identify whether the patient has bone marrow edema, the method further includes: Historical MRI image data of different patients were obtained and annotated to construct a fourth dataset; the annotation content was bone marrow edema status. The third image recognition model is trained based on the fourth dataset to obtain the trained third image recognition model; the third image recognition model is used to identify the patient's bone marrow edema.
[0010] Optionally, before inputting the wound image data into the corresponding wound recognition model to obtain the patient's wound status, the method further includes: Historical osteomyelitis wound image data of different patients were obtained and annotated to construct the fifth dataset; the annotation content included wound pus condition, wound infection bacterial type, wound granulation condition, and wound sinus tract condition. The wound recognition model is trained based on the fifth dataset to obtain the trained wound recognition model; the wound recognition model is used to identify the patient's wound pus condition, wound infection bacterial type, wound granulation condition, and wound sinus tract condition.
[0011] According to another aspect of the present invention, a clinical decision-making apparatus for osteomyelitis is provided, comprising: The data acquisition module is used to acquire patients' medical record text data, wound image data, and medical imaging data; The bone condition recognition module is used to input medical image data into the corresponding image recognition model to obtain the patient's bone condition; The wound recognition module is used to input the wound image data into the corresponding wound recognition model to obtain the patient's wound status; The osteomyelitis assessment module is used to determine whether a patient has osteomyelitis based at least on the medical record text data and the bone condition assessment of the patient. The osteomyelitis staging module is used to assess and stage the osteomyelitis of a patient based on the medical record text data, the bone condition, and the wound condition if the patient has osteomyelitis. The osteomyelitis classification module is used to classify osteomyelitis in patients based on the patient's medical record text data, medical imaging data, and wound condition. The treatment recommendation module is used to output treatment recommendations based on the patient's osteomyelitis diagnosis, osteomyelitis staging, and osteomyelitis subtype. According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the clinical decision-making method for osteomyelitis as described in any embodiment of the present invention.
[0012] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the clinical decision-making method for osteomyelitis as described in any embodiment of the present invention.
[0013] The technical solution of this invention involves acquiring patient medical record text data, wound image data, and medical imaging data; inputting the medical imaging data into a corresponding image recognition model to obtain the patient's bone status; inputting the wound image data into a corresponding wound recognition model to obtain the patient's wound status; assessing whether the patient has osteomyelitis based on the medical record text data, bone status, and wound status; if the patient has osteomyelitis, staging the osteomyelitis based on the medical record text data, bone status, and wound status; classifying the osteomyelitis based on the patient's medical record text data, medical imaging data, and wound status; and outputting treatment recommendations based on the patient's osteomyelitis diagnosis, staging, and classification results. This invention predicts osteomyelitis, its stage, and classification by using a constructed multi-input clinical decision-making system. This quantitative prediction method eliminates the need for manual judgment, outputting the clinical stage and classification of osteomyelitis, thereby assisting doctors in making clinical treatment decisions and significantly reducing the misdiagnosis rate of osteomyelitis.
[0014] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description
[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0016] Figure 1 This is a flowchart of a clinical decision-making method for osteomyelitis according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of a clinical decision-making method for osteomyelitis according to Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of a clinical decision-making device for osteomyelitis according to Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the clinical decision-making method for osteomyelitis according to an embodiment of the present invention. Detailed Implementation
[0017] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.
[0018] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.
[0019] Example 1 Figure 1 This is a flowchart illustrating a clinical decision-making method for osteomyelitis, as provided in Embodiment 1 of the present invention. Figure 1 As shown, the method includes: S101. Obtain the patient's medical record text data, wound image data, and medical imaging data.
[0020] The patient's case data may include basic information, symptom duration, symptom location, underlying diseases, medical history, trauma history, surgical history, wound condition, pain score, and laboratory test data. Basic information may include the patient's name, age, and gender; symptom duration may include duration of symptoms less than 4 weeks, greater than 4 weeks, or no symptoms; symptom location may include upper limbs, lower limbs, or trunk; underlying diseases may include presence of hypertension, diabetes, or serious underlying diseases; medical history may include whether there was fever within 3 days, duration of fever, and highest temperature (in °C); trauma history may include whether there was trauma and the location of trauma (options: upper limbs, lower limbs, trunk); surgical history may include whether surgery was performed and the surgical location (options: upper limbs, lower limbs, trunk); wound condition may include presence of a wound and presence of foul odor; pain score may include the VAS pain score, ranging from 0 to 10 (0 indicates no pain, 10 indicates the most painful); laboratory test data may include complete blood count (white blood cell count (×10⁻¹⁰)). 9 The data included: neutrophil percentage (%), hemoglobin (g / L), CRP (mg / L), procalcitonin (ng / mL), albumin, glycated hemoglobin, blood culture data (positive / negative, bacterial type, and drug sensitivity results), and secretion culture data (positive / negative, bacterial type, and drug sensitivity results).
