CN116259414B - Metastatic lymph node distinguishing model, construction method and application - Google Patents

Metastatic lymph node distinguishing model, construction method and application Download PDF

Info

Publication number
CN116259414B
CN116259414B CN202310511551.3A CN202310511551A CN116259414B CN 116259414 B CN116259414 B CN 116259414B CN 202310511551 A CN202310511551 A CN 202310511551A CN 116259414 B CN116259414 B CN 116259414B
Authority
CN
China
Prior art keywords
lymph nodes
dimensional
index
metastatic
aminolevulinic acid
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
Application number
CN202310511551.3A
Other languages
Chinese (zh)
Other versions
CN116259414A (en
Inventor
蔡惠明
王子阳
王毅庆
马陈
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Nuoyuan Medical Devices Co Ltd
Original Assignee
Nanjing Nuoyuan Medical Devices Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Nanjing Nuoyuan Medical Devices Co Ltd filed Critical Nanjing Nuoyuan Medical Devices Co Ltd
Priority to CN202310511551.3A priority Critical patent/CN116259414B/en
Publication of CN116259414A publication Critical patent/CN116259414A/en
Application granted granted Critical
Publication of CN116259414B publication Critical patent/CN116259414B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/0059—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence
    • A61B5/0071—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence by measuring fluorescence emission
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/003—Navigation within 3D models or images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images
    • G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2046—Tracking techniques
    • A61B2034/2055—Optical tracking systems
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B34/00—Computer-aided surgery; Manipulators or robots specially adapted for use in surgery
    • A61B34/20—Surgical navigation systems; Devices for tracking or guiding surgical instruments, e.g. for frameless stereotaxis
    • A61B2034/2046—Tracking techniques
    • A61B2034/2065—Tracking using image or pattern recognition
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2219/00—Indexing scheme for manipulating 3D models or images for computer graphics
    • G06T2219/20—Indexing scheme for editing of 3D models
    • G06T2219/2012—Colour editing, changing, or manipulating; Use of colour codes

Abstract

The invention relates to the technical field of fluorescence image navigation surgery, in particular to a metastatic lymph node distinguishing model, a construction method and application. The method for distinguishing the metastatic lymph nodes in real time and high efficiency by using the two-dimensional you index and the double fluorescent tracer (especially indocyanine green and 5-aminolevulinic acid) in combination has the following advantages: 1. the method can judge the benign and malignant properties of the lymph nodes in real time in the operation, and fills the blank of evaluating the lymph nodes in the traditional operation; 2. the method of the invention can keep benign lymph nodes as much as possible while ensuring no false negative, and is expected to improve the prognosis of patients and reduce postoperative complications; 3. the method is simple in calculation, programmable and capable of continuously optimizing model prediction accuracy along with accumulation of clinical data.

Description

Metastatic lymph node distinguishing model, construction method and application
Technical Field
The invention relates to the technical field of fluorescence image navigation surgery, in particular to a metastatic lymph node distinguishing model, a construction method and application.
Background
As an independent risk factor, residual metastatic lymph nodes can severely affect the prognosis of cancer patients in multiple types of cancers, such as melanoma, gastric, colon, breast, prostate and head and neck. However, current preoperative medical imaging techniques, such as MRI, CT, US or PET, do not accurately assist the surgeon in determining metastatic lymph nodes (sensitivity not exceeding 90%), and do not detect positive lymph nodes intraoperatively in real time. Because of the lack of specific identification methods for detecting metastatic lymph nodes either preoperatively or intraoperatively, surgeons often employ nonselective regional lymph node cleaning procedures in many procedures. While the broader lymph node dissection may increase the recurrence-free survival rate, especially for advanced cancer and the elderly, it also leads to a significant increase in the incidence of complications such as oedema, pain and infection and more post-operative deaths, ultimately leading to no significant improvement in overall survival. Thus, finding all metastatic lymph nodes in real time and retaining as many normal lymph nodes as possible is critical to improving the prognosis and quality of life of the patient.
Fluorescence navigation surgery (FGS) has been used as an emerging technique for tumor localization, tumor margin assessment and tumor metastasis localization, greatly promoting the development of accurate surgery. As for application in lymph node detection, sentinel lymph node localization is the most predominant method, mainly using indocyanine green (ICG) as fluorescent tracer. After injection into living tissue, ICG molecules bind to plasma proteins, which behave like macromolecules, are difficult to spread in tissue, but are easily drained through lymphatic vessels and remain in lymph nodes. However, this metastasis and retention process is non-selective between metastatic lymph nodes (metastatic lymph nodes) and normal lymph nodes, resulting in a sensitivity of detection of metastatic lymph nodes approaching 100%, but a specificity of 0%. Today, it is necessary to find 100% metastatic lymph nodes while retaining as many normal lymph nodes as possible, in other words, first obtaining 100% sensitivity and then obtaining as high specificity as possible. Naoki et al used the SBR-based near-infrared FGS technique for intraoperative detection of metastatic lymph nodes by using a near-infrared fluorescent dye conjugated with Panitumumab (EGFR antibody) with a sensitivity of 87.2% and a specificity of 86.1%. However, such active targeting tracers are only sensitive to those tumor types that highly express EGFR (such as head and neck cancer) and are expensive.
In view of this, the present invention has been made.
Disclosure of Invention
The invention aims to provide a model for distinguishing metastatic lymph nodes in real time and high efficiency in operation by using a two-dimensional you-on index and a dual-fluorescence tracer in combination and a construction method thereof, wherein the dual-fluorescence tracer with an advantage complementary relationship in an enrichment principle is used, and the two-dimensional you-on index is combined as a new threshold dividing index to improve the efficiency of distinguishing the metastatic lymph nodes from normal lymph nodes, so that the real-time high-efficiency distinguishing of the metastatic lymph nodes in operation is realized, 100% sensitivity (namely no false negative) is ensured, and meanwhile, as many normal lymph nodes are reserved as possible, and a new intra-operation means is provided for reducing postoperative recurrence and postoperative complications in a metastatic lymph node cleaning operation.
In order to solve the technical problems and achieve the purposes, the invention provides the following technical scheme:
in a first aspect, the present invention provides a method for constructing a lymph node metastasis risk assessment model, comprising: the lymph node metastasis risk assessment model adopts a double-fluorescence tracer to indicate the obtained two-dimensional you index to determine metastatic lymph nodes; the construction method comprises the following steps:
s1, acquiring training data: training (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section conventional pathology detection data of normal lymph nodes and metastatic lymph nodes in a sample;
s2, respectively arranging TBR thresholds obtained by calculating the spectral data of the dual fluorescent tracer obtained in the step S1 according to the order of magnitude as an abscissa axis and an ordinate axis, drawing a two-dimensional Euclidean index matrix chart, and taking a TBR threshold combination of the dual fluorescent tracer corresponding to the maximum value of the two-dimensional Euclidean index as an evaluation index of the metastatic lymph node;
the calculation formula of the TBR threshold value is as follows:
；
wherein TBR is the signal-to-back ratio of the dual fluorescent tracer per lymph node and INT () is a rounding function;
the two-dimensional you-den index calculation formula is:
；
wherein i=1, 2, …, total; j=1, 2, …, total;
TPR i,j true positive rate of lymph nodes judged under a certain two-dimensional threshold value is combined for the dual fluorescent tracer;
P G the number of metastatic lymph nodes judged for pathological diagnosis;
N IMI i,j the number of negative lymph nodes judged under a certain two-dimensional threshold value is used for the dual fluorescent tracer;
N G the number of negative lymph nodes judged by the pathological gold standard;
s3, acquiring verification data: verifying (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section routine pathology detection data of normal lymph nodes and metastatic lymph nodes in the sample;
and S4, adjusting the association model adopting the two-dimensional Euonyx index by using the verification data.
In an alternative embodiment, the dual fluorescent tracer comprises indocyanine green and 5-aminolevulinic acid.
In an alternative embodiment, the method for acquiring spectral data of the dual fluorescent tracer in step S1 includes injecting indocyanine green and 5-aminolevulinic acid into a training sample respectively, exposing lymph nodes in skin or fat of the training sample, acquiring fluorescence spectra of the lymph nodes (T) and surrounding normal tissues (B) by using an indocyanine green fluorescence spectrum probe and a 5-aminolevulinic acid fluorescence spectrum probe respectively, and further calculating respective signal-to-back ratios TBR of indocyanine green fluorescence and 5-aminolevulinic acid fluorescence.
Preferably, the training sample comprises a nude mouse or a human body.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid injected into the nude mice are 1-10 mg/kg and 100-500 mg/kg respectively.
Further preferably, the doses of indocyanine green and 5-aminolevulinic acid are injected into nude mice at 5.6 mg/kg and 250 mg/kg, respectively.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid administered to a human body are 0.2-2 mg/kg and 10-50 mg/kg, respectively.
It is further preferred that the doses of indocyanine green and 5-aminolevulinic acid administered to the human are 0.62mg/kg and 27.8 mg/kg, respectively.
Preferably, indocyanine green and 5-aminolevulinic acid are administered to the training sample, and the lymph nodes under the skin or in fat of the training sample are then exposed.
Preferably, the injection time of the 5-aminolevulinic acid is 10-30 hours later than the injection time of indocyanine green, and more preferably 20 hours later.
Preferably, the injection time of the 5-aminolevulinic acid is 2-6 hours before operation observation, more preferably 3-4 hours before operation;
preferably, the injection is by pushing blood through the peripheral vein using an intravenous pump.
Preferably, the mode of administration of 5-aminolevulinic acid for the human is oral.
In an alternative embodiment, the paraffin section is prepared by embedding and fixing the formalin-fixed isolated lymph node using paraffin, then slicing, and staining the slice.
Preferably, the concentration of formalin is 10%.
Preferably, the staining comprises HE staining.
In an alternative embodiment, the conventional pathology detection data comprises diagnostic results obtained by a pathologist based on paraffin section observations.
Preferably, the pathologist uses an inverted microscope to view paraffin sections.
In an alternative embodiment, in step S2, TPR i,j The calculation method of (1) is as follows:
the false negative rate FNR of the lymph node determined by the dual fluorescent tracer combination was first calculated according to the following formula:
；
wherein i and j represent matrix elements in the ith row and the jth column, total is Total number of lymph nodes, and P G Number of metastatic lymph nodes, noX, judged by pathological diagnosis i And NoY j Is the number of the ith and jth lymph nodes in the x-dimension and y-dimension, noM 1 、NoM 2 ……Is the number of metastatic lymph node judged by pathological diagnosis, whereinFNR 1,1 =0；
AND () is a logical function that returns 1 if AND only if both elements in brackets are 1, or 0 otherwise;
COUNTIF () is a logical function, two elements are included in brackets, the first element being a range, the second element being a scalar, the function returning to 1 if this scalar occurs in the range, otherwise returning to 0;
and then according to the formula
。
In an alternative embodiment, in step S2, N is calculated according to the following formula IMI ：
；
Where i=2, 3, …, total; j=1, 2, …, total-1,。
in an alternative embodiment, the construction method further comprises expanding the training samples, and then correcting the two-dimensional you index calculation formula using the following formula:
correcting two-dimensional you-den index =
；
The way to determine the optimized value of ω is to calculate the loss function:
。
MAX function is the extended full datasetAdjusted 2D Youden indexIs 1, and is incremented or decremented by one step value delta each time, such that the value of omega results from the last iteration 0 Becomes omega 1 The method comprises the steps of carrying out a first treatment on the surface of the If the calculation result of the loss function sigma is not less than 0, the omega value is maintained as the omega of the last iteration result 0 Unchanged; if the result of the calculation of the loss function sigma is smaller than 0, omega is calculated 1 Assign a value to ω 0 And repeating the judgment until the calculation result of the loss function sigma is not less than 0, and then taking the new omega value as the corrected model parameter.
Preferably 0.01<<0.05。
In a second aspect, the invention provides a TBR threshold combination of dual fluorescent tracers obtained using the construction method as described in any of the previous embodiments for assessing risk of lymph node metastasis.
In an alternative embodiment, the dual fluorescent tracer is indocyanine green and 5-aminolevulinic acid, the TBR threshold combination is that the TBR threshold of indocyanine green is 10-25, and the TBR threshold of 5-aminolevulinic acid is 5-15.
In a third aspect, the present invention provides an application of the TBR threshold combination in any one of (a) to (d) according to the foregoing embodiment:
(a) Developing a medicament for treating or preventing metastatic lymph nodes;
(b) Developing a contrast agent or combination thereof for tumor imaging, metastatic lymph node identification or tumor margin assessment;
(c) Training or checking metastatic lymph node extirpation surgical skills;
(d) And evaluating the efficacy of metastatic lymph nodes.
Compared with the prior art, the method for distinguishing the metastatic lymph nodes in real time and high efficiency in operation by using the two-dimensional eudon index and the double fluorescent tracer (especially indocyanine green and 5-aminolevulinic acid) in combination has the following advantages: 1. the method can judge the benign and malignant properties of the lymph nodes in real time in the operation, and fills the blank of evaluating the lymph nodes in the traditional operation; 2. the method of the invention can keep benign lymph nodes as much as possible while ensuring no false negative, and is expected to improve the prognosis of patients and reduce postoperative complications; 3. the method is simple in calculation, programmable and capable of continuously optimizing model prediction accuracy along with accumulation of clinical data. Therefore, the medicine for treating or preventing the metastatic lymph nodes can be developed more efficiently and accurately, the training or the assessment of the metastatic lymph node removal operation skill of an operator can be performed, or the medicine effect evaluation of the metastatic lymph nodes can be performed.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic illustration of an iterative equation for calculating FNR at a two-dimensional threshold combination;
FIG. 2 is a ROC plot of 5-ALA fluorescence TBR of example 2;
FIG. 3 is a ROC graph of ICG fluorescence TBR of example 2;
FIG. 4 is a heat map of the two-dimensional Euclidean index of the combination of 5-ALA and ICG in example 2;
FIG. 5 shows the result of model verification in example 2.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention.
In a specific embodiment, the present invention provides a method for constructing a lymph node metastasis risk assessment model, including: the lymph node metastasis risk assessment model adopts a double-fluorescence tracer to indicate the obtained two-dimensional you index to determine metastatic lymph nodes; the construction method comprises the following steps:
s1, acquiring training data: training (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section conventional pathology detection data of normal lymph nodes and metastatic lymph nodes in a sample;
s2, respectively arranging TBR thresholds obtained by calculating the spectral data of the dual fluorescent tracer obtained in the step S1 according to the order of magnitude as an abscissa axis and an ordinate axis, drawing a two-dimensional Euclidean index matrix chart, and taking a TBR threshold combination of the dual fluorescent tracer corresponding to the maximum value of the two-dimensional Euclidean index as an evaluation index of the metastatic lymph node;
the calculation formula of the TBR threshold value is as follows:
；
wherein TBR is the signal-to-back ratio of the dual fluorescent tracer per lymph node and INT () is a rounding function;
the two-dimensional you-den index calculation formula is:
；
wherein i=1, 2, …, total; j=1, 2, …, total;
TPR i,j true positive rate of lymph nodes judged under a certain two-dimensional threshold value is combined for the dual fluorescent tracer;
P G the number of metastatic lymph nodes judged for pathological diagnosis;
N IMI i,j the number of negative lymph nodes judged under a certain two-dimensional threshold value is used for the dual fluorescent tracer;
N G the number of negative lymph nodes judged by the pathological gold standard;
s3, acquiring verification data: verifying (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section routine pathology detection data of normal lymph nodes and metastatic lymph nodes in the sample;
and S4, adjusting the association model adopting the two-dimensional Euonyx index by using the verification data.
In an alternative embodiment, the dual fluorescent tracer comprises indocyanine green and 5-aminolevulinic acid.
In an alternative embodiment, the method for acquiring spectral data of the dual fluorescent tracer in step S1 includes injecting indocyanine green and 5-aminolevulinic acid into a training sample respectively, exposing lymph nodes in skin or fat of the training sample, acquiring fluorescence spectra of the lymph nodes (T) and surrounding normal tissues (B) by using an indocyanine green fluorescence spectrum probe and a 5-aminolevulinic acid fluorescence spectrum probe respectively, and further calculating respective signal-to-back ratios TBR of indocyanine green fluorescence and 5-aminolevulinic acid fluorescence.
Preferably, the training sample comprises a nude mouse or a human body.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid injected into the nude mice are 1-10 mg/kg and 100-500 mg/kg respectively.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid are injected into nude mice at 5.6 mg/kg and 250 mg/kg, respectively.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid administered to a human body are 0.2-2 mg/kg and 10-50 mg/kg, respectively.
Preferably, the doses of indocyanine green and 5-aminolevulinic acid are administered to the human at 0.62mg/kg and 27.8 mg/kg, respectively.
Preferably, indocyanine green and 5-aminolevulinic acid are administered to the training sample, and the lymph nodes in the skin or fat of the training sample are then exposed.
Preferably, the injection time of the 5-aminolevulinic acid is 10-30 hours later than the injection time of indocyanine green, and more preferably 20 hours later.
Preferably, the injection time of the 5-aminolevulinic acid is 2-6 hours before operation observation, more preferably 3-4 hours before operation;
preferably, the injection is by pushing blood through the peripheral vein using an intravenous pump.
Preferably, the mode of administration of 5-aminolevulinic acid for the human is oral.
In an alternative embodiment, the paraffin section is prepared by embedding and fixing the formalin-fixed isolated lymph node using paraffin, then slicing, and staining the slice.
Preferably, the concentration of formalin is 10%.
Preferably, the staining comprises HE staining.
In an alternative embodiment, the conventional pathology detection data comprises diagnostic results obtained by a pathologist based on paraffin section observations.
Preferably, the pathologist uses an inverted microscope to view paraffin sections.
In an alternative embodiment, in step S2, TPR i,j The calculation method of (1) is as follows:
the false negative rate FNR of the lymph node determined by the dual fluorescent tracer combination was first calculated according to the following formula:
；
wherein i and j represent matrix elements in the ith row and the jth column, total is Total number of lymph nodes, and P G Number of metastatic lymph nodes, noX, judged by pathological diagnosis i And NoY j Is the number of the ith and jth lymph nodes in the x-dimension and y-dimension, noM 1 、NoM 2 ……Is the number of metastatic lymph node judged by pathological diagnosis, whereinFNR 1,1 =0；
AND () is a logical function that returns 1 if AND only if both elements in brackets are 1, or 0 otherwise;
COUNTIF () is a logical function, two elements are included in brackets, the first element being a range, the second element being a scalar, the function returning to 1 if this scalar occurs in the range, otherwise returning to 0;
and then according to the formula
；
Calculating TPR i,j 。
In an alternative embodiment, in step S2, N is calculated according to the following formula IMI ：
；
Where i=2, 3, …, total; j=1, 2, …, total-1,。
in an alternative embodiment, the construction method further comprises expanding the training samples, and then correcting the two-dimensional you index calculation formula using the following formula:
correcting two-dimensional you-den index =
；
The way to determine the optimized value of ω is to calculate the loss function:
；
MAX function is the extended full datasetAdjusted 2D Youden indexIs 1, and is incremented or decremented by one step value delta each time, such that the value of omega results from the last iteration 0 Becomes omega 1 The method comprises the steps of carrying out a first treatment on the surface of the If the calculation result of the loss function sigma is not less than 0, the omega value is maintained as the omega of the last iteration result 0 Unchanged; if the result of the calculation of the loss function sigma is smaller than 0, omega is calculated 1 Assign a value to ω 0 And repeating the judgment until the calculation result of the loss function sigma is not less than 0, and then taking the new omega value as the corrected model parameter.
Preferably, 0.01< delta <0.05.
In a second aspect, the invention provides a TBR threshold combination of dual fluorescent tracers obtained using the construction method as described in any of the previous embodiments for assessing risk of lymph node metastasis.
In an alternative embodiment, the dual fluorescent tracer is indocyanine green and 5-aminolevulinic acid, the TBR threshold combination is that the TBR threshold of indocyanine green is 10-25, and the TBR threshold of 5-aminolevulinic acid is 5-15.
In a third aspect, the present invention provides an application of the TBR threshold combination in any one of (a) to (d) according to the foregoing embodiment:
(a) Developing a medicament for treating or preventing metastatic lymph nodes;
(b) Developing a contrast agent or combination thereof for tumor imaging, metastatic lymph node identification or tumor margin assessment;
(c) Training or checking metastatic lymph node extirpation surgical skills;
(d) And evaluating the efficacy of metastatic lymph nodes.
Some embodiments of the present invention are described in detail below with reference to the accompanying drawings. The following embodiments and features of the embodiments may be combined with each other without conflict.
Example 1
The embodiment provides a method for distinguishing metastatic lymph nodes in real time and efficiently in operation by using a two-dimensional eudon index and indocyanine green and 5-aminolevulinic acid in combination, which comprises the following steps: A. surgical lymph node data acquisition; B. a conventional combined use of a you-den index analysis list and a two-dimensional you-den index analysis; C. and the clinical big data is collected, the model data is expanded, and the accuracy is improved.
Step a is surgical lymph node data acquisition, comprising the steps of:
(A1) Preoperative injections of 5-ALA and ICG. In order for the fluorescent tracer to produce sufficient signal differences in the metastatic lymph nodes and normal lymph nodes, it takes appropriate time for the tracer to accumulate and metabolize. 5.6 The ICG of mg/kg is dissolved in 100 mu L of deionized water for nude mice injection, the injection time is 24 h before operation, and the injection is carried out through tail vein once; or 0.62mg/kg ICG is dissolved in 30 mL of sterilized water for injection into human body, the injection time is 24 h before operation, and blood is pushed into the body at constant speed within 30 min by using an intravenous pump through peripheral vein. 250 mg/kg of 5-ALA is dissolved in 100 mu L of deionized water for nude mice injection, the injection time is 4h before operation, and the injection is carried out through tail vein for one time; or 27.8 mg/kg of 5-ALA is dissolved in pure drinking water for oral administration to human body, and the administration time is 4h before operation. 5-ALA requires that the subject be protected from direct sunlight within days after injection in order to avoid photo-induced drug allergy.
(A2) Intraoperatively detecting 5-ALA and ICG spectra of lymph nodes;
this example employs another method comprising steps (a 2.1) to (a 2.3):
(A2.1) free exposure of lymph nodes at the target site from skin and fat (in nude mice, only skin needs to be cut and free under anesthesia);
(A2.2) the ICG fluorescence spectrometer probe and the 5-ALA fluorescence spectrometer probe were brought into close proximity to the target lymph node in sequence and their spectra measured, all set parameters of the spectrometer remained unchanged throughout the experiment except for the integration time, which was typically kept constant (e.g., 100 ms), and if fluorescence intensity was too high, the spectrometer integration time was scaled down (e.g., 50ms,25ms or 20 ms) and then the signal measurements were multiplied by corresponding multiples (e.g., 2, 4 or 5) at the later data processing. And repeating the measurement 5-8 times at different positions of the same target each time, and taking an average value of the result.
(A2.3) after the lymph node measurement of each individual, the fluorescence signal of the surrounding normal tissue (e.g., muscle, fat) was measured. And repeating the measurement 5-8 times at different positions of the same target each time, and taking an average value of the result.
(A3) Conventional pathological detection of postoperative lymph node paraffin sections;
this example employs another method comprising steps (a 3.1) to (a 3.5):
(A3.1) for lymph nodes with measured spectra, excised with a scalpel, the isolated lymph nodes were fixed with 10% formalin overnight;
(A3.2) taking out the fixed isolated lymph nodes, and embedding the lymph nodes by paraffin;
(A3.3) cutting out 3 paraffin sections with a thickness of 5 μm at the anterior, middle and posterior parts of the lymph node with a microtome;
(a 3.4) obtaining 3 HE staining sections of each lymph node sample through the HE staining standard step of the conventional pathological section;
(a 3.5) allowing an experienced pathologist to observe HE stained sections with an inverted microscope and let it judge whether the lymph node is a metastatic lymph node (i.e., is invaded by cancer cells) without informing it of the fluorescence results.
Step B is a combination of conventional and two-dimensional Ewing index analysis, comprising the steps of:
(B1) Optimal TBR thresholds for 5-ALA, ICG single use are obtained using conventional eudragit indices, respectively, comprising steps (B1.1) to (B1.4):
(B1.1) numbering all lymph nodes, and then sorting the lymph nodes according to the peak value of the 5-ALA fluorescence spectrum and the peak value of the ICG fluorescence spectrum from small to large respectively;
(B1.2) defining a TBR threshold:
；
wherein TBR is the signal-to-back ratio of 5-ALA fluorescence or ICG fluorescence for each lymph node,
a TBR threshold value of 5-ALA fluorescence or ICG fluorescence for each lymph node above which the lymph node is judged positive by the tracer alone;INT() Is a rounding function;
(B1.3) each lymph node can be correspondingly provided with a TBR threshold value of 5-ALA fluorescence or ICG fluorescence, and the true positive rate (sensibility), the true negative rate (specificity) and the conventional Euclidean index (sensibility+specificity-1) under each threshold value are respectively calculated by combining the pathological detection result in the step C;
and (B1.4) drawing a conventional Eudragit index-TBR threshold relation image of the two tracer single use methods, and finding a TBR threshold corresponding to the maximum value of the conventional Eudragit index, namely the optimal TBR threshold of the 5-ALA and ICG single use method.
(B2) Obtaining an optimal TBR threshold for a combination of 5-ALA and ICG using a two-dimensional eudon index, the steps comprising (B2.1) to (B2.8):
(B2.1) numbering all lymph nodes, and then sorting the lymph nodes according to the peak value of the 5-ALA fluorescence spectrum and the peak value of the ICG fluorescence spectrum from small to large respectively;
(B2.2) calculating TBR threshold:
formula (1):；
(B2.3) taking the ordered 5-ALA fluorescence TBR threshold as the x dimension (column direction), taking the ordered ICG fluorescence TBR threshold as the y dimension (row direction), x and y forming two dimensions of a two-dimensional matrix;
(B2.4) calculating a False Negative Rate (FNR) at each two-dimensional threshold combination;
another method is adopted in the embodiment, which comprises the steps (B2.4.1) to (B2.4.2);
(B2.4.1) Here, the definition of the negative of the combination is established if and only if the determination results of both tracers are negative, and the combination is determined to be positive in the other cases. So the first column and first row in the two-dimensional matrix can be set to 0 according to the definition of the TBR threshold;
(B2.4.2) calculating the FNR for all two-dimensional threshold combinations using an iterative equation:
formula (2):
；
as shown in fig. 1. Wherein, the liquid crystal display device comprises a liquid crystal display device,、/>representing matrix elements at->Line, th->Column (/ -)>=2，3，…，Total；/>=1, 2, …, total-1), total is Total number of lymph nodes, P G Metastatic lymph node judged by pathological diagnosis gold standardNumber, noX i And NoY j Is the number of the ith and jth lymph nodes in the x-and y-dimensions,NoM 1 、NoM 2 ……/>is the number of the metastatic lymph node of the pathological diagnosis decision, AND () is a logical function, returning 1 if AND only if both elements in the brackets are 1, AND returning 0 otherwise. COUNTIF () is a logical function comprising two elements in brackets, the first element being a range (e.gNoM 1 ：/>And (3) representing. From the slaveNoM 1 To->The set of all numbers), the second element is a scalar, the function returns a1 if this scalar occurs in the range, and a 0 otherwise.
(B2.5) calculating the True Positive Rate (TPR) under each two-dimensional threshold combination, wherein the rule that the sum of the TPR and the FNR is 1 can be directly calculated by the FNR in the step E4 to obtain the TPR under all two-dimensional threshold combinations:
formula (3):；
wherein the method comprises the steps of=2，3，…，Total；/>=1，2，…，Total-1；
(B2.6) calculating the number of negative lymph nodes judged by the tracer combination method under each two-dimensional threshold combinationN IMI ）；
In this embodiment, another method is used, which includes steps (B2.6.1) to (B2.6.2)
(B2.6.1) the first column and first row in the two-dimensional matrix can be set to 0 according to the definition of the TBR threshold;
(B2.6.2) calculating all two-dimensional threshold combinations using iterative equationsN IMI ：
Formula (4):
；
wherein the method comprises the steps of=2，3，…，Total；/>=1，2，…，Total-1；
(B2.7) calculating a two-dimensional you index (2D you index) for each two-dimensional threshold combination, combining equations (1), (2) and (3), and calculating 2D you index for all two-dimensional threshold combinations using equation (4):
；
wherein the method comprises the steps of=1，2，…，Total；/>=1，2，…，Total；
(B2.8) finding out a two-dimensional TBR threshold value combination corresponding to the maximum value in a matrix list of the 2D Youden index, namely, an optimal TBR threshold value (two-dimensional combination form) of a 5-ALA and ICG combined method;
and C, collecting clinical big data, expanding model data, and improving accuracy.
After each model data expansion, the two components of the 2D Youden index, sensitivity and specificity, are given weighting coefficients, respectively, namely:
;
determination ofThe way of optimizing the values of (a) is to calculate the loss function:
；
MAX function is the extended full datasetAdjusted 2D Youden indexThe initial value of ω is 1, and is increased or decreased by one step value δ at a time, (preferably, 0.01)<δ<0.05 So that the ω value results from the last iteration ω 0 Becomes omega 1 The method comprises the steps of carrying out a first treatment on the surface of the If the calculation result of the loss function sigma is not less than 0, the omega value is maintained as the omega of the last iteration result 0 Unchanged; if the result of the calculation of the loss function sigma is smaller than 0, omega is calculated 1 Assign a value to ω 0 And repeating the judgment until the calculation result of the loss function sigma is not less than 0, then taking the new omega value as the corrected model parameter until the next expansion of the model data, and executing the step C again.
Although the method is not an essential step for realizing the method in a single experiment, the method has important significance for continuously improving the accuracy of judging the metastatic lymph nodes, and belongs to a step for further expanding the clinical application of the method. When the method is used for clinical practice, the step B and the step C are programmed for use, and the newly added clinical data are directly input into the program each time, so that the sample content of the method model can be continuously expanded, and the prediction accuracy of the model can be continuously improved. When model data are accumulated to a certain large data scale and judgment accuracy is fully verified, the method is expected to replace the mode of regional lymph node cleaning and postoperative pathological diagnosis adopted in clinic, and objective basis is provided for doctors to resect metastatic lymph nodes and reserve normal lymph nodes directly in operation.
Example 2
The procedure provided in example 1 was followed for model construction and validation using nude mouse samples as follows:
2.1 Model construction
The 25 nude mice metastatic lymph node model was selected, and since the metastatic lymph node could only be pathologically verified after removal, it was not known whether the target lymph node was the metastatic lymph node at the time of fluorescence judgment, and the calculation was performed according to the procedure provided in example 1, and fluorescence and pathologic diagnosis data were as follows:
(1) Nude mouse fluorescence and pathology diagnostic data
ROC curves of the 5-ALA fluorescence TBR and the ICG fluorescence TBR are shown in FIG. 2 and FIG. 3 respectively, and TBR thresholds of the 5-ALA and the ICG are calculated as follows: 12.69 and 13.84. The conventional you's index maximum for both fluorescent tracers alone was 0.597 and 0.565, respectively.
(2) 2D Youden index data
A two-dimensional Euclidean index heat map of the combination of 5-ALA and ICG is shown in FIG. 4, and the optimal TBR threshold is obtained (12.69, 18.95).
2.2 model verification
Another batch of nude mice was taken, the TBR values of 5-ALA and ICG from each lymph node were directly detected, metastatic assessment was performed using the optimal TBR threshold obtained in step 2.1, and then compared with the pathology results, with a final sensitivity of 88%, as shown in fig. 5.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention.

Claims (15)

1. The method for constructing the metastatic lymph node differentiation model is characterized by comprising the following steps: the metastatic lymph node differentiation model adopts a double-fluorescence tracer to indicate the obtained two-dimensional you index to determine the metastatic lymph node; the construction method comprises the following steps:
s1, acquiring training data: training (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section conventional pathology detection data of normal lymph nodes and metastatic lymph nodes in a sample;
s2, respectively arranging TBR thresholds obtained by calculating the spectral data of the dual fluorescent tracer obtained in the step S1 according to the order of magnitude as an abscissa axis and an ordinate axis, drawing a two-dimensional Euclidean index matrix chart, and taking a TBR threshold combination of the dual fluorescent tracer corresponding to the maximum value of the two-dimensional Euclidean index as an evaluation index of the metastatic lymph node;
the calculation formula of the TBR threshold value is as follows:
；
where TBR is the signal-to-back ratio of the dual fluorescent tracer for each lymph node and INT () is a rounding function;
the two-dimensional you-den index calculation formula is:
；
wherein i and j represent matrix elements in the ith row and the jth column, total is the Total number of lymph nodes, i=1, 2, … and Total; j=1, 2, …, total;
TPR i,j is a double fluorescent tracerCombining true positive rates of lymph nodes determined below a certain two-dimensional threshold;
P G the number of metastatic lymph nodes judged for pathological diagnosis;
N IMI i,j the number of negative lymph nodes judged under a certain two-dimensional threshold value is used for the dual fluorescent tracer;
N G the number of negative lymph nodes judged by the pathological gold standard;
s3, acquiring verification data: verifying (a) dual fluorescent tracer spectral data of normal lymph nodes and metastatic lymph nodes and (b) paraffin section routine pathology detection data of normal lymph nodes and metastatic lymph nodes in a training sample;
and S4, adjusting the association model adopting the two-dimensional Euonyx index by using the verification data.
2. The method of claim 1, wherein the dual fluorescent tracer comprises indocyanine green and 5-aminolevulinic acid.
3. The construction method according to claim 2, wherein the method for obtaining spectral data of the dual fluorescent tracer in step S1 comprises injecting indocyanine green and 5-aminolevulinic acid into the training sample, exposing lymph nodes in skin or fat of the training sample, collecting fluorescence spectra of the lymph nodes and surrounding normal tissues by using indocyanine green fluorescence spectrum probe and 5-aminolevulinic acid fluorescence spectrum probe, respectively, and further calculating respective signal-to-back ratio TBR of indocyanine green fluorescence and 5-aminolevulinic acid fluorescence.
4. A method of constructing as claimed in claim 3 wherein said training sample comprises nude mice or humans.
5. The method according to claim 4, wherein the doses of indocyanine green and 5-aminolevulinic acid are 5.6 mg/kg and 250 mg/kg, respectively, in nude mice.
6. The method of claim 4, wherein indocyanine green and 5-aminolevulinic acid are administered to the human in doses of 0.62mg/kg and 27.8 mg/kg, respectively.
7. A method of construction according to claim 3, wherein indocyanine green and 5-aminolevulinic acid are administered to the training sample and the lymph nodes under the skin or in fat of the training sample are then exposed.
8. A method of construction according to claim 3, wherein the 5-aminolevulinic acid injection time is 20 hours later than indocyanine green injection time.
9. The construction method according to claim 3, wherein the injection time of 5-aminolevulinic acid is 3-4 hours before surgical observation.
10. The method of claim 1, wherein the paraffin section is prepared by embedding and fixing the formalin-fixed isolated lymph node using paraffin, then slicing, and staining the slice.
11. The method of claim 1, wherein the conventional pathology detection data comprises diagnostic results obtained from a pathologist viewing paraffin sections.
12. The construction method according to claim 1, wherein in step S2, TPR i,j The calculation method of (1) is as follows:
the false negative rate FNR of the lymph node determined by the dual fluorescent tracer combination was first calculated according to the following formula:
；
P G number of metastatic lymph nodes, noX, judged by pathological diagnosis i And NoY j Is the number of the ith and jth lymph nodes in the x-dimension and y-dimension, noM 1 、NoM 2 、⋯、Is the number of metastatic lymph node judged by pathological diagnosis, whereinFNR 1,1 =0；
AND () is a logical function that returns 1 if AND only if both elements in brackets are 1, or 0 otherwise;
COUNTIF () is a logical function, two elements are included in brackets, the first element being a range, the second element being a scalar, the function returning to 1 if this scalar occurs in the range, otherwise returning to 0;
and then according to the formula
；
Calculating TPR i,j 。
13. The construction method according to claim 1, wherein in step S2, N is calculated according to the following formula IMI ：
；
Where i=2, 3, …, total; j=1, 2, …, total-1,。
14. the method of claim 1, further comprising expanding the training samples and then correcting the two-dimensional euler-index calculation formula using the following formula:
correcting two-dimensional you-den index =
；
The way to determine the optimized value of ω is to calculate the loss function:
；
MAX function is the extended full datasetAdjusted 2D Youden indexIs 1, and is incremented or decremented by one step value delta each time, such that the value of omega results from the last iteration 0 Becomes omega 1 The method comprises the steps of carrying out a first treatment on the surface of the If the calculation result of the loss function sigma is not less than 0, the omega value is maintained as the omega of the last iteration result 0 Unchanged; if the result of the calculation of the loss function sigma is smaller than 0, omega is calculated 1 Assign a value to ω 0 And repeating the judgment until the calculation result of the loss function sigma is not less than 0, and then taking the new omega value as the corrected model parameter.
15. The method of claim 14, wherein 0.01< δ <0.05.
CN202310511551.3A 2023-05-09 2023-05-09 Metastatic lymph node distinguishing model, construction method and application Active CN116259414B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310511551.3A CN116259414B (en) 2023-05-09 2023-05-09 Metastatic lymph node distinguishing model, construction method and application

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310511551.3A CN116259414B (en) 2023-05-09 2023-05-09 Metastatic lymph node distinguishing model, construction method and application

Publications (2)

Publication Number Publication Date
CN116259414A CN116259414A (en) 2023-06-13
CN116259414B true CN116259414B (en) 2023-08-18

Family

ID=86679617

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310511551.3A Active CN116259414B (en) 2023-05-09 2023-05-09 Metastatic lymph node distinguishing model, construction method and application

Country Status (1)

Country Link
CN (1) CN116259414B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108882896A (en) * 2015-09-23 2018-11-23 诺瓦达克技术公司 For evaluating the method and system of the healing of tissue
WO2019245009A1 (en) * 2018-06-22 2019-12-26 株式会社Ａｉメディカルサービス Method of assisting disease diagnosis based on endoscope image of digestive organ, diagnosis assistance system, diagnosis assistance program, and computer-readable recording medium having said diagnosis assistance program stored thereon
CN111602174A (en) * 2018-01-08 2020-08-28 普罗热尼奇制药公司 System and method for rapidly segmenting images and determining radiopharmaceutical uptake based on neural network
CN112837815A (en) * 2021-02-04 2021-05-25 复旦大学附属中山医院 Model for evaluating gastric cancer prognosis based on Lauren typing and postoperative residual lymph node and application
CN114588268A (en) * 2022-03-25 2022-06-07 清华大学 Methods for increasing activation of STING pathway and anti-tumor T cell response
CN115222992A (en) * 2022-08-05 2022-10-21 中国科学院自动化研究所 Preoperative intraoperative image-based breast cancer lymph node benign and malignant classification system and method

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200342600A1 (en) * 2018-01-08 2020-10-29 Progenics Pharmaceuticals, Inc. Systems and methods for rapid neural network-based image segmentation and radiopharmaceutical uptake determination
US20200005901A1 (en) * 2018-06-30 2020-01-02 20/20 Genesystems, Inc Cancer classifier models, machine learning systems and methods of use

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108882896A (en) * 2015-09-23 2018-11-23 诺瓦达克技术公司 For evaluating the method and system of the healing of tissue
CN111602174A (en) * 2018-01-08 2020-08-28 普罗热尼奇制药公司 System and method for rapidly segmenting images and determining radiopharmaceutical uptake based on neural network
WO2019245009A1 (en) * 2018-06-22 2019-12-26 株式会社Ａｉメディカルサービス Method of assisting disease diagnosis based on endoscope image of digestive organ, diagnosis assistance system, diagnosis assistance program, and computer-readable recording medium having said diagnosis assistance program stored thereon
CN112837815A (en) * 2021-02-04 2021-05-25 复旦大学附属中山医院 Model for evaluating gastric cancer prognosis based on Lauren typing and postoperative residual lymph node and application
CN114588268A (en) * 2022-03-25 2022-06-07 清华大学 Methods for increasing activation of STING pathway and anti-tumor T cell response
CN115222992A (en) * 2022-08-05 2022-10-21 中国科学院自动化研究所 Preoperative intraoperative image-based breast cancer lymph node benign and malignant classification system and method

Also Published As

Publication number Publication date
CN116259414A (en) 2023-06-13

Similar Documents

Publication Publication Date Title
Zhang et al. Liver tumor boundaries identified intraoperatively using real-time indocyanine green fluorescence imaging
Shioyama et al. Intraoperative flow cytometry analysis of glioma tissue for rapid determination of tumor presence and its histopathological grade
US10188297B2 (en) Intra-operative molecular imaging
US20060173362A1 (en) Methods of medical imaging using quantum dots
Kawashima et al. Transpapillary biliary forceps biopsy to distinguish benign biliary stricture from malignancy: how many tissue samples should be obtained?
WO2008144760A1 (en) Optical methods to intraoperatively detect positive prostate and kidney cancer margins
Abramov et al. Intraoperative confocal laser endomicroscopy: prospective in vivo feasibility study of a clinical-grade system for brain tumors
Cao et al. A pilot study of near-infrared fluorescence guided surgery for primary tumor localization and lymph node mapping in colorectal cancer
CN116259414B (en) Metastatic lymph node distinguishing model, construction method and application
RU2350262C2 (en) Method of differential diagnostics of eyelid skin neoplasms
RU2743221C1 (en) Method for determining type of gastric cancer by lauren classification by computed tomography
Lim et al. An optimized 3T MRI scan protocol to assess iris melanoma with subsequent histopathological verification–a prospective study
Fu et al. Improving oral squamous cell carcinoma diagnosis and treatment with fluorescence molecular imaging
RU2522388C1 (en) Method for determining therapeutic approach to differentiated thyroid carcinoma combined with hashimoto&#39;s thyroiditis with junction
CN114601938B (en) Fluorescence/nuclear magnetic probe for guiding breast cancer breast conservation cutting edge accurate assessment
CN104805079B (en) The VX2 tumor markers probe of based on PCR technology
Khokhar et al. Discussion of Uveal Melanoma Metastasis-Using Novel Biomarkers for Improved Tracking & Treatment
Fernandes et al. Applicability of autofluorescence and fluorescent probes in the trans-surgical of oral carcinomas: A systematic review
Yousri Targeted axillary dissection using carbon-tattooing for clinically node-positive breast cancer converting to node negativity after neoadjuvant systemic therapy: Initial experience
Song et al. Tu1628 Initial Experiences of EUS-Guided Radiofrequency Ablation of Unresectable Pancreatic Cancer
Jin et al. Near-Infrared Intraoperative Imaging of Pelvic Autonomic Nerves: Clinical Trial
RU2157528C1 (en) Method for predicting malignant neoplasm to be the case
Nijboer et al. Fluorescence‐guided surgery using cetuximab‐800CW in patients with penile carcinoma
Long et al. Predicting Tertiary Lymphoid Structures in Head and Neck Contrast-Enhanced CT Images using a 3D CNN
El-Sharkawy Automated hyperspectral imaging for non-invasive characterization of human eye vasculature: A potential tool for ocular vascular evaluation

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