CN118098520B - Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient - Google Patents
Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient Download PDFInfo
- Publication number
- CN118098520B CN118098520B CN202410489842.1A CN202410489842A CN118098520B CN 118098520 B CN118098520 B CN 118098520B CN 202410489842 A CN202410489842 A CN 202410489842A CN 118098520 B CN118098520 B CN 118098520B
- Authority
- CN
- China
- Prior art keywords
- tumor
- residual
- postoperative
- factor
- cell carcinoma
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000002980 postoperative effect Effects 0.000 title claims abstract description 36
- 206010061534 Oesophageal squamous cell carcinoma Diseases 0.000 title claims abstract description 24
- 208000036765 Squamous cell carcinoma of the esophagus Diseases 0.000 title claims abstract description 24
- 208000007276 esophageal squamous cell carcinoma Diseases 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 title claims abstract description 15
- 206010028980 Neoplasm Diseases 0.000 claims abstract description 48
- 208000035346 Margins of Excision Diseases 0.000 claims abstract description 27
- 208000007660 Residual Neoplasm Diseases 0.000 claims abstract description 25
- 238000007477 logistic regression Methods 0.000 claims abstract description 11
- 238000001356 surgical procedure Methods 0.000 claims abstract description 8
- 210000001519 tissue Anatomy 0.000 claims description 15
- 239000013598 vector Substances 0.000 claims description 11
- 239000011159 matrix material Substances 0.000 claims description 8
- 230000001186 cumulative effect Effects 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 201000011510 cancer Diseases 0.000 claims description 3
- 230000002792 vascular Effects 0.000 claims description 3
- 238000009795 derivation Methods 0.000 claims description 2
- 230000008595 infiltration Effects 0.000 claims description 2
- 238000001764 infiltration Methods 0.000 claims description 2
- 210000001165 lymph node Anatomy 0.000 claims description 2
- 210000005036 nerve Anatomy 0.000 claims description 2
- 238000000513 principal component analysis Methods 0.000 claims description 2
- 238000002271 resection Methods 0.000 claims description 2
- 230000007170 pathology Effects 0.000 claims 3
- 238000012847 principal component analysis method Methods 0.000 claims 2
- 238000004140 cleaning Methods 0.000 claims 1
- 230000009545 invasion Effects 0.000 claims 1
- 230000006870 function Effects 0.000 description 8
- 230000001575 pathological effect Effects 0.000 description 7
- 208000002847 Surgical Wound Diseases 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 2
- 206010066902 Surgical failure Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004027 cell Anatomy 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010837 poor prognosis Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 238000005406 washing Methods 0.000 description 1
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention relates to the technical field of postoperative incisal margin positive prediction, and provides a method for constructing a postoperative surgical incisal margin positive prediction model of an esophageal squamous cell carcinoma patient, which comprises the following steps: collecting historical full-cycle data of an esophageal squamous cell carcinoma patient subjected to operation, wherein indexes of the historical full-cycle data comprise photographed postoperative tissue pictures and postoperative scanning slices; identifying whether tumor residues exist in the postoperative tissue picture and the postoperative scanning slice, and dividing the identification result into three categories of complete excision, visible residual tumor under the mirror and macroscopic residual tumor; constructing a surgical margin positive prediction model based on a logistic regression algorithm by using the historical full-period data; the full cycle data of the new esophageal squamous cell carcinoma patient is entered into the surgical margin positive predictive model prior to surgery to analyze the risk of surgical margin positivity. According to the invention, a positive prediction model of the surgical margin is constructed through logistic regression, so that whether the esophageal squamous cell carcinoma has the risk of positive surgical margin after the operation is judged, and the pressure of doctors is relieved.
Description
Technical Field
The invention relates to the technical field of postoperative incisal margin positive prediction, in particular to a method for constructing a postoperative surgical incisal margin positive prediction model of esophageal squamous cell carcinoma patients.
