Infant food allergy noninvasive diagnosis model and system and application thereof

Abstract

The invention discloses a noninvasive diagnosis model for infant food allergy. The noninvasive diagnosis model is used for early recognition of infant food allergy. The model is established through multi-factor Logistic regression based on clinical common symptoms and risk factors, and comprises six variables: eczema severity, stool character, hemafecia, respiratory system symptoms, growth retardation and allergic disease family history. Various variables are endowed with different weights according to objective statistical results to form a model scoring formula: score = 0.23 * eczema (mild) + 1.02 * eczema (moderate) + 1.35 * eczema (severe) + 0.47 * stool (loose stool) + 1.42 * stool (watery stool) + 1.60 * hemafecia + 0.78 * growth retardation + 1.19 * respiratory system symptoms + 1.36 * allergic disease family history. Through verification of a training set and an independent test set, the model has relatively high diagnosis efficiency (AUC is about 0.86-0.87), when the model score is greater than or equal to 3.19, the food allergy is judged, and the optimal diagnosis effect can be achieved. The model is non-invasive, simple and convenient in evaluation process, small in error and high in compliance, and can assist clinicians in diagnosing infant food allergy.

Description

Technical Field

[0001] This invention relates to the field of medical diagnostics, specifically to a non-invasive diagnostic model, system, and application of infant food allergies. Background Technology

[0002] The prevalence of food allergies in infants is as high as 5%-11%. [1,2] Delayed diagnosis of this disease can lead to growth retardation, developmental delays, and even life-threatening situations in infants. The oral food challenge (OFC) test is the gold standard for diagnosing food allergies, but it must be performed by experienced physicians and carries the risk of inducing severe allergic reactions, making it difficult to implement in primary care hospitals. Clinical history is readily available diagnostic data in primary care hospitals, but the value of different clinical histories is currently difficult to quantify.

[0003] Previous studies have revealed the indicative significance of certain symptoms (such as growth retardation) and risk factors (such as a family history of allergic diseases) for food allergies, but there is no diagnostic tool that integrates these clinical data. [3,4] In 2014, a European expert conference developed the Cow's Milk Related Symptom Score (CoMiSS), which comprises digestive, respiratory, and skin symptoms, as well as crying. The values ​​assigned to different symptoms are determined collaboratively by experts. [5] CoMiSS has improved clinicians' awareness of cow's milk protein allergy (CMPA) and simplified the diagnostic process. However, studies in China and other countries have shown that CoMiSS has an AUC of less than 0.75 for diagnosing CMPA and low diagnostic sensitivity. [6-8] This indicates that its application value in Asian countries is limited, and CoMiSS is only used to identify CMPA patients whose allergen is milk protein, and has no identification effect on FA patients with other types of food allergies.

[0004] In recent years, some food allergy diagnostic models based on risk factors have been explored. A 2019 Chinese study established a food allergy risk score for infants and young children based on multiple risk factors, including family history of allergic diseases, cesarean section, pet feeding, and tobacco exposure. This score model may have some predictive value for food allergies, but it has not been validated. [9]A large-scale Japanese study in 2021 collected data through questionnaires and established and validated a food allergy prediction model based on risk factors. Multivariate logistic regression analysis showed that high-risk birth months (August-December), first child, eczema, family history of atopic dermatitis on both parents, and family history of food allergies on both mother and siblings were independent risk factors for food allergy (P = 0.003). Based on these factors, a logistic prediction model for food allergy was established, with an AUC of 0.75 on both the training and validation sets, indicating that the model has some predictive ability. However, in this study, the diagnosis of food allergy was completed by family members, and it is unclear whether the diagnosis was based on the gold standard OFC (Obstructive Cognition), thus its diagnostic value is debatable.

[10] In 2021, a Japanese study established a food allergy prediction model based on risk factors. This model had an AUC of 0.75. However, the study did not use OFC (Oil-Free Concentration) as the gold standard for diagnosing food allergies, therefore the true value of the prediction model obtained in this study remains to be seen.