[0021] The wound image data consists of local photographs of the patient's wound, which can support a variety of common image formats.
[0022] Medical imaging data can include X-ray, CT, and MRI images.
[0023] S102. Input the medical image data into the corresponding image recognition model to obtain the patient's bone condition.
[0024] It should be noted that different medical imaging data can be input into corresponding image recognition models to identify the patient's bone condition. Bone condition can include bone destruction, sequestrum, periosteal reaction, bone defect volume, and bone marrow edema.
[0025] S103. Input the wound image data into the corresponding wound recognition model to obtain the patient's wound status.
[0026] The wound recognition model is used to identify the wound status based on the input wound image data. The wound status can include whether there is pus in the wound, the type of bacteria causing the infection if pus is present, whether there is granulation tissue in the wound, whether there is a sinus tract in the wound and the predicted depth of the sinus tract.
[0027] S104. Determine whether the patient has osteomyelitis based at least on the medical record text data and the bone condition assessment of the patient.
[0028] The system can be used to assess whether a patient has osteomyelitis by comparing their medical records, bone condition, and wound condition against pre-defined osteomyelitis assessment criteria. For example, if the medical records indicate a long period of symptom onset, and the bone shows periosteal reaction and bone marrow edema, the patient can be diagnosed with osteomyelitis; otherwise, osteomyelitis is not present.
[0029] S105. If the patient has osteomyelitis, the patient's osteomyelitis is staged based on the medical record text data, the bone condition, and the wound condition.
[0030] The osteomyelitis can be staged by comparing the patient's medical record data, bone condition, and wound condition against a pre-defined staging standard. The staging results can include acute osteomyelitis and chronic osteomyelitis.
[0031] S106. Based on the patient's medical record text data, medical imaging data, and wound condition, the patient's osteomyelitis is classified.
[0032] It should be noted that the classification of osteomyelitis can include anatomical classification and host classification. Pre-set classification methods can be used to classify osteomyelitis based on the patient's medical record text data, medical imaging data, and wound condition to obtain the patient's anatomical classification and host classification.
[0033] S107. Based on the patient's osteomyelitis diagnosis, osteomyelitis staging, and osteomyelitis subtyping results, output treatment recommendations. In this embodiment, corresponding treatment suggestions can be output based on the preset treatment decision criteria, the patient's osteomyelitis diagnosis results, osteomyelitis staging results, and osteomyelitis subtype results.
[0034] It should be noted that medical record text data, wound image data, and medical imaging data need to be input into the osteomyelitis clinical decision system. The system can automatically output bone status, wound status, osteomyelitis condition, osteomyelitis stage, osteomyelitis type, and provide treatment suggestions.
[0035] The technical solution of this invention involves acquiring patient medical record text data, wound image data, and medical imaging data; inputting the medical imaging data into a corresponding image recognition model to obtain the patient's bone status; inputting the wound image data into a corresponding wound recognition model to obtain the patient's wound status; assessing whether the patient has osteomyelitis based on the medical record text data, bone status, and wound status; if the patient has osteomyelitis, staging the osteomyelitis based on the medical record text data, bone status, and wound status; classifying the osteomyelitis based on the patient's medical record text data, medical imaging data, and wound status; and outputting treatment recommendations based on the patient's osteomyelitis diagnosis, staging, and classification results. This invention predicts osteomyelitis, its stage, and classification by using a constructed multi-input clinical decision-making system. This quantitative prediction method eliminates the need for manual judgment, outputting the clinical stage and classification of osteomyelitis, thereby assisting doctors in making clinical treatment decisions and significantly reducing the misdiagnosis rate of osteomyelitis.