Background
Surgical margin positive means that there are also cancerous cells at the margin of the incision where the tissue is cut, that is, there is a residual tumor tissue that has not been completely excised. The esophageal squamous cell carcinoma has high morbidity and mortality, and poor prognosis is often indicated if there is a margin tumor residue after surgery, and although positive surgical margin does not represent surgical failure, clinical avoidance is required. There are many factors that lead to the positive of the surgical margin of esophageal squamous cell carcinoma, and it is difficult for the doctor to visually find the residual tumor at the margin after the operation in the past, and although there are many factors that lead to the positive of the surgical margin, the doctor cannot clearly analyze the influence degree of various factors, and can only judge through experience. Therefore, it is difficult to control whether there is a positive risk of surgical incision after esophageal squamous cell carcinoma surgery.
Disclosure of Invention
The invention aims to solve two technical problems, namely the first technical problem of avoiding inaccuracy caused by naked eye observation and delineating a tumor residual target by a doctor, the second technical problem of judging whether the esophageal squamous cell carcinoma has a positive surgical margin or not after operation by logistic regression, relieving the working pressure of the doctor and providing a construction method of a positive prediction model of the surgical margin of the esophageal squamous cell carcinoma patient after operation.
In order to achieve the above object, the embodiment of the present invention provides the following technical solutions:
The method for constructing the positive predictive model of the postoperative surgical incisal margin of the esophageal squamous cell carcinoma patient comprises the following steps:
Step 1, collecting historical full-cycle data of an esophageal squamous cell carcinoma patient subjected to operation, wherein indexes of the historical full-cycle data comprise shot postoperative tissue pictures and postoperative scanning slices;
step 2, identifying whether tumor residues exist in the postoperative tissue picture and the postoperative scanning slice, and dividing the identification result into three categories of complete excision, visible residual tumor under the mirror and macroscopic residual tumor;
Step 3, constructing a surgical margin positive prediction model based on a logistic regression algorithm by using the historical full-period data;
Step 4, inputting full cycle data of a new esophageal squamous cell carcinoma patient into a positive predictive model of surgical margin before surgery to analyze the risk of positive surgical margin.
Compared with the prior art, the invention has the beneficial effects that:
According to the invention, whether a tumor residual target exists in a tissue picture or a scanning slice after operation or not is identified through a model constructed by a deep learning algorithm, so that inaccuracy caused by visual observation and sketching of the tumor residual target by a doctor is avoided; and (3) constructing a surgical margin positive prediction model through logistic regression, judging whether the esophageal squamous cell carcinoma has a risk of positive surgical margin after operation, and relieving the working pressure of doctors.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present invention.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Also, in the description of the present invention, the terms "first," "second," and the like are used merely to distinguish one from another, and are not to be construed as indicating or implying a relative importance or implying any actual such relationship or order between such entities or operations.
Example 1:
the invention is realized by the following technical scheme, as shown in figure 1, the method for constructing the positive prediction model of the postoperative surgical margin of the esophageal squamous cell carcinoma patient comprises the following steps:
step 1, collecting historical full-cycle data of an esophageal squamous cell carcinoma patient subjected to operation, wherein indexes of the historical full-cycle data comprise photographed postoperative tissue pictures and postoperative scanning slices.
The indicators of the historical full cycle data also include sex, age, functional status score, body mass index, tumor location, pathological T stage, pathological N stage, pathological TNM stage, number of lymph node washings, whether vascular infiltrates, whether nerves are invading, whether multifocal cancer is present, surgical mode, pre-operative treatment.
The post-operation tissue picture in the historical full-period data is shot by a high-resolution camera or is shot by a cavity mirror; post-operative scan slices are obtained by tomographic scanning of the patient post-operatively.
And 2, identifying whether tumor residues exist in the postoperative tissue picture and the postoperative scanning slice, and dividing the identification result into three categories of complete excision, microscopic residual tumor and macroscopic residual tumor.