[10] A 2023 study developed a model to predict OFC outcomes, but the resulting model had an AUC of ≤0.68, indicating poor diagnostic efficacy.

[11] Therefore, there is currently no validated diagnostic model for infant food allergies with good diagnostic efficacy. Summary of the Invention

[0005] To address the problems of invasiveness, reliance on experience, and difficulty in quantification in existing diagnostic methods for infant food allergies, this invention provides a non-invasive diagnostic model for infant food allergies that can identify food allergies caused by various allergens. This model integrates common clinical manifestations and risk factors in infants, employing statistical methods to assign objective weights to each variable, facilitating clinical application.

[0006] The technical solution of this invention includes: establishing an infant food allergy diagnostic model based on clinical data. The model includes six variables: severity of infant eczema, stool characteristics, rectal bleeding, respiratory symptoms, growth retardation, and family history of allergic diseases. Specifically, by performing multivariate logistic regression analysis on the training set, the regression coefficients of each variable are obtained as weights, and a scoring formula is formed (see example). When using this model to assess infants, the presence and severity of eczema are assessed according to the SCORAD scale, and the stool is assessed as formed, loose, or watery using the Brussels Infant Stool Scale. The presence or absence of rectal bleeding, respiratory symptoms, growth retardation, and family history are recorded. Each indicator is substituted into the model formula to calculate the score. If the score reaches or exceeds a preset threshold (3.19 points), the infant is judged to have a possible food allergy.

[0007] Specifically, the present invention provides the following technical solution:

[0008] A first aspect of the present invention provides a system for assessing the risk of food allergies in infants, the system comprising: an input module for receiving clinical indicators of the infant, including the severity of eczema, stool characteristics, rectal bleeding, respiratory symptoms, growth retardation, and family history of allergic diseases;

[0009] The calculation module calculates the infant food allergy diagnostic model score based on the clinical indicators according to the following formula:

[0010] Score = 0.23 × Eczema (mild) + 1.02 × Eczema (moderate) + 1.35 × Eczema (severe) + 0.47 × Stool (loose stool) + 1.42 × Stool (watery stool) + 1.60 × Rectal bleeding + 0.78 × Growth retardation + 1.19 × Respiratory symptoms + 1.36 × Family history of allergic diseases;

[0011] The output module is used to determine whether an infant has a food allergy based on the calculated score.

[0012] In one embodiment, the severity of the eczema is determined according to the SCORAD assessment scale, including four levels: "no eczema", "mild", "moderate" and "severe"; the stool characteristics are determined according to the Brussels Infant Stool Scale, including three categories: "formed or soft stool", "loose stool" and "watery stool"; the rectal bleeding, respiratory symptoms, growth retardation and family history of allergic diseases are binary indicators, and their presence or absence is recorded as 1 or 0.

[0013] In one implementation, the model threshold is set to 3.19, and the output module determines that the infant has a food allergy when the model score is ≥3.19.

[0014] In one implementation, the input module includes a user interface or interface for collecting consultation information, and the system further includes a storage module for storing model weights and thresholds.

[0015] In one implementation, the computing module includes a microprocessor or computing unit configured to retrieve model parameters from a storage module and perform a scoring operation.

[0016] In one embodiment, a display module is also included for displaying the infant's model score and evaluation results in real time.

[0017] A second aspect of the present invention provides a diagnostic model for assessing the risk of food allergies in infants, the model calculating a diagnostic model score for infant food allergies based on clinical indicators according to the following formula:

[0018] Score = 0.23 × Eczema (mild) + 1.02 × Eczema (moderate) + 1.35 × Eczema (severe) + 0.47 × Stool (loose stool) + 1.42 ×

[0019] Stool (watery stool) + 1.60 × bloody stool + 0.78 × growth retardation + 1.19 × respiratory symptoms + 1.36 × family history of allergic diseases;

[0020] The model threshold is 3.19. When the model score is ≥3.19, it is determined that the infant may have a food allergy.