[0036] Example 2 Figure 2 This is a flowchart illustrating a clinical decision-making method for osteomyelitis, as provided in Embodiment 2 of the present invention. Figure 2 As shown, the method includes: S201. Obtain the patient's medical record text data, wound image data, and medical imaging data.
[0037] The patient's case data may include basic information, symptom duration, symptom location, underlying diseases, medical history, trauma history, surgical history, wound condition, pain score, and laboratory test data. Basic information may include the patient's name, age, and gender; symptom duration may include duration of symptoms less than 4 weeks, greater than 4 weeks, or no symptoms; symptom location may include upper limbs, lower limbs, or trunk; underlying diseases may include presence of hypertension, diabetes, or serious underlying diseases; medical history may include whether there was fever within 3 days, duration of fever, and highest temperature (in °C); trauma history may include whether there was trauma and the location of trauma (options: upper limbs, lower limbs, trunk); surgical history may include whether surgery was performed and the surgical location (options: upper limbs, lower limbs, trunk); wound condition may include presence of a wound and presence of foul odor; pain score may include the VAS pain score, ranging from 0 to 10 (0 indicates no pain, 10 indicates the most painful); laboratory test data may include complete blood count (white blood cell count (×10⁻¹⁰)). 9The data included: neutrophil percentage (%), hemoglobin (g / L), CRP (mg / L), procalcitonin (ng / mL), albumin, glycated hemoglobin, blood culture data (positive / negative, bacterial type, and drug sensitivity results), and secretion culture data (positive / negative, bacterial type, and drug sensitivity results).
[0038] The wound image data consists of local photographs of the patient's wound, which can support a variety of common image formats.
[0039] Medical imaging data can include X-ray, CT, and MRI images.
[0040] S202. Input the patient's X-ray image into the first image recognition model to identify whether the patient has bone destruction, dead bone, or periosteal reaction.
[0041] In one embodiment, prior to inputting the patient's X-ray image into a first image recognition model to identify whether the patient has bone destruction, sequestrum, or periosteal reaction, the method further includes: S2021. Obtain historical X-ray image data of different patients, and annotate the historical X-ray image data to construct a second dataset; the annotation content includes bone destruction, sequestrum, and periosteal reaction. S2022. The first image recognition model is trained based on the second dataset to obtain the trained first image recognition model; the first image recognition model is used to identify the patient's bone destruction, necrotic bone, and periosteal reaction.
[0042] Anonymized historical X-ray images of osteomyelitis patients from multiple top-tier hospitals can be used. Several senior orthopedic surgeons annotated the historical X-ray images, noting whether bone destruction was observed, the presence of sequestrum, and periosteal reaction, thus constructing a second dataset. Furthermore, data augmentation can be performed using random rotation, contrast adjustment, and Gaussian noise.
[0043] It should be noted that the second dataset can be divided into training and testing sets in a 4:1 ratio. A ResNet-50 architecture is used as the base network to build the first image recognition model, and transfer learning is employed to output multi-label classifications. This first image recognition model is then trained until a preset number of iterations is reached. The number of iterations can be set manually. The algorithm is considered successfully trained when the classification accuracy is >80% and the AUC is >0.8.
[0044] S203. Input the patient's CT image into the second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume.
[0045] In one embodiment, prior to inputting the patient's CT image into a second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume, the method further includes: S2031. Obtain historical CT image data of different patients, and annotate the historical CT image data to construct a third dataset; the annotation content is the segmentation contour of bone defects. S2032. The second image recognition model is trained based on the third dataset to obtain the trained second image recognition model; the second image recognition model is used to identify the segmentation contour and volume of the patient's bone defect.
[0046] Anonymized historical CT image data of osteomyelitis patients from multiple top-tier hospitals can be used. Several senior orthopedic surgeons can annotate the historical CT image data, specifically the segmentation contours of bone defects, to construct a third dataset. The historical CT image data can be converted into fixed-size slices, such as 512x512 pixels.
[0047] It should be noted that the third dataset can be divided into training and testing sets in a 4:1 ratio. The U-Net architecture is used as the base network to build the second image recognition model. Transfer learning is used to output a binary segmentation mask of the bone defect area. The second image recognition model is then trained. The algorithm is considered to be qualified when DICE > 0.85.