Since the doctor in the past judges whether the surgical incision is positive or not through naked eyes or microscopic observation and relies on technicians to draw residual targets, but is influenced by human experiences, the judgment result may be inaccurate, and therefore, the step builds a surgical incision positive identification model based on a deep learning algorithm and is used for identifying the residual targets in postoperative tissue pictures and postoperative scanning slices.
A large number of post-operative tissue pictures and post-operative scan slices, which have been labeled with residual tumors, are retrieved as training sets from various medical databases. And training the ResNet-101 network model by adopting the ResNet-101 network model and using the training set and the labels thereof. Because the data in the training set may relate to privacy problems, the training set is encrypted by using a linear kernel function to obtain a kernel training set:
；
Wherein K represents a core training set; x i denotes the i-th sample data, and x i' denotes the i-th sample data after encryption; n is the total number of sample data. The binary response vector of x i is r i; constructing an objective function:
；
Wherein, Is an objective function; /(I)Representing minimizing the objective function; j epsilon [1, N ], r j is the binary response vector of the j-th sample data x j; /(I)、/>The i and j solution vectors are respectively; k (i, j) represents the elements of the ith row and the jth column in the core training set; /(I)Is a regularization parameter.
First-order derivation is performed on the objective function:
；
when training ResNet-101 network models, the loss function is minimized by combining the symmetrical tree as a predictor:
；
Recording device For/>, under the t-th iterationTaylor expansion is performed:
；
；
；
Wherein L (F (x i),yi) is a loss function, F (x i) represents a predictive label of sample data x i, y i represents a real label of sample data x i, and F i represents a leaf value loss function; Representation She Zhi; /(I) Representation pair/>Is a L2 canonical constraint of (2); /(I)、/>Representing regularization parameters; she Zhi/>The optimal solution of (a) is:
；
And (3) obtaining a surgical incising edge positive identification model after training, and identifying the postoperative tissue picture and the postoperative scanning slice in the historical full-period data collected in the step (1) by using the surgical incising edge positive identification model, so as to identify whether tumor residues exist or not, wherein the identification result is divided into three categories of complete excision, microscopic residual tumor and macroscopic residual tumor.
And 3, constructing a surgical margin positive prediction model based on a logistic regression algorithm by using the historical full-period data.
The collected historical full cycle data is shown in tables 1-1 and 1-2:
table 1-1 historical full period data (k=1, 2,., 5)
Table 1-2 historical full cycle data (k=6, 7,., 14)
Taking the index of the historical full-period data as a factor influencing the tumor residual category, and establishing a factor probability matrix P:
；
Wherein P k,r represents the probability of influence of the kth factor on the kth tumor residual category, k=1, 2,..14, e.g., k=1 represents the factor as gender; r=1, 2,3, r=1 indicates that the tumor residual category is complete resection, r=2 indicates that the tumor residual category is microscopic residual tumor, and r=3 indicates that the tumor residual category is macroscopic residual tumor.
The TOPSIS method is used for calculating the score index of each factor, the probability score index used by the TOPSIS method is shown in table 2, and the calculated factor score index is shown in tables 3-1 and 3-2:
TABLE 2 probability score indicator
Table 3-1 factor score index (k=1, 2.,. 5)
Table 3-2 factor score index (k=6, 7.,. 14)
The average score for all factors under each tumor residual category was calculated:
；
Wherein, Mean scores for all factors under category r tumor residual, r=1, 2,3; s cr represents the score of the c-th factor under the r-th tumor residual category, c=1, 2. If/>Then note S' cr = 0; if it isThe notation S' cr=1,S`cr indicates the influence degree score.
Solving the characteristic equation based on the factor probability matrix PWherein/>And B is a feature vector. Solving to obtain the characteristic value/>, of the k-th factorAnd feature vector/>, of class r tumor residual. Here, the general classes k, k=1, 2, 14, listed as factors in table 1-1 and table 1-2, 1, 2; subclass c, c=1, 2, listed as factors 3, 4 of tables 1-1 and 1-2; it will be readily appreciated that subclasses c1, c2 belong to the large class k1.