[0021] In one embodiment, the severity of the eczema is determined according to the SCORAD assessment scale, including four levels: "no eczema", "mild", "moderate" and "severe"; the stool characteristics are determined according to the Brussels Infant Stool Scale, including three categories: "formed or soft stool", "loose stool" and "watery stool"; the rectal bleeding, respiratory symptoms, growth retardation and family history of allergic diseases are binary indicators, and their presence or absence is recorded as 1 or 0.

[0022] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0023] 1. Non-invasive diagnosis: All variables used are clinically readily available information obtained through history taking or physical manifestations (such as eczema, stool characteristics, etc.), without involving blood draws or other invasive procedures, resulting in high infant compliance.

[0024] 2. The weights of each variable were determined based on statistical results from a sample of Chinese infants, and the evaluation criteria were standardized (using validated scales such as SCORAD and the Brussels Infant Stool Scale) to avoid relying solely on subjective judgment based on experience.

[0025] 3. Excellent diagnostic performance: The model has been validated internally and externally. The training set ROC curve shows an AUC of 0.86, sensitivity of 0.72, and specificity of 0.88; the external validation test set shows an AUC of 0.87, sensitivity of 0.72, and specificity of 0.89. The calibration curve closely matches the ideal situation, with a Brier score of 0.143, indicating high predictive accuracy. The decision curve shows that applying this model over a wide range of thresholds yields high clinical net benefits.

[0026] 4. High Promotional Value: The model is easy to operate, requiring only the collection of medical history and basic physical examination information to calculate a score, making it suitable for widespread use in primary hospitals and outpatient clinics. Furthermore, this invention can be further implemented as a calculator or mobile app to assist doctors and parents in conducting rapid assessments. Attached Figure Description

[0027] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0028] Figure 1This diagram illustrates the enrollment and analysis process of the training and test sets in this invention. The training set included 572 infants suspected of food allergies (286 in the FA group and 286 in the non-FA group) for model building and internal validation; the test set included 260 infants (137 in the FA group and 123 in the non-FA group) for external validation. The enrollment, grouping, and analysis process are shown in the figure.

[0029] Figure 2 Receiver operating characteristic (ROC) curve for an infant food allergy diagnostic model. Figure 2 (Left) is the ROC curve of the training set; Figure 2 (Middle) is the ROC curve for 10-fold cross-validation (internal validation); Figure 2 (Right) shows the ROC curve for external validation on the test set. The training set model had an AUC of 0.86, an optimal diagnostic threshold of 3.19, a sensitivity of 0.72, and a specificity of 0.88; the external validation on the test set had an AUC of 0.87, a sensitivity of 0.72, and a specificity of 0.89.

[0030] Figure 3 A schematic diagram of the calibration curve and decision curve for a diagnostic model of infant food allergies. Figure 3 (Left) The calibration curve shows that the predicted probability is close to the actual probability, indicating that the model is well calibrated; Figure 3 (Right) The decision curve shows that significant net benefits can be obtained by clinical intervention based on this model within a wide threshold range.

[0031] Figure 4 This is a schematic diagram of the SCORAD scale for atopic dermatitis (left image) and the Brussels Infant Stool Scale (right image), used to assess the severity of infantile eczema and stool characteristics. The SCORAD scale scores based on the area of ​​eczema affected and the intensity of inflammation; the Brussels Infant Stool Scale categorizes infantile stool into formed, loose, and watery stools based on appearance, for a more consistent assessment of stool characteristics. Detailed Implementation

[0032] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0033] Example 1: Data Collection and Model Building

[0034] (1) Collection of clinical data and enrollment of study subjects

[0035] Infants with suspected food allergies who visited Peking University Third Hospital between January 2022 and September 2023 were selected as the training set, and clinical data were collected. Infants were divided into a food allergy (FA) group and a non-FA group based on the gold standard OFC results. A preliminary diagnostic model for infant food allergies was established using the training set data, and internal validation was conducted to evaluate the model's reproducibility and diagnostic efficacy. Infants with suspected food allergies who visited Peking University Third Hospital between January 2016 and December 2021 were selected as the test set. The diagnostic model established based on the training set data and the food allergy diagnosis made based on OFC results were compared, i.e., external validation was conducted to further evaluate the application value and generalization value of the diagnostic model to other clinical scenarios.

