CN-122000051-A - System and computer storage medium for endometriosis diagnosis

CN122000051ACN 122000051 ACN122000051 ACN 122000051ACN-122000051-A

Abstract

A system and computer storage medium for endometriosis diagnosis are disclosed. Aiming at the important adjustment of endometriosis diagnosis standard, the invention detects biomarker and ploidy information of circulating endometriosis cells in the patient by collecting easily available samples such as peripheral blood of the patient, and utilizes an optimization algorithm model to carry out comprehensive risk assessment, provides an objective and quantifiable noninvasive diagnosis tool which is suitable for the adjusted standard, solves the problem of missed diagnosis caused by false negative in the existing operation diagnosis method, and has great clinical application value, social benefit and market prospect for improving the overall diagnosis level of endometriosis.

Inventors

CHANG XIAOHONG
WANG JIANLIU
LIU HUIPING
ZHANG CHEN
ZHOU XIAOHUA
ZHU HONGLAN
CHENG HONGYAN
YE XUE
LI YI
WANG SHANG
LI BUYUN
Tai Qianchen

Assignees

北京大学人民医院

Dates

Publication Date: 20260508
Application Date: 20260119

Claims (10)

1. A system for the diagnosis of endometriosis, comprising: a data acquisition unit for acquiring biomarker information and ploidy information of at least circulating endometrial cells in a sample of a subject; a risk score calculation unit for calculating based on the cell count of different biomarkers and the model of the cell count of different ploidy information to obtain a risk score result; A judging unit for comparing the risk score result with a diagnostic threshold value and outputting a risk of the subject for endometriosis; Wherein the model is: Y=a (er+) +b (cd10+) +c (bp+) +d (cec1) +e (cec2) +f (cec3) +j (cec4) +h (cec5+), where Y represents a risk score value, a-h each represents a coefficient of a corresponding variable, er+ represents a cell number positive for estrogen receptor, cd10+ represents a cell number positive for CD10 molecule, bp+ represents a cell number positive for both estrogen receptor and CD10 molecule, CEC1 represents a cell number of haploid, CEC2 represents a cell number of diploid, CEC3 represents a cell number of triploid, CEC4 represents a cell number of tetraploid, CEC5 represents a cell number of five-fold or more.
2. The system for diagnosis of endometriosis according to claim 1, characterized in that the coefficients in the model are obtained by calculating the diagnostic accuracy and disease prevalence without gold standards based on a binary potential class model framework.
3. The system for endometriosis diagnosis of claim 1, wherein the threshold is derived by plotting a ROC curve and maximizing according to a about log index.
4. The system for diagnosis of endometriosis according to claim 1, characterized in that the model is: Y=0.133×(ER+)+0.400×(CD10+)+0.081×(BP+)+0.479×(CEC1)+0.545×(CEC2)+0.554×(CEC3)+0.239×(CEC4)+0.346×(CEC5+).
5. The system for diagnosis of endometriosis according to claim 1, wherein the determination unit is configured to make a determination that the subject is classified as a high probability patient of endometriosis when the risk score is greater than the diagnostic threshold.
6. The system for diagnosis of endometriosis according to claim 1, further comprising a display unit for displaying the result of the judgment.
7. The system for endometriosis diagnosis of claim 1, wherein the data acquisition unit is communicatively connected to a hospital information system or autonomously entered by a subject for selectively acquiring biomarker information and ploidy information of circulating endometrial cells from raw clinical data of a hospital or data autonomously entered by a subject.
8. A method for constructing a diagnostic model of endometriosis, comprising the step of calculating the diagnostic accuracy and prevalence of the disease without perfect gold standards, based on a binary latent class model framework: (1) Symbol and data setting Considering an observation dataset with a sample size of N, for each subject i (i=1..once., N) a K-term binary test, denoted T i1 ,…,T iK , covariate vector X i , containing a set of CEC-related measurements including information on biomarkers er+, cd10+, bp+, and ploidy information; (2) Parameter definition Assuming that there is a binary potential class D i representing a true disease state, where D i =1 represents a disease, i.e. endometriosis, d=0 represents a non-disease, the key parameters are defined as follows: Sensitivity (SE k ) probability of the kth test result being positive in case the subject suffers from a disease: SE k =P(T iK =1|D i =1); Specificity (SP k ) probability of negative k test results in the absence of disease in the subject: SP k =P(T iK =0 |D i =0); (3) Independence assumption Assuming that the results of the K binary tests (T i1 ,…,T iK ) are independent of each other under the condition of the true disease state D i ; (4) Disease prevalence modeling Modeling the probability of disease prevalence under covariate X using a logical connection function: where β is the vector of regression coefficients; (5) Likelihood function Likelihood functions for data of unknown parameters are constructed as a mixture of disease and non-disease distributions: , expanding probability terms of binary results (T ik epsilon {0,1 }), and writing the log likelihood as: ; (6) Calculation of A numerical optimization algorithm is adopted to maximize likelihood functions and obtain all parameters # ,..., , ,..., And ) Is used for the estimation of the estimated value of (a).
9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the method of claim 8 when the computer program is executed or the steps of: a data acquisition unit for acquiring biomarker information and ploidy information of at least circulating endometrial cells in a sample of a subject; a risk score calculation unit for calculating based on the cell count of different biomarkers and the model of the cell count of different ploidy information to obtain a risk score result; A judging unit for comparing the risk score result with a diagnostic threshold value and outputting a risk of the subject for endometriosis; Wherein the model is: Y=a (er+) +b (cd10+) +c (bp+) +d (cec1) +e (cec2) +f (cec3) +j (cec4) +h (cec5+), where Y represents a risk score value, a-h each represents a coefficient of a corresponding variable, er+ represents a cell number positive for estrogen receptor, cd10+ represents a cell number positive for CD10 molecule, bp+ represents a cell number positive for both estrogen receptor and CD10 molecule, CEC1 represents a cell number of haploid, CEC2 represents a cell number of diploid, CEC3 represents a cell number of triploid, CEC4 represents a cell number of tetraploid, CEC5 represents a cell number of five-fold or more.
10. A computer storage medium or cloud having stored therein a computer program which when executed by a computer performs the method of claim 8 or performs the steps of: a data acquisition unit for acquiring biomarker information and ploidy information of at least circulating endometrial cells in a sample of a subject; a risk score calculation unit for calculating based on the cell count of different biomarkers and the model of the cell count of different ploidy information to obtain a risk score result; A judging unit for comparing the risk score result with a diagnostic threshold value and outputting a risk of the subject for endometriosis; Wherein the model is: Y=a (er+) +b (cd10+) +c (bp+) +d (cec1) +e (cec2) +f (cec3) +j (cec4) +h (cec5+), where Y represents a risk score value, a-h each represents a coefficient of a corresponding variable, er+ represents a cell number positive for estrogen receptor, cd10+ represents a cell number positive for CD10 molecule, bp+ represents a cell number positive for both estrogen receptor and CD10 molecule, CEC1 represents a cell number of haploid, CEC2 represents a cell number of diploid, CEC3 represents a cell number of triploid, CEC4 represents a cell number of tetraploid, CEC5 represents a cell number of five-fold or more.

Description

System and computer storage medium for endometriosis diagnosis Technical Field The present invention relates to the field of disease diagnosis, and in particular to a system and a computer storage medium for endometriosis diagnosis. Background Endometriosis (EM) is a common gynaecological disease whose core pathology is manifested by abnormal growth of endometrial tissue at sites outside the uterine cavity. The disease often causes clinical problems such as chronic pelvic pain, dysmenorrhea, infertility, remarkably reduced quality of life and the like. Diagnosis of endometriosis includes clinical symptoms, imaging examination, serum marker detection, laparoscopic pathological biopsy judgment, etc. Among them, laparoscopic diagnosis is an invasive procedure, the diagnosis efficiency is highly dependent on the experience of the operator, and since the surgical field and the exploration scope are mainly concentrated in the pelvic region, the systematic evaluation of the lesions outside the pelvic region (such as the lung, the peritoneum, the mesentery, the diaphragm, or the distant peritoneum) is insufficient, and the recognition of atypical, tiny, or subperitoneal lesions is significantly limited, resulting in a clinically non-negligible false negative rate (i.e., missed diagnosis). With the deep understanding of its high rate of missed diagnosis (false negative) and the development of non-invasive diagnostic and therapeutic strategies, the international consensus of diagnosis and treatment has radically shifted. In particular, the latest guidelines of the European Society of Human Reproduction (ESHRE) in 2022 clearly state that endometriosis is no longer diagnosed as a single standard based on laparoscopic results. Diagnosis based on clinical symptoms often lacks specificity and is prone to misdiagnosis/missed diagnosis. Clinical symptoms are taken as a basic reference for diagnosis, but are affected by individual differences and focal sites/sizes, and have significant limitations. For example, the specificity is extremely low, and confusion with various diseases, such as core symptoms (dysmenorrhea, pelvic pain, menstrual abnormalities, infertility) is not characteristic of endometriosis. And the ratio of asymptomatic patients is high, and diagnosis is easy to be missed. About 20% -30% of endometriosis patients have no obvious symptoms, are only found by accident during infertility examination, gynecological operation or physical examination, especially early stage micro lesions (diameter <5 mm) or non-pelvic ectopic lesions (such as intestinal tract, urinary system), and are usually ignored due to atypical symptoms. In addition, subjectivity is strong, and quantization standards are lacking. Pain degree and abnormal menstrual flow are dependent on subjective description of patients, tolerance degree of different patients is large, and clinical evaluation is difficult to objectively and quantitatively evaluate through symptoms, so that diagnosis consistency is low. Diagnosis based on imaging examinations is affected by equipment accuracy, operator experience, and lesion characterization, with significantly shorter plates. For example, transvaginal ultrasound has a high rate of missed diagnosis. The missing diagnosis rate of the transparent/white scar-like focus reaches 6-13% for the peritoneal micro focus (diameter <5 mm), and the detection rate is only about 70% for the deep infiltration type focus. The magnetic resonance detection has high cost and long time consumption, and is not suitable for screening. Whereas computed tomography has very low specificity and is at risk of radiation exposure, it is rarely used for the examination of endometriosis. Diagnosis based on serum markers is generally based on CA125, CA199, EMAB and the like at present, and is only an auxiliary diagnosis means, has no diagnosis value, lacks specificity, and cannot be independently diagnosed at present, and particularly has poor specificity. For example, with CA125, except for EM, pelvic inflammatory disease, endometrial cancer, ovarian cancer, uterine fibroids, pregnancy, liver disease, etc. all cause the rise, and the CA125 positive rate in EM patients is only 40% -60%, which cannot be distinguished by single detection. For CA199 and EMAB, the specificity is lower, CA199 can also rise in gastrointestinal tract diseases and biliary tract diseases, the positive rate of EMAB is only 30% -50%, and the clinical diagnosis value is limited. In addition, the markers lack of effective unified threshold values, and the detection methods and reference standards of different laboratories are inconsistent, so that the marker results are poor in comparability. The information in the background section is only for the purpose of illustrating the general background of the invention and is not to be construed as an admission or any form of suggestion that such information forms the prior art that is well known to those of ordinary skill in the