Principal component probability is calculated using principal component analysis:
；
calculating the cumulative contribution rate of the principal components:
；
Extracting main components:
；
wherein I k represents the principal component probability of the kth factor; i represents the cumulative contribution rate of the principal component; f r represents the main component of the r-th tumor residual category.
Constructing a regression model based on a logistic regression algorithm:
；
Wherein ln (hass r) represents a regression model of the r-th tumor residual class; Representing the intercept of the r-th regression model; /(I) Indicating that the r-th tumor residual category will/>S cr of (C); Indicating that the r-th tumor residual category will/> Is added at S cr.
And (3) fitting a regression relation by using SPSS software to obtain a positive prediction model of the surgical margin under each tumor residual category. The positive predictive model of surgical incision edge for complete excision is:
the surgical margin positive prediction model for the visible residual tumor under the lens is as follows:
the surgical incisal margin positive prediction model for macroscopic residual tumor is:
It can be seen that for the case of complete excision after surgery, pathological N-staging, vascular infiltration, and preoperative treatment are important; for the situation that residual tumor is visible under the postoperative scope, the tumor position, multifocal cancer, operation mode and preoperative treatment are important; for the situation that residual tumor is visible by naked eyes after operation, the tumor position, pathological T stage, pathological N stage, pathological TNM stage, operation mode and preoperative treatment are important. It is also demonstrated that preoperative treatment is important for the risk of residual tumor after surgery.
Step 4, inputting full cycle data of a new esophageal squamous cell carcinoma patient into a positive predictive model of surgical margin before surgery to analyze the risk of positive surgical margin.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (2)
1. The method for constructing the positive predictive model of the postoperative surgical incisal margin of the esophageal squamous cell carcinoma patient is characterized by comprising the following steps of: the method comprises the following steps:
Step 1, collecting historical full-cycle data of an esophageal squamous cell carcinoma patient subjected to operation, wherein indexes of the historical full-cycle data comprise shot postoperative tissue pictures and postoperative scanning slices;
step 2, identifying whether tumor residues exist in the postoperative tissue picture and the postoperative scanning slice, and dividing the identification result into three categories of complete excision, visible residual tumor under the mirror and macroscopic residual tumor;
the step 2 specifically comprises the following steps:
acquiring a plurality of post-operative tissue pictures and post-operative scan slices marked with residual tumors as training sets from various medical databases; encrypting the training set by using a linear kernel function to obtain a kernel training set:
；
Wherein K represents a core training set; x i denotes the i-th sample data, and x i' denotes the i-th sample data after encryption; n is the total number of sample data; the binary response vector of x i is r i; constructing an objective function:
；
Wherein, Is an objective function; /(I)Representing minimizing the objective function; j epsilon [1, N ], r j is the binary response vector of the j-th sample data x j; /(I)、/>The i and j solution vectors are respectively; k (i, j) represents the elements of the ith row and the jth column in the core training set; /(I)Is a regularization parameter;
first-order derivation is performed on the objective function:
；
when the core training set is used for training ResNet-101 network models, the symmetrical tree is combined as a predictor to minimize the loss function:
；
Recording device For/>, under the t-th iterationTaylor expansion is performed:
；
；
；
Wherein L (F (x i),yi) is a loss function, F (x i) represents a predictive label of sample data x i, y i represents a real label of sample data x i, and F i represents a leaf value loss function; Representation She Zhi; /(I) Representation pair/>Is a L2 canonical constraint of (2); /(I)、/>Representing regularization parameters; she Zhi/>The optimal solution of (a) is:
；
Obtaining a surgical incising edge positive identification model after training is completed, identifying a post-operation tissue picture and a post-operation scanning slice in the historical full-period data collected in the step 1 by using the surgical incising edge positive identification model, and identifying whether tumor residues exist or not, wherein the identification result is divided into three categories of complete excision, microscopic residual tumors and macroscopic residual tumors;
Step 3, constructing a surgical margin positive prediction model based on a logistic regression algorithm by using the historical full-period data;
the step 3 specifically comprises the following steps:
taking the index of the historical full-period data as a factor influencing the tumor residual category, and establishing a factor probability matrix;
calculating factor score indexes by using a TOPSIS method, and calculating average scores of all factors under each tumor residual category;
solving eigenvalues and eigenvectors based on the factor probability matrix;
extracting principal components by a principal component analysis method based on the eigenvalues and eigenvectors;
Constructing a regression model by a logistic regression algorithm based on the average score and the extracted principal components of all factors, and fitting a regression relation by using SPSS software to obtain a positive prediction model of the surgical margin under each tumor residual category;
The step of establishing a factor probability matrix by taking the index of the historical full-period data as a factor influencing the tumor residual category comprises the following steps:
taking the index of the historical full-period data as a factor influencing the tumor residual category, and establishing a factor probability matrix P:
；
Where P k,r represents the probability of impact of the kth factor on the kth tumor residual category, k=1, 2,..14; r=1, 2,3, r=1 indicates that the tumor residual category is complete resection, r=2 indicates that the tumor residual category is microscopic residual tumor, and r=3 indicates that the tumor residual category is macroscopic residual tumor;
the step of calculating factor score index by TOPSIS method and calculating average score of all factors under each tumor residual category comprises the following steps:
；
Wherein, Mean scores for all factors under category r tumor residual, r=1, 2,3; s cr represents the score of the c-th factor under the r-th tumor residual category, c=1, 2..38, and the 14 factors k include 38 factors c; if/>Then note S' cr = 0; if/>Then the notation S' cr=1,S`cr represents the impact level score;
The step of solving eigenvalues and eigenvectors based on the factor probability matrix comprises the following steps:
Solving the characteristic equation based on the factor probability matrix P Wherein/>The characteristic value is B, and the characteristic vector is B; solving to obtain the characteristic value/>, of the k-th factorAnd feature vector/>, of class r tumor residual；
The step of extracting the principal component by a principal component analysis method based on the feature value and the feature vector comprises the following steps:
Principal component probability is calculated using principal component analysis:
；
calculating the cumulative contribution rate of the principal components:
；
Extracting main components:
；
Wherein I k represents the principal component probability of the kth factor; i represents the cumulative contribution rate of the principal component; f r represents the main component of the r-th tumor residual category;
The step of constructing a regression model by a logistic regression algorithm based on the average score and the extracted principal components of all the factors comprises the following steps:
Constructing a regression model based on a logistic regression algorithm:
；
Wherein ln (hass r) represents a regression model of the r-th tumor residual class; Representing the intercept of the r-th regression model; Indicating that the r-th tumor residual category will/> S cr of (C); Indicating that the r-th tumor residual category will/> S cr of (C);
Step 4, inputting full cycle data of a new esophageal squamous cell carcinoma patient into a positive predictive model of surgical margin before surgery to analyze the risk of positive surgical margin.