[0048] S204. Input the patient's MRI image into the third image recognition model to identify whether the patient has bone marrow edema.
[0049] In one embodiment, prior to inputting the patient's MRI image into a third image recognition model to identify whether the patient has bone marrow edema, the method further includes: S2041. Obtain historical MRI image data of different patients, and annotate the historical MRI image data to construct a fourth dataset; the annotation content is the bone marrow edema status; S2042. The third image recognition model is trained based on the fourth dataset to obtain the trained third image recognition model; the third image recognition model is used to identify the patient's bone marrow edema.
[0050] Anonymized historical MRI images of osteomyelitis patients from multiple top-tier hospitals can be used. Several senior orthopedic surgeons annotated the historical MRI images, specifying whether bone marrow edema was present, thus constructing a fourth dataset. Furthermore, data augmentation can be performed using random rotation, contrast adjustment, and Gaussian noise.
[0051] It should be noted that the third dataset can be divided into training and testing sets in a 4:1 ratio. A 3dResNet-18 architecture is used as the base network to build the third image recognition model, outputting a multi-label classification. Five-fold cross-validation is used to ensure the model's generalization ability. The third image recognition model is trained until a preset number of iterations is reached. The number of iterations can be set manually. The algorithm is considered successfully trained when the classification accuracy is >80% and the AUC is >0.8.
[0052] S205. Input the wound image data into the corresponding wound recognition model to obtain the patient's wound status.
[0053] In one embodiment, before inputting the wound image data into the corresponding wound recognition model to obtain the patient's wound status, the method further includes: S2051. Obtain historical osteomyelitis wound image data from different patients, and annotate the historical osteomyelitis wound image data to construct the fifth dataset; the annotation content includes wound pus status, wound infection bacterial type, wound granulation status, and wound sinus tract status. S2052. The wound recognition model is trained based on the fifth dataset to obtain the trained wound recognition model; the wound recognition model is used to identify the patient's wound pus condition, wound infection bacterial type, wound granulation condition, and wound sinus tract condition.
[0054] Anonymized historical wound image data of osteomyelitis patients from multiple top-tier hospitals can be used. Several senior orthopedic surgeons annotated the historical wound image data, including: presence of pus, bacterial type (Staphylococcus aureus, Pseudomonas aeruginosa, anaerobic bacteria, others), presence of granulation tissue, and presence of sinus tracts, thus constructing a fifth dataset. Furthermore, data augmentation can be performed using random rotation, contrast adjustment, and Gaussian noise, while preserving the geometric transformations of color during data augmentation.
[0055] It should be noted that the fifth dataset can be divided into training and testing sets in a 4:1 ratio. An EfficientNet-B4 architecture can be used as the backbone network to build a wound recognition model, with the output being a multi-label classification. This model can then be trained on the first image recognition model until a preset number of iterations is reached. The number of iterations can be set manually. The algorithm is considered successfully trained when the classification accuracy is >80% and the AUC is >0.8. It is important to note that class imbalance should be addressed (using Focal Loss for rare classes).
[0056] S206. Determine whether the patient has osteomyelitis based at least on the medical record text data and the bone condition assessment of the patient.
[0057] In this embodiment, if the symptom duration in the patient's medical record text data is less than or greater than 4 weeks, and the periosteum shows a reaction and the bone marrow shows edema in the bone status output by the imaging recognition model, the patient is determined to have osteomyelitis. Otherwise, the patient is determined not to have osteomyelitis.
[0058] S207. If the patient has osteomyelitis, the patient's osteomyelitis is staged based on the medical record text data, the bone condition, and the wound condition.
[0059] In this embodiment, when a patient is identified as having osteomyelitis, the patient's osteomyelitis can be staged based on the patient's medical record text data, bone condition, and wound condition.
[0060] Specifically, if any of the following conditions are met: ①, ②, ③, ④, ⑤, ⑥, ⑦, ⑧, ⑨, ⑩, the condition is classified as acute osteomyelitis: ① Diagnosed as osteomyelitis; ②The duration of symptoms is less than 4 weeks; ③ History of fever within 3 days, with the highest body temperature >38.5℃; ④ History of trauma or surgery; ⑤ Pain score > 5 points; ⑥ In laboratory test data, the white blood cell count is >10×10 9 / L, neutrophil percentage >75%; ⑦ In laboratory test data, CRP > 5 mg / L or procalcitonin > 0.05 ng / mL; ⑧ In laboratory test data, blood culture is positive, or wound secretion culture is positive; ⑨ The bone condition shows bone destruction, bone defects, and periosteal reaction; ⑩ In the wound condition, there is no granulation tissue and no sinus tract.
[0061] If the following conditions ①, ②, ③, and ④ are met, the condition is classified as chronic osteomyelitis: ① Diagnosed as osteomyelitis; ② Symptoms lasted for more than 4 weeks; ③ The bone structure shows bone destruction, periosteal reaction, and bone marrow edema; ④ In the wound condition, there is pus, granulation tissue, and sinus tracts.
[0062] S208. Input the patient's X-ray images into the osteomyelitis classification model to identify the patient's anatomical classification.
[0063] Specifically, the Cierny-Mader classification includes two parts: anatomical classification and host classification. The anatomical classification includes: Type I: Intramedullary, infection confined to the medullary cavity; Type II: Superficial, infection on the surface of the cortical bone; Type III: Localized, full-thickness cortical destruction + sequestrum formation; Type IV: Diffuse, ring-shaped bone destruction with mechanical instability. The host classification includes: Category A: Normal immunity + good soft tissue blood supply; Category B: Systemic / local deficiencies; Category C: Treatment risks > benefits.
[0064] In one embodiment, prior to inputting the patient's X-ray images into the osteomyelitis classification model to identify the patient's anatomical classification, the method further includes: S2081. Obtain historical X-ray image data of different patients, and annotate the historical X-ray image data to construct the first dataset; the annotation content is the anatomical classification type; S2082. The osteomyelitis subtyping model is trained based on the first dataset to obtain the trained osteomyelitis subtyping model; the osteomyelitis subtyping model is used to identify the anatomical subtyping of patients.
[0065] Anonymized historical X-ray image data of osteomyelitis patients from multiple top-tier hospitals can be used. Several senior orthopedic surgeons annotated the historical X-ray image data, labeling it with CM classification (anatomical types I, II, III, and IV), thus constructing the first dataset. Furthermore, data augmentation can be performed using random rotation, contrast adjustment, and Gaussian noise.
[0066] It should be noted that the first dataset can be divided into training and testing sets in a 4:1 ratio. A ResNet-50 architecture is used as the base network to build an osteomyelitis subtyping model, and transfer learning is employed to output multi-label classifications. This model is then trained until a preset number of iterations is reached. The number of iterations can be set manually. The algorithm is considered successfully trained when the classification accuracy is >80% and the AUC is >0.8.
[0067] S209. Based on the patient's medical record text data, the wound status is used to perform host typing of the patient's osteomyelitis.
[0068] When the laboratory test data shows a white blood cell count <10×10 9 The host type is classified as A when the following criteria are met: neutrophil percentage <75%, CRP <5 mg / L, procalcitonin <0.05 ng / mL, hemoglobin >110 g / L, albumin ≥35 g / L, 4% ≤ glycated hemoglobin ≤6%, and the wound contains pus, granulation tissue, and sinus tracts. When the patient has severe underlying diseases, the host type is classified as C. If neither category A nor C is met, the patient is classified as B.
[0069] S210. Based on the patient's osteomyelitis diagnosis, osteomyelitis staging, and osteomyelitis subtyping results, output treatment recommendations. In one embodiment, based on the results of the osteomyelitis diagnosis module, osteomyelitis staging module, and osteomyelitis typing module, treatment suggestions for osteomyelitis can be output, as follows: ① If it is determined that it is not osteomyelitis, output: symptomatic and supportive treatment, and further relevant examinations; ② When diagnosed with acute osteomyelitis, the treatment should include: symptomatic and supportive care + systemic antibiotic therapy; ③ When diagnosed with chronic osteomyelitis and classified as type I, category A, the treatment is: medullary cavity irrigation for 6 weeks; ④ When diagnosed with chronic osteomyelitis and classified as type I, category B, the treatment plan is: 6 weeks of medullary cavity irrigation + prolonged systemic antibiotic treatment; ⑤ When diagnosed with chronic osteomyelitis and classified as type II, category A, the treatment should be: debridement + local antibiotic delivery. ⑥ When diagnosed with chronic osteomyelitis and classified as type II, category B, the output should be: debridement + vascularized muscle flap; ⑦ When diagnosed with chronic osteomyelitis and classified as type III, category A, the output is: resection of the sequestrum + primary suturing; ⑧ When diagnosed with chronic osteomyelitis and classified as type III, category B, the recommended treatment is: Papineau procedure + negative pressure drainage; ⑨ When diagnosed with chronic osteomyelitis and classified as type IV, category A, the output should be: thorough debridement + Ilizarov bone transport. ⑩ When diagnosed with chronic osteomyelitis and classified as type IV, category B, output: amputation assessment; When diagnosed with chronic osteomyelitis and classified as type C, the output should be: Palliative care.
[0070] In one embodiment, a pathogen output module is also included, which outputs the pathogen type based on laboratory test data, and displays it as: "Blood culture pathogen: xx" or "Secrecy culture pathogen: xx".
[0071] The system can also output the pathogen type based on the wound image as input and the output of the wound recognition model, displaying it as: "Pus pathogen: xx". It can also output the bone defect volume based on the output results of the second image recognition model, which is displayed as: "Bone defect volume: xx cm" 3 ".
[0072] It also includes an antibiotic assessment output module, which outputs an antibiotic assessment based on the drug sensitivity test results of laboratory test data, displaying it as: "It is recommended to use antibiotic xx".
[0073] It may also include a treatment plan output module, which outputs the final treatment recommendation and displays it as: "Recommended treatment plan: xx".
[0074] Example 3 Figure 3 This is a schematic diagram of a clinical decision-making device for osteomyelitis provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: Data acquisition module 301 is used to acquire patient medical record text data, wound image data and medical imaging data; The bone condition recognition module 302 is used to input medical image data into the corresponding image recognition model to obtain the patient's bone condition; The wound recognition module 303 is used to input the wound image data into the corresponding wound recognition model to obtain the patient's wound status; Osteomyelitis assessment module 304 is used to determine whether a patient has osteomyelitis based at least on the medical record text data and the bone condition assessment of the patient. Osteomyelitis staging module 305 is used to assess and stage the osteomyelitis of a patient based on the medical record text data, the bone condition and the wound condition if the patient has osteomyelitis. Osteomyelitis classification module 306 is used to classify osteomyelitis in patients based on the patient's medical record text data, medical image data and wound condition; The treatment recommendation module 307 is used to output treatment recommendations based on the patient's osteomyelitis diagnosis, osteomyelitis staging, and osteomyelitis typing results. The message data processing resource scheduling device provided in this embodiment of the invention can execute the clinical decision-making method for osteomyelitis provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.
[0075] Example 4 Figure 4 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.
[0076] like Figure 4As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.
[0077] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.
[0078] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as a clinical decision-making approach for osteomyelitis.
[0079] In some embodiments, a clinical decision-making method for osteomyelitis may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or mounted on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the clinical decision-making method for osteomyelitis described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform a clinical decision-making method for osteomyelitis by any other suitable means (e.g., by means of firmware).
[0080] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.
[0081] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.
[0082] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.
[0083] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).
[0084] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.
[0085] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.
[0086] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.
[0087] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.
Claims
1. A clinical decision-making method for osteomyelitis, characterized in that, include: Acquire patient medical record text data, wound image data, and medical imaging data; Medical imaging data is input into the corresponding image recognition model to obtain the patient's bone condition; The wound image data is input into the corresponding wound recognition model to obtain the patient's wound status; The determination of whether a patient has osteomyelitis is based at least on the medical record text data and the bone condition assessment of the patient. If the patient has osteomyelitis, the patient's osteomyelitis is staged based on the medical record text data, the bone condition, and the wound condition. Based on the patient's medical record text data, medical imaging data, and wound condition, the patient's osteomyelitis was classified. Treatment recommendations are generated based on the patient's osteomyelitis diagnosis, staging, and subtype results.
2. The clinical decision-making method for osteomyelitis according to claim 1, characterized in that, The classification of osteomyelitis based on the patient's medical record text data, medical imaging data, and wound condition includes: The patient's X-ray images were input into the osteomyelitis classification model to identify the patient's anatomical type; Based on the patient's medical record text data, the wound status is used to perform host typing of the patient's osteomyelitis.
3. The clinical decision-making method for osteomyelitis according to claim 2, characterized in that, Before inputting the patient's X-ray images into the osteomyelitis classification model to identify the patient's anatomical type, the following steps are also included: Historical X-ray image data of different patients were obtained and annotated to construct the first dataset; the annotation content was the anatomical classification type. The osteomyelitis subtyping model is trained based on the first dataset to obtain the trained osteomyelitis subtyping model; the osteomyelitis subtyping model is used to identify the anatomical subtype of the patient.
4. The clinical decision-making method for osteomyelitis according to claim 1, characterized in that, The medical imaging data includes X-ray images, CT images, and MRI images. The step of inputting the medical imaging data into the corresponding image recognition model to obtain the patient's bone condition includes: The patient's X-ray image is input into the first image recognition model to identify whether the patient has bone destruction, sequestrum, or periosteal reaction. The patient's CT images are input into a second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume; The patient's MRI images are input into a third image recognition model to identify whether the patient has bone marrow edema.
5. The clinical decision-making method for osteomyelitis according to claim 4, characterized in that, Before inputting the patient's X-ray image into the first image recognition model to identify whether the patient has bone destruction, sequestrum, or periosteal reaction, the method further includes: Historical X-ray image data of different patients were obtained and annotated to construct a second dataset; the annotation content included bone destruction, sequestrum, and periosteal reaction. The first image recognition model is trained based on the second dataset to obtain the trained first image recognition model; the first image recognition model is used to identify the patient's bone destruction, necrotic bone, and periosteal reaction.
6. The clinical decision-making method for osteomyelitis according to claim 4, characterized in that, Before inputting the patient's CT image into the second image recognition model to identify the patient's bone defect segmentation contour and bone defect volume, the method further includes: Historical CT image data of different patients were acquired and annotated to construct a third dataset; the annotation content was the segmentation contour of bone defects. The second image recognition model is trained based on the third dataset to obtain the trained second image recognition model; the second image recognition model is used to identify the segmentation contour and volume of the patient's bone defect.
7. The clinical decision-making method for osteomyelitis according to claim 4, characterized in that, Prior to inputting the patient's MRI images into a third image recognition model to identify whether the patient has bone marrow edema, the following steps are also included: Historical MRI image data of different patients were obtained and annotated to construct a fourth dataset; the annotation content was bone marrow edema status. The third image recognition model is trained based on the fourth dataset to obtain the trained third image recognition model; the third image recognition model is used to identify the patient's bone marrow edema.
8. The clinical decision-making method for osteomyelitis according to claim 1, characterized in that, Before inputting the wound image data into the corresponding wound recognition model to obtain the patient's wound status, the method further includes: Historical osteomyelitis wound image data of different patients were obtained and annotated to construct the fifth dataset; the annotation content included wound pus condition, wound infection bacterial type, wound granulation condition, and wound sinus tract condition. The wound recognition model is trained based on the fifth dataset to obtain the trained wound recognition model; the wound recognition model is used to identify the patient's wound pus condition, wound infection bacterial type, wound granulation condition, and wound sinus tract condition.
9. A clinical decision-making device for osteomyelitis, characterized in that, include: The data acquisition module is used to acquire patients' medical record text data, wound image data, and medical imaging data; The bone condition recognition module is used to input medical image data into the corresponding image recognition model to obtain the patient's bone condition; The wound recognition module is used to input the wound image data into the corresponding wound recognition model to obtain the patient's wound status; The osteomyelitis assessment module is used to determine whether a patient has osteomyelitis based at least on the medical record text data and the bone condition assessment of the patient. The osteomyelitis staging module is used to assess and stage the osteomyelitis of a patient based on the medical record text data, the bone condition, and the wound condition if the patient has osteomyelitis. The osteomyelitis classification module is used to classify osteomyelitis in patients based on the patient's medical record text data, medical imaging data, and wound condition. The treatment recommendation module is used to output treatment recommendations based on the patient's osteomyelitis diagnosis, osteomyelitis staging, and osteomyelitis subtype.
10. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the clinical decision-making method for osteomyelitis according to any one of claims 1-8.