[0036] In the preliminary stage of the study, common clinical manifestations and risk factors of infant food allergies were identified through literature review. Detailed clinical data were obtained through outpatient consultations, medical record reviews, and regular follow-ups, and case report forms were completed. The collected data included: 1) Demographic data: age, sex, gestational age, delivery method, length, and weight; 2) Symptoms and severity in various systems: ① Skin: eczema, Scoring Atopic Dermatitis (SCORAD) score; ② Digestive system: vomiting, regurgitation, constipation, hematochezia, mucus in stool, feeding difficulties, growth retardation, defecation frequency, stool characteristics, and the Brussels Infant Stool Scale; ③ Respiratory system: runny nose, nasal congestion, sneezing, coughing, wheezing, etc.; ④ Poor comfort: repeated eye rubbing, crying, and frequent nighttime awakenings; 3) Risk factors: ① Genetic factors, such as family history of allergic diseases and autoimmune diseases; ② Environmental factors: including history of infection and antibiotic exposure. All data were collected in accordance with internationally recognized guidelines and diagnostic criteria.

[0037] The enrollment information of the study subjects can be found in [link to study participant information]. Figure 1 The training set (discovery set) included 572 infants, with 286 in the FA group and 286 in the non-FA group; the test set (validation set) included 260 infants, with 137 in the FA group and 123 in the non-FA group. The inclusion and grouping process for the study subjects can be found in [link to relevant documentation]. Figure 1 The ratio of training set to test set sample size is 2.2, which is in line with expectations. The training set data is used to initially build the diagnostic model and perform internal model validation, while the test set data is used for external validation to test the model's generalization and application capabilities.

[0038] (2) Univariate regression analysis was used to preliminarily screen variables for the diagnostic model of infant food allergy.

[0039] The age, sex, and gestational age of the FA and non-FA groups were comparable and showed no statistically significant differences. In the training set, infants in the FA group had slightly shorter birth lengths and slightly higher birth weights than those in the non-FA group. Regarding clinical symptoms, infants in the FA group were more likely than those in the non-FA group to exhibit eczema, vomiting, regurgitation, increased bowel frequency, abnormal stool consistency, bloody stools, mucus stools, growth retardation, respiratory symptoms, and poor comfort (all P > 0.05), while there were no significant differences in constipation and feeding difficulties between the two groups. Regarding risk factors, infants in the FA group had a higher probability of having a family history of allergic diseases and autoimmune diseases than those in the non-FA group, with statistically significant differences between the groups.

[0040] (3) Multivariate regression analysis to establish a diagnostic model for infant food allergy

[0041] Using the occurrence of food allergies as the dependent variable, and incorporating variables with significant differences between groups from the univariate regression analysis above as independent variables, a stepwise binary multivariate logistic regression analysis was first used to screen model variables and establish a diagnostic model for infant food allergies. An initial diagnostic model for infant food allergies was established using R 4.3.2 (see Table 1).

[0042] Table 1 - Results of binary multivariate logistic regression analysis of the training set and diagnostic model for infant food allergy

[0043]

[0044] In Table 1, the β values ​​for different variables represent their respective weights. Since the constant term is fixed, it was not included in the formula for ease of calculation. The final formula for the model is as follows:

[0045] Infant food allergy diagnostic model score = 0.23 × eczema (mild) + 1.02 × eczema (moderate) + 1.35 × eczema Rash (severe) +0.47 × Stool consistency (loose stool) +1.42 × Stool consistency (watery stool) +1.60 × Rectal bleeding +0.78 × Growth retardation Slower symptoms +1.19 × Respiratory symptoms +1.36 × Family history of allergic diseases

[0046] In this model, the severity of eczema is assessed using the SCORAD scale, stool characteristics are assessed using the Brussels Infant Stool Scale, and the presence or absence of other variables is determined according to the symptom definition section above; the presence of a variable is recorded as 1, and the absence of a variable is recorded as 0. The assessment results of each variable are substituted into the above formula to calculate the infant food allergy diagnostic model score. A score of 3.19 indicates the highest Youden's index for diagnosis, meaning that an infant food allergy diagnostic model score ≥3.19 is considered a food allergy, achieving the best diagnostic effect.

[0047] Example 2: Model Evaluation and Validation

[0048] First, the accuracy of the food allergy diagnostic model used on the training set of infants was evaluated using ROC curves, see [link to ROC curve analysis]. Figure 2(Left). The AUC of the training set ROC curve was 0.86 (0.83–0.90), the optimal threshold for diagnosing food allergies was 3.19, the sensitivity for diagnosing food allergies was 0.72 (0.71–0.73), and the specificity reached 0.88 (0.87–0.89). This indicates that the infant food allergy diagnostic model has good diagnostic efficacy and high specificity, and has good differential diagnostic ability. The Hosmer-Lemeshow test showed that the diagnostic model established based on the training set data had a good fit (χ2 = 11.30, P = 0.19) and the model was stable.

[0049] Then, the model was internally validated using a machine learning 10-fold cross-validation method, see [link / reference]. Figure 2 (Chinese). The AUC of the internal validation was 0.86 (0.76–0.95), the sensitivity for diagnosing food allergies was 0.72 (0.66–0.78), and the specificity was 0.92 (0.88–0.95), which were largely consistent with the training set, indicating that the diagnostic model for infant food allergies was stable. External validation of the diagnostic model was conducted using test set data to evaluate its diagnostic value in other clinical scenarios and to assess its potential for widespread application. The external validation results are shown below. Figure 2 (Right): The AUC of the test set was 0.87 (0.83–0.92), the sensitivity was 0.72 (0.65–0.80), and the specificity was 0.89 (0.83–0.95). The optimal threshold for diagnosing food allergies when the Youden index was at its maximum was 3.09 points, which was not significantly different from the training set. This indicates that the infant food allergy diagnostic model still has high diagnostic accuracy when applied to other clinical scenarios, and the diagnostic thresholds in different clinical scenarios are not significantly different, demonstrating the model's good applicability. The Hosmer-Lemeshow test showed that the diagnostic model had a good fit on the test set (χ2 = 10.95, P = 0.20), indicating model stability.

[0050] Finally, the predictive accuracy and net clinical benefit of the diagnostic model were evaluated using calibration and decision curves:

[0051] ① The calibration curve is shown below. Figure 3 (Left): The predicted probability curve of the food allergy diagnostic model is very close to the ideal situation, indicating that the diagnostic model obtained from clinical research has good calibration for the diagnosis of food allergies; Boolean score is used to evaluate the difference between the evaluation results of the diagnostic model and the actual results. The smaller the better. The Boolean score of the infant food allergy diagnostic model is 0.143, which suggests that the model has good accuracy in predicting the occurrence of food allergies and is close to the actual situation.

[0052] ② A good diagnostic model should aid in clinical medical decision-making. The clinical benefits of using a food allergy diagnostic model to determine whether to conduct clinical intervention can be assessed by plotting decision curves. (See...) Figure 3 (Right): The results show that when the threshold probability of the training set is 21%-98%, clinical intervention based on the food allergy diagnostic model can achieve high net clinical benefits.

[0053] Example 3: How to use the model

[0054] (1) Obtain the clinical data required for the diagnosis of infant food allergy and assess the severity.

[0055] The symptoms recorded in the clinical section represent the most severe stages of the disease. Detailed symptom definitions and assessments are as follows:

[0056] ① The Chinese Atopic Dermatitis SCORAD Scale (see [link to eczema assessment]) is used as a reference for eczema assessment. Figure 4 Severity was assessed according to guidelines: no eczema (SCORAD 0 points), mild eczema (SCORAD 0–24 points), moderate eczema (SCORAD 25–50 points) and severe eczema (SCORAD >50 points)

[12] .

[0057] ② Stool characteristics are described according to the Brussels Infant Stool Scale.

[13] Formed or soft stools (BSFS type 4), loose stools (BSFS types 5 and 6), and watery stools (BSFS type 7).

[0058] ③Hematochezia includes visible blood in the stool and a positive fecal occult blood test. Abnormal stool color alone without a negative occult blood test is not considered hematochezia.

[0059] ④ Growth retardation: Must meet at least one of the following criteria: ① Age-specific weight below the third percentile for children of the same sex in two consecutive monitoring sessions; ② Height-specific weight below the fifth percentile for children of the same age and sex in two consecutive monitoring sessions; ③ Weight loss exceeding two percentiles on the age-specific weight percentile chart in two consecutive monitoring sessions. Percentile charts for height, weight, height-specific weight, and age-specific weight are sourced from the official website of the World Health Organization.

[14] .

[0060] ⑤ Respiratory symptoms: Persistent cough, wheezing, sneezing, runny nose, or nasal congestion may indicate allergic diseases.

[15] Short-term respiratory symptoms may not be related to food allergies; respiratory symptoms associated with food allergies are often chronic. Only respiratory symptoms lasting more than one week will be recorded.

[16] .

[0061] ⑥ Family history of allergic diseases: including but not limited to allergic rhinitis, food allergies, allergic asthma, and other allergic family history.

[0062] (2) Determine whether it is a food allergy based on the infant food allergy diagnostic model formula.

[0063]

[0064] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

[0065] References:

[0066] [1]Peters RL,Koplin JJ,Gurrin LC,et al.The prevalence of foodallergy and other allergic diseases in early childhood in a population-basedstudy:HealthNuts age 4-year follow-up[J].J Allergy Clin Immun,2017,140(1):145-+.

[0067] [2]Spolidoro GCI,Amera YT,Ali MM,et al.Frequency of food allergyin Europe:An updated systematic review and meta-analysis[J].Allergy,2023,78(2):351-68.

[0068] [3] Saito-Abe M, Yamamoto-Hanada K, Pak K, et al. How a Family History ofAllergic Diseases Influences Food Allergy in Children: The Japan Environmentand Children's Study[J].Nutrients,2022,14(20)

[0069] [4]Diaferio L,Caimmi D,Verga M C,et al.May Failure to Thrive inInfants Be a Clinical Marker for the Early Diagnosis of Cow's Milk Allergy?[J].Nutrients,2020,12(2).

[0070] [5]Saad K,Elgenidy A,Atef M,et al.Cow's Milk-related Symptom Scorefor cow's milk allergy assessment:a meta-analysis for test accuracy[J].Pediatr Res,2023,93(4):772-9.

[0071] [6]El-Shafie A M,Omar Z A,El Zefzaf H M S,et al.Evaluation of Cow'sMilk Related Symptom Score[CoMiSS]accuracy in cow's milk allergy diagnosis[J].Pediatr Res,2023.

[0072] [7]Bajerova K,Salvatore S,Dupont C,et al.The Cow's Milk-RelatedSymptom Score(CoMiSS()):A Useful Awareness Tool[J].Nutrients,2022,14(10).

[0073] [8]Vandenplas Y,Zhao Z Y,Mukherjee R,et al.Assessment of the Cow'sMilk-related Symptom Score(CoMiSS)as a diagnostic tool for cow's milk proteinallergy:a prospective,multicentre study in China(MOSAIC study)[J].Bmj Open,2022,12(2).

[0074] [9] Bai Pingping, Xing Mengjuan, Hu Yan. Exploration of methods for early prediction of high-risk children with food allergies [J]. Journal of Clinical Pediatrics, 2019, 37(12):936-9.

[0075]

[10] Sugiura S,Hiramitsu Y,Futamura M,et al.Development of aprediction model for infants at high risk of food allergy[J].Asia PacAllergy,2021,11(1).

[0076]

[11] Zhang J, Lee D, Jungles K, et al. Prediction of oral food challenge outcomes via ensemble learning [J]. Informatics in Medicine Unlocked, 2023, 36.

[0077]

[12] Yao

[0078]

[13] Huysentruyt K,Koppen I,Benninga M,et al.The Brussels Infant and Toddler Stool Scale: A Study on Interobserver Reliability[J].J Pediatr GastrNutr,2019,68(2):207-13.

[0079]

[14] Diaferio L, Caimmi D, Verga MC, et al. May Failure to Thrive in Infants Be a Clinical Marker for the Early Diagnosis of Cow's Milk Allergy? [J].Nutrients,2020,12(2).

[0080]

[15] Fiocchi A,Brozek J,Schunemann H,et al.World Allergy Organization(WAO)Diagnosis and Rationale for Action against Cow's Milk Allergy(DRACMA)Guidelines[J].Pediat Allerg Imm-Uk,2010,21:1-125.

[0081]

[16] Fiocchi A,Bognanni A,Brozek J,et al.World Allergy Organization(WAO)Diagnosis and Rationale for Action against Cow's Milk Allergy(DRACMA)Guidelines update-I-Plan and definitions[J].World Allergy Organ,2022,15(1).

Claims

1. A system for assessing the risk of food allergies in infants, characterized in that, The system includes: The input module is used to receive the infant's clinical indicators, including the severity of eczema, stool characteristics, rectal bleeding, respiratory symptoms, growth retardation, and family history of allergic diseases. The calculation module calculates the infant food allergy diagnostic model score based on the clinical indicators according to the following formula: Score = 0.23 × Eczema (mild) + 1.02 × Eczema (moderate) + 1.35 × Eczema (severe) + 0.47 × Stool (loose stool) + 1.42 × Stool (watery stool) + 1.60 × Rectal bleeding + 0.78 × Growth retardation + 1.19 × Respiratory symptoms + 1.36 × Family history of allergic diseases; The output module is used to determine whether an infant has a food allergy based on the calculated score.

2. The system according to claim 1, wherein the severity of eczema is determined according to the SCORing Atopic Dermatitis (SCORAD) scale, including four levels: "no eczema", "mild", "moderate" and "severe"; the stool characteristics are determined according to the Brussels Infant Stool Scale, including three categories: "formed or soft stool", "loose stool" and "watery stool"; the hematochezia, respiratory symptoms, growth retardation and family history of allergic diseases are binary indicators, and their presence or absence is recorded as 1 or 0.

3. The system according to claim 1, wherein the model threshold is set to 3.19, and the output module determines that the infant may have a food allergy when the model score is ≥3.

19.

4. The system according to claim 1, wherein the input module includes a user interface or interface for collecting consultation information, and the system further includes a storage module for storing model weights and thresholds.

5. The system according to claim 1, wherein the computing module includes a microprocessor or computing unit configured to call model parameters from the storage module and perform scoring calculations.

6. The system according to claim 1, further comprising a display module for displaying the infant's model score and evaluation results in real time.

7. A diagnostic model for assessing the risk of food allergies in infants, characterized in that, The model calculates the infant food allergy diagnostic model score based on clinical indicators using the following formula: Score = 0.23 × Eczema (mild) + 1.02 × Eczema (moderate) + 1.35 × Eczema (severe) + 0.47 × Stool (loose stool) + 1.42 × Stool (watery stool) + 1.60 × Rectal bleeding + 0.78 × Growth retardation + 1.19 × Respiratory symptoms + 1.36 × Family history of allergic diseases; The model threshold is 3.

19. When the model score is ≥3.19, it is determined that the infant may have a food allergy.

8. The diagnostic model as described in claim 7, characterized in that, The severity of eczema was determined according to the SCORAD assessment scale, including four levels: "no eczema", "mild", "moderate" and "severe". The stool characteristics were determined according to the Brussels Infant Stool Scale, including three categories: "formed or soft stool", "loose stool" and "watery stool". The blood in stool, respiratory symptoms, growth retardation and family history of allergic diseases were binary indicators, and their presence or absence was recorded as 1 or 0.

Images