2. The method for constructing a positive predictive model of postoperative surgical margin for patients with esophageal squamous cell carcinoma according to claim 1, wherein the method comprises the following steps: in the step 1, the indexes of the historical full cycle data also comprise sex, age, function state score, body quality index, tumor position, pathology T stage, pathology N stage, pathology TNM stage, lymph node cleaning number, vascular infiltration, nerve invasion, multifocal cancer, operation mode and preoperative treatment of the esophageal squamous cell carcinoma patient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410489842.1A CN118098520B (en) | 2024-04-23 | 2024-04-23 | Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410489842.1A CN118098520B (en) | 2024-04-23 | 2024-04-23 | Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118098520A CN118098520A (en) | 2024-05-28 |
| CN118098520B true CN118098520B (en) | 2024-06-21 |
Family
ID=91144229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410489842.1A Active CN118098520B (en) | 2024-04-23 | 2024-04-23 | Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118098520B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2708349A1 (en) * | 2001-09-18 | 2003-03-27 | Genentech, Inc. | Compositions and methods for the diagnosis and treatment of tumor |
| CN115830005A (en) * | 2023-01-09 | 2023-03-21 | 电子科技大学 | Prostate margin positive prediction method fusing 3D feature calculation |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11875903B2 (en) * | 2018-12-31 | 2024-01-16 | Tempus Labs, Inc. | Method and process for predicting and analyzing patient cohort response, progression, and survival |
| CN114913124B (en) * | 2022-04-13 | 2023-04-07 | 中南大学湘雅医院 | Incisional edge path generation method and system for tumor surgery and storage medium |
| CN117611806B (en) * | 2024-01-24 | 2024-04-12 | 北京航空航天大学 | Prostate cancer operation incisal margin positive prediction system based on images and clinical characteristics |
-
2024
- 2024-04-23 CN CN202410489842.1A patent/CN118098520B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CA2708349A1 (en) * | 2001-09-18 | 2003-03-27 | Genentech, Inc. | Compositions and methods for the diagnosis and treatment of tumor |
| CN115830005A (en) * | 2023-01-09 | 2023-03-21 | 电子科技大学 | Prostate margin positive prediction method fusing 3D feature calculation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118098520A (en) | 2024-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20190286652A1 (en) | Surgical video retrieval based on preoperative images | |
| US8503749B2 (en) | Method and system for the diagnosis of disease using retinal image content and an archive of diagnosed human patient data | |
| CN111936989A (en) | Similar medical image search | |
| Kauppi | Eye fundus image analysis for automatic detection of diabetic retinopathy | |
| CN113768461B (en) | Fundus image analysis method, fundus image analysis system and electronic equipment | |
| Hu et al. | Automated segmentation of geographic atrophy in fundus autofluorescence images using supervised pixel classification | |
| Zhang et al. | MRMR optimized classification for automatic glaucoma diagnosis | |
| Sikder et al. | Early blindness detection based on retinal images using ensemble learning | |
| Gupta et al. | An efficient model for detection and classification of internal eye diseases using deep learning | |
| US10430905B2 (en) | Case search device and method | |
| US20180365833A1 (en) | System and method for automated detection and measurement of corneal haze and demarcation line in optical coherence tomography images | |
| Mahadi et al. | Machine learning assisted decision support system for prediction of prostrate cancer | |
| Guru et al. | Brain tumor identification and classification using a novel extraction method based on adapted alexnet architecture | |
| CN118098520B (en) | Method for constructing positive predictive model of postoperative surgical incisal margin of esophageal squamous cell carcinoma patient | |
| US20150363930A1 (en) | Cost-sensitive linear reconstruction based optic cup localization | |
| Gunasinghe et al. | Comparison of pretrained feature extractors for glaucoma detection | |
| Al-Gburi et al. | Optical disk segmentation in human retina images with golden eagle optimizer | |
| Kumar et al. | Prediction of Lung Cancer Using Machine Learning Technique: A Survey | |
| Lincke et al. | AI-Based Decision-Support System for Diagnosing Acanthamoeba Keratitis Using In Vivo Confocal Microscopy Images | |
| Amiri et al. | Detection of topographic images of keratoconus disease using machine vision | |
| Theissen et al. | Learning cellular phenotypes through supervision | |
| Baiju et al. | Glaucoma and diabetic retinopathy diagnosis using image mining | |
| Kumari et al. | Automated process for retinal image segmentation and classification via deep learning based cnn model | |
| Shahalinejad et al. | Macular hole detection using a new hybrid method: using multilevel thresholding and derivation on optical coherence tomographic images | |
| Kuppusamy | An artificial intelligence formulation and the investigation of glaucoma in color fundus images by using BAT algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |