CN111739615A - AI medical diagnosis image picture computer input method - Google Patents
AI medical diagnosis image picture computer input method Download PDFInfo
- Publication number
- CN111739615A CN111739615A CN202010631295.8A CN202010631295A CN111739615A CN 111739615 A CN111739615 A CN 111739615A CN 202010631295 A CN202010631295 A CN 202010631295A CN 111739615 A CN111739615 A CN 111739615A
- Authority
- CN
- China
- Prior art keywords
- computer
- logic
- model
- picture
- medical
- 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.)
- Pending
Links
Classifications
-
- 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
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- 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
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Abstract
The invention provides a computer input method of AI medical image pictures. The method comprises the steps of establishing codes for basic pathological changes and symptoms of each system of a human body and corresponding image expressions one by one, inputting coded pictures into a computer, and establishing a reliable internal logic model in the computer through a drawing tool, so that the technical problem of inputting medical image picture codes into the computer is solved. The invention forms the picture element by the medical image through the picture coding, and then inputs the picture element to the computer, establishes the logic model for the computer to identify, realizes the computer identification and interpretation of the medical image, thereby reducing the time cost, having the advantages of clear picture coding information expression, easy storage and interpretation, wide application range and accurate identification result. The method is suitable for being used as a medical diagnosis image picture computer input method.
Description
Technical Field
The invention provides a computer input method suitable for AI medical image pictures, which is used in the field of image processing.
Background
Modern medicine develops to date, an examination instrument is advanced and rapid, but a diagnosis process which needs medical images as diagnosis and reference is related, the image performance still needs to be judged by doctors to read and diagnose, but the medical images of a plurality of diseases are different, and aiming at the characterization result of the medical image of the same disease, doctors need to repeatedly read the medical images of different patients, so that time and labor are wasted, and a large amount of time is occupied for doctors and patients.
The existing image diagnosis report is divided into 4 parts: 1) general items including basic information such as name, sex, age, etc. of the patient;
2) a description part which narrates the observed images by using the professional terms, namely basic lesions, signs and associated words of each system;
3) image diagnosis, namely, a conclusion obtained by combining image data with clinically relevant data;
4) signatures, i.e. signatures of the consultant and the reviewing physician.
The disadvantages in the image diagnosis process are:
after the medical image is obtained, a doctor is usually required to perform medical image interpretation, the workload of the doctor is large, and generally, medical image results cannot be obtained on the same day, so that the time of patients is occupied.
In view of the above disadvantages, it is urgently needed to provide a method capable of rapidly performing medical image recognition.
Disclosure of Invention
The invention provides a computer input and identification method suitable for AI medical image pictures, aiming at the defect that the current medical image can not be suitable for AI identification and classified input into a computer. The method comprises the steps of establishing codes for basic pathological changes and symptoms of each system of a human body and corresponding image expressions one by one, inputting coded pictures into a computer, and establishing a reliable internal logic model in the computer through a drawing tool, so that the technical problem of inputting medical image picture codes into the computer is solved.
The technical scheme adopted by the invention for solving the technical problems is as follows:
drawing a picture code on the medical image through drawing software, establishing a computer identification model, forming a logic model, and performing supplementary information logic and application verification on the logic model.
Establishing a computer recognition model: establishing a unique picture code by marking the existing medical image and the interpretation result, establishing basic pathological changes and symptoms of each organ of a human body with the corresponding image data expression image through a drawing tool, forming a picture code file according to the sequence of each organ system by the picture code, and inputting the picture code file to a computer to complete the establishment of a computer identification model;
logic model: by combining with computer picture coding, the established model of the computer is edited in a logic sequence, and conventional interpretation terms for medical image recognition are supplemented to form a logic model;
and supplementary information logic: the verified logic model is combined with other patient data recorded by a computer, and the basic identity information, the past medical history, the existing symptoms or other biological verification results of the patient are logically supplemented to further perfect the logic interpretation of the logic model;
application verification: comparing, identifying and verifying the edited logic model with the manual interpretation result of the existing medical image database, verifying the reliability of the logic model, and completing verification of the logic model; and applying the verified model, comparing and verifying the model with the medical image manual identification interpretation of the actual clinical image department, further verifying whether the established logic model is reliable and accurate, and if the established logic model is not deviated from the manual identification, proving that the established logic model is accurate.
The positive effects are as follows: the invention forms the picture element by the medical image through the picture coding, and then inputs the picture element to the computer, establishes the logic model for the computer to identify, realizes the computer identification and interpretation of the medical image, thereby reducing the time cost, having the advantages of clear picture coding information expression, easy storage and interpretation, wide application range and accurate identification result. The method is suitable for being applied as a computer input identification method of medical diagnosis images.
Detailed Description
The following will clearly and completely describe the technical solutions in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Establishing a computer recognition model: establishing a unique picture code by marking the existing medical image and the interpretation result, establishing basic pathological changes and symptoms of each organ of a human body with the corresponding image data expression image through a drawing tool, forming a picture code file according to the sequence of each organ system by the picture code, and inputting the picture code file to a computer to complete the establishment of a computer identification model;
establishing a logic model: by combining with computer picture coding, the established model of the computer is edited in a logic sequence, and conventional interpretation terms for medical image recognition are supplemented to form a logic model;
and supplementary information logic: the verified logic model is combined with other patient data recorded by a computer, and the basic identity information, the past medical history, the existing symptoms or other biological verification results of the patient are logically supplemented to further perfect the logic interpretation of the logic model;
application verification: comparing, identifying and verifying the edited logic model with the manual interpretation result of the existing medical image database, verifying the reliability of the logic model, and completing verification of the logic model; and applying the verified model, comparing and verifying the model with the medical image manual identification interpretation of the actual clinical image department, further verifying whether the established logic model is reliable and accurate, and if the established logic model is not deviated from the manual identification, proving that the established logic model is accurate.
The medical images include lung, liver, brain, chest, head, biliary X-ray, CT, ultrasound scan, and nuclear magnetic resonance images.
Example 1: taking the establishment of a lung CT medical image model as an example, inputting the existing medical image and medical image interpretation, drawing basic image codes by adopting 3Dslicer software for example as a processing object, grouping and coding image representations according to symptoms, establishing basic logic rules by matching colors, recording and storing;
if a computer logic model is established: Lung-LU, inputs sub-categories under logic, such as: LU-1 exudation, LU-2 proliferation, LU-3 fibrosis, LU-5 calcification, LU-6 cavity, LU-7 cavity, LU-8 nodule, LU-9 lump, LU-10 emphysema, LU-11 atelectasis, LU-12 lobular sign, LU-13 pleural depression, LU-15 peripheral burr, LU-16 hydrops, etc.;
a branch subclass is established under each subclass, such as: left lung calcification LU-5-L, right lung calcification LU-5-R, diffuse emphysema LU-10-1, and tubiform pneumothorax LU-10-2;
supplementary entry color logic: GREEN-mild, YELLOW-moderate, RED-severe;
supplementary entry color logic: and (3) encoding the abnormal part image, such as: BLUE-image anomaly;
inputting and storing the logic sequence into a computer, judging and analyzing according to the logic sequence after the computer acquires medical influence, and outputting and identifying logic = LU-slight tuberculosis when the computer identifies that the image code simultaneously has LU-1 & LU-2 & LU-3 & GREEN; if the computer identification image codes are LU-1 & LU-2 & LU-3 & BLUE, outputting fault identification logic = LU-tuberculosis-fault, and at the moment, switching to the existing image identification system to perform doctor manual judgment and performing logic model supplementary interpretation verification;
supplementary input information logic: combining the verified logic model with other patient data recorded by a computer, such as basic information, past medical history, existing symptoms or other biological detection results of patients, performing logic supplement on the logic model, and further perfecting logic interpretation of the logic model;
final output = patient baseline + LU-mild tuberculosis + other bioassay + other medical information.
Because the image representation of the disease is complex, one image symptom corresponds to less of a disease condition. In most cases, different pathological anatomy, pathology and physiology characterization changes occur in different stages from occurrence to regression of the same disease, and individual differences in size, form, position and the like exist in the same stage, so that multiple images of the same disease occur simultaneously, which is called 'same disease and different shadow'; similarly, the phenomenon of "same disease in different appearances" can occur, i.e., different image manifestations mark the same disease. When a logic rule is established, the image segments and the image interpretation of the existing known database are extracted, so that the computer acquires the image segments of one or n nodes from different stages of disease occurrence to regression, and the accuracy of establishing the image coding logic is further ensured.
Each system is divided into:
1. basic lesion image representation of craniocerebral disease (CT):
gas: trauma, postoperative, artificial qi and brain, CT value-800 to-1000 HU;
fat: CT value-10 to-120 HU, normal fat, lipoma, fat accumulation, liposarcoma, wherein the liposarcoma has uneven density, and the sarcoma part has higher density than normal fat;
normal cerebrospinal fluid: CT value is 0-15 HU;
equal density: normal brain tissue density and lesions that are consistent or similar to normal brain tissue density. a-normal white matter HU, b-normal gray matter HU;
cerebral edema: the density is lower than the white matter of the brain and higher than the cerebrospinal fluid. The water content in brain tissue increases, causing an increase in brain volume.
Classification and mechanism: a-angiogenetic edema, BBB damage, part, diffusion around focus along white matter fiber, b-cytotoxic edema, Na-K pump inactivation calcium inflow, part, intracellular swelling, common gray matter, c-interstitial edema, hydrocephalus caused, part, white matter around ventricle.
Brain atrophy: brain tissue loss from a variety of causes and subsequent enlargement of the cerebral or subarachnoid space. Classification and etiology, a, idiopathic encephalatrophy, olivopontocerebellar atrophy, b, diffuse encephalatrophy, senile arteriosclerotic encephalopathy AD (alzheimer's disease), cerebral hypoxia, tumors and decompensation, c, localized encephalatrophy, trauma, infection, infarction.
Cerebral hematoma and hemorrhage: a-acute density, b-subacute hematoma density, c-chronic hematoma density, d-softened focus density, disease of e-linear bleeding, subarachnoid hemorrhage and suspected subarachnoid hemorrhage; high density and soft tissue density, higher than normal grey matter density.
Calcification and bone density, classification and etiology: a-physiological calcification, pineal bodies, choroid plexus, and brain sickle; b-pathological calcification, classification and etiology, 1) parasites, cysticercus, toxoplasma, hydatid, schistosomes, 2) vascular, arteriosclerosis AVM, cavernous hemangioma, Sturge-Weber syndrome, 3) inflammatory, tuberculosis, abscesses, encephalitis, 4) tumorous, oligodendroglioma, craniopharyngioma, pinealoid tumors, metastases. B-bilateral basal ganglia calcification, etiology and classification: a-congenital, tuberous sclerosis, neurofibromatosis, b-secretory, hyperparathyroidism, hypothyroidism, c-compensatory, Fahr's disease, MELAS syndrome, d-infective, tuberculosis, toxoplasma, viral infections (rubella, cytomegalovirus, etc.).
Mixed density, the focus of the mixture of two or more of the above densities.
Fuzzy focus edge or clear focus edge.
Normal structural shift: a-congenital malformation, B-white matter extrusion.
Cerebral hernia: the elevated intracranial pressure is unbalanced, causing a portion of the brain tissue to be displaced into another portion or through the damaged skull to extracranially. And (4) classification: a-hiatus hernia, b-occipital macroporous hernia, c-sickle cerebri-cranial hernia, d-extracranial hernia; space occupying effects or tumor effects.
Hydrocephalus: morphological changes in the ventricular brain pool: 1 or more of the 3 links of secretion, circulation and absorption of cerebrospinal fluid are obstructed and can cause; and (4) classification: a-communicating hydrocephalus, the blockage or malabsorption of the normal CSF pathways below the four ventricular outlets, causes: subarachnoid hemorrhage, meningitis, craniocerebral injury, venous sinus embolism and b-normbaric hydrocephalus occur on the basis of traffic hydrocephalus, the CSF circulation function is in a compensation range, the hydrocephalus is mostly seen in the elderly, c-obstructive hydrocephalus occurs, obstruction occurs at any part above the outlet of the fourth ventricle, and the hydrocephalus is mostly seen in congenital, infectious and neoplastic properties.
Strengthening: signs of blood-brain barrier destruction of tumors in the brain. a is obviously strengthened, b is strengthened, c is not strengthened, d is strengthened by wall nodules, e is strengthened by annular and flower ring, f is strengthened unevenly, and g is used for blood supply of external cerebral arteries.
The signs are as follows: a-off-white mass blur characteristic: the normal grey white matter has density difference, and can distinguish fuzzy boundary, disease state, disappearance of grey white matter fuzzy boundary, meningeal tail sign and meningioma; b-butterfly signs, generally located on both sides of the midline, distributed according to the corpus callosum fibers, often characteristic signs of corpus callosum tumors; c-white matter extrusion; d-trigonometric sign, high density after superior sagittal sinus embolization.
Fog sign or blur effect: it refers to the special CT expression period of cerebral infarction in which the focus is raised from low density to equal density and the space effect disappears after 2-3 weeks of cerebral infarction.
Comma sign: the method is characterized in that when the upper and lower concrescence of the brain curtain is invaded, the focus is intercepted by the brain curtain to form a special form similar to comma sample application; it is commonly seen in supratentorially developing meningiomas.
2. Lung disease (CT) basic lesion image representation:
A. exudative lesions: 1) the etiology is as follows: physical, chemical and biological factors, ischemia, anoxia, poisoning and the like, and are commonly seen in diseases such as phthisis, pulmonary hemorrhage, pulmonary edema and the like. The exudate can be serous fluid, blood, and leukocyte, erythrocyte and cellulose. The exudates are mainly liquid and can spread among alveoli without obvious boundary with normal lung tissue, the volume of the affected lung tissue is not or slightly reduced, and the lung consolidation change caused by expansion can be seen in Klebsiella pneumonia.
2) The performance is as follows:
different morphologies of lesion range are different, a, spot blur with lesion of 6-8mm oozes in alveolus, b, lobules of 1-2.5cm, c, subsegments, d, and segmental leaves;
the exudates have different densities, wherein a is mainly serous fluid or edema fluid and has lower density, b is mainly pus cell exudation and has higher density, and c is mainly cellulose exudation and has highest density.
3) Exudative lesions are classified as the main lesion, and are absorbed quickly after proper treatment, most of the exudative lesions can be absorbed within 1-2 weeks, and the exudative lesions around the pulmonary tuberculosis lesion can be absorbed obviously within about 4 weeks. A few are aggravated as described later.
The density of the image lower than that of the adjacent blood vessel is called frosted glass density, and the density equal to or higher than that of the adjacent blood vessel is called smear or real change. The process by which gas within the alveoli is replaced by exuding fluid, protein and cells is called consolidation. The bronchi shadow containing gas is seen in the solid-variable area and is called as the air bronchi.
Key words: lung excess, smear, patch, spot, mass, ground glass density, air bronchia.
B. Value-added lesions: 1) and (4) classification: a. Takes fibroblast, vascular endothelial cell and histiocyte hyperplasia as main components, has pathological change of chronic inflammation for infiltrating focus formed by lymphocyte and plasma cell, and has obvious granuloma at the boundary formed by local histiocyte hyperplasia, and the main components are macrophage, c and inflammatory pseudotumor;
2) the performance is as follows: a. nodular, lump, high density of lung segments or lung lobes, the latter is usually smaller than normal lung segments and lung lobes, b is different from exudative lesions, the lesion boundary is clear, most lesions are gathered together without fusion tendency, c, dynamic observation changes slowly, no obvious change exists for months or even years, some lesions can grow slowly, d, the boundary of acinus nodular shadow diameter is less than 1cm (more than 4-7 mm) is clear, the shape is plum petal, and the change is equivalent to the actual change in the range of alveoli;
proliferative lesions are where lung tissue forms granulation tissue dominated by cells and fibers, often localized within the alveoli, clearly demarcated from surrounding normal tissues, including: a, (b) proliferative inflammatory granulomas such as tuberculosis and silicosis nodules, c, inflammatory pseudonodules;
key words: nodules, acinar nodules, bumps.
C. Fibrosis: in the proliferative lesion, the fibrous component can gradually replace the cellular component, and when the focus component is mainly fibrous tissue, the focus is called fibrous focus. Pathologically, it is a reaction to interstitial lung disease.
Classification and etiology: 1) limitations, the healing consequences of chronic pneumonia and tuberculosis;
2) diffuse, complex etiology, collagen disease, scleroderma, rheumatoid, pneumoconiosis, asbestosis, hypersensitivity pneumonitis, chronic bronchitis, b, unexplained pulmonary interstitial fibrosis, also known as idiopathic pulmonary interstitial fibrosis;
3) the result of fibrosis is: may cause dilation of the bronchi and respiratory air cavities.
The performance is as follows: 1) limitation: nodular, lumpy, reticular, linear and cord-like shadows; when the localized fibrosis is expressed as nodules, lumps, lung segments and lung lobe shadows, the fibrosis and the value-added lesion cannot be identified; 2) diffuse fibrosis manifestations, diffusely distributed nodules, networks, linear and cellular shadows, varying in size with emphysema, can be used to provide a diagnosis of diffuse interstitial fibrosis, but it is difficult to determine the cause.
Key words: nodules, masses, lung segments, lung lobe shadows, thickened lung texture, thickened bronchial tube wall, strip shadow, mesh shadow, honeycomb shadow, localized strip shadow (caused by lesion drainage or diffusion), lobular space lines: line A, line B, line C, honeycomb sign.
D. Calcification:
pathologically, it is a degenerative disease, which refers to the abnormal deposition of calcium salts in lung tissue.
1) Generally in degenerative or necrotic tissue, a, mostly in the healing phase of pulmonary or lymph node caseous tuberculosis, partial fungal infections, b, the wall of certain tumors such as pulmonary hamartoma, mediastinal teratoma, c, cysts or parasite cysts;
2) the shuangfei powder is mainly used for treating diseases with calcification, which are mostly seen in a alveolar microsclerosis, primary hyperthyroidism, vitamin D poisoning, and lung metastasis such as osteosarcoma and chondrosarcoma;
3) extrapulmonary pleura such as tuberculous pleurisy, calcification of vessel wall, atherosclerosis, calcification of coronary artery, calcification of pulmonary vascular malformation, etc.;
the performance is as follows: high density, clear and sharp edge, and different sizes and shapes; popcorn pattern calcification is seen in pulmonary hamartoma; can be spot-shaped, block-shaped or spherical and is distributed in a limited and/or dispersed way; portulaca lymphadenopathy eggshell-like calcification found in pneumoconiosis; a lung cyst or a wall of a parasite has a curved calcification or a discontinuous line-like calcification distributed along the wall.
Key words: calcification, intrapulmonary calcification, lymph node, pleural calcification, heart valve calcification, vascular calcification, scattered amorphous, eccentric, punctate calcification in the focus, popcorn pattern calcification, and layered calcification;
E. basic lung lesions:
1) bronchial obstructive changes:
the etiology is as follows: a. in the cavity: including benign and malignant tumors, foreign bodies, inflammation, tuberculosis, congenital stenosis and the like;
b. outside the cavity: including lymph node enlargement and lumpy effect compression of periluminal organ disease.
As a result: a. emphysema: due to partial obstruction, emphysema can cause the alveolar walls to break and fuse into bullae; but also can press the alveolar wall capillaries to cause alveolar wall blood supply disorder.
Emphysema classification and presentation: diffuse, barrel chest, decreased density of both lungs, sparse and thin lung texture, pendulous heart pattern, low and flat diaphragm. Limitation, decreased lung density in the distal end of the occluded bronchus, and sparse lung texture.
b. The lung is atelectasis, also known as lung atrophy. And (4) classification: the etiology is roughly classified into four categories: complete obstruction of bronchus, passive atelectasis, cicatricial atelectasis, adhesive atelectasis;
b.1, complete obstruction of bronchus: after 18-24 hours, the air in the alveolus is absorbed, the alveolus is atrophied, the volume of the corresponding lung tissue is reduced, the lung is changed to be solid, and the lung atrophies faces to the direction of the pulmonary ligament. The mucus retention in bronchus of the lesion part can produce bronchiectasis and also can cause pneumonia;
b.2, passive atelectasis: this is due to external compression of the lungs by the fluid and gas in the chest cavity. The lung tissue close to the pathological changes become more obvious;
b.3, cicatricial atelectasis: the pulmonary fibrosis is caused in a pulmonary fibrosis area, lung atrophy is caused by fiber traction, and the pulmonary fibrosis is mostly manifested as poor lung expansion in different degrees, but simultaneously a great deal of fibrosis, tractive branch and enlargement, localized emphysema, bullous lung and other changes occur;
b.4, adhesive atelectasis: may be plaque-like due to diffuse alveolar collapse, such as discoid atelectasis at the bottom of the lung, and pleural linear atelectasis.
Classifying the anatomical positions: one side of the lung is atelectasis, and the secondary lobule (1-2.5 cm).
Sheet separation: the alveoli of the lung (6-10 mm) are atelectasis.
The concept of real variation: exudative lesions are seen.
Key words:
1-bronchial obstruction, a space occupation, b external pressure, c sputum embolism or blood clot.
2-broncholuminal stenosis.
3-bronchial foreign body obstruction, a positive, b negative.
4-emphysema; diffuse emphysema, pulmonary emphysema of lobes and lung segments, emphysema of subsegments, emphysema of lobules and emphysema of acinus of lung, with or without different sizes.
5-pulmonary bullae.
6-atelectasis: one side of lung atelectasis, b lung lobes and lung segments atelectasis, and c subsegments, lobules and alveolar atelectasis.
7-poor lung distention: a, poor lung expansion on one side, b, poor lung expansion on the lung lobes and lung segments, and c, poor lung expansion on the sub-segment, the lung lobules and the lung acinus.
8-lung excess change.
The method is characterized in that:
the medical image is processed by the method, and the drawing code is input to a computer for doctors to diagnose the state of illness and the category and judge the development degree for treatment.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments or portions thereof without departing from the spirit and scope of the invention.
Claims (6)
1. A computer input method for AI medical diagnosis image picture is characterized in that:
drawing a picture code on the medical image through drawing software, establishing a computer identification model, forming a logic model, and performing supplementary information logic and application verification on the logic model.
2. The AI medical diagnostic image picture computer input and recognition method of claim 1, wherein:
the computer identifies a model: the method comprises the steps of marking existing medical images and interpretation results, establishing unique picture codes for basic lesions and symptoms of each organ of a human body through a drawing tool and corresponding image data expression images, forming picture code files according to the sequence of each organ system by the picture codes, inputting the picture code files to a computer, and completing establishment of a computer identification model.
3. The AI medical diagnostic image picture computer input method according to claim 1, wherein:
the logic model is as follows: by combining with computer picture coding, the logic sequence editing is carried out on the established model of the computer, and the special interpretation terms for conventional medical image recognition are supplemented to form a logic model.
4. The AI medical diagnostic image picture computer input method according to claim 1, wherein:
the supplemental information logic: the verified logic model is combined with other patient data recorded by a computer, such as basic identity information, past medical history, existing symptoms or other biological verification results of the patient, so as to perform logic supplement of the logic model and further improve the interpretation of the logic model.
5. The AI medical diagnostic image picture computer input method according to claim 1, wherein:
the application verification comprises the following steps: comparing, identifying and verifying the edited logic model with the manual interpretation result of the existing medical image database, verifying the reliability of the logic model, and completing verification of the logic model; the verified model is applied, namely the method is compared with the medical image manual identification interpretation of the actual clinical image department for verification, whether the established logic model is reliable and accurate is further verified, and if the established logic model is not deviated from the manual identification, the established logic model is proved to be accurate.
6. The AI medical diagnostic image picture computer input method according to claim 1, wherein: the medical images include lung, liver, brain, chest, head, biliary X-ray, CT, ultrasound scan, and nuclear magnetic resonance images.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010631295.8A CN111739615A (en) | 2020-07-03 | 2020-07-03 | AI medical diagnosis image picture computer input method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010631295.8A CN111739615A (en) | 2020-07-03 | 2020-07-03 | AI medical diagnosis image picture computer input method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111739615A true CN111739615A (en) | 2020-10-02 |
Family
ID=72652734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010631295.8A Pending CN111739615A (en) | 2020-07-03 | 2020-07-03 | AI medical diagnosis image picture computer input method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111739615A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113011462A (en) * | 2021-02-22 | 2021-06-22 | 广州领拓医疗科技有限公司 | Classification and device of tumor cell images |
| CN113838559A (en) * | 2021-09-15 | 2021-12-24 | 王其景 | Medical image management system and method |
| CN114280014A (en) * | 2021-11-30 | 2022-04-05 | 杭州迪英加科技有限公司 | Independent accounting reagent for AI interpretation |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030095692A1 (en) * | 2001-11-20 | 2003-05-22 | General Electric Company | Method and system for lung disease detection |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN106909778A (en) * | 2017-02-09 | 2017-06-30 | 北京市计算中心 | A kind of Multimodal medical image recognition methods and device based on deep learning |
-
2020
- 2020-07-03 CN CN202010631295.8A patent/CN111739615A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030095692A1 (en) * | 2001-11-20 | 2003-05-22 | General Electric Company | Method and system for lung disease detection |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN106909778A (en) * | 2017-02-09 | 2017-06-30 | 北京市计算中心 | A kind of Multimodal medical image recognition methods and device based on deep learning |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113011462A (en) * | 2021-02-22 | 2021-06-22 | 广州领拓医疗科技有限公司 | Classification and device of tumor cell images |
| CN113011462B (en) * | 2021-02-22 | 2021-10-22 | 广州领拓医疗科技有限公司 | Classification and device of tumor cell images |
| CN113838559A (en) * | 2021-09-15 | 2021-12-24 | 王其景 | Medical image management system and method |
| CN114280014A (en) * | 2021-11-30 | 2022-04-05 | 杭州迪英加科技有限公司 | Independent accounting reagent for AI interpretation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111739615A (en) | AI medical diagnosis image picture computer input method | |
| Park et al. | Texture-based automated quantitative assessment of regional patterns on initial CT in patients with idiopathic pulmonary fibrosis: relationship to decline in forced vital capacity | |
| Silva et al. | Pulmonary quantitative CT imaging in focal and diffuse disease: current research and clinical applications | |
| EP2962280A1 (en) | Automatic measurement of lesions on medical images | |
| Balci et al. | Current surgical therapy for bronchiectasis: surgical results and predictive factors in 86 patients | |
| Rolighed et al. | Sternotomy for substernal goiter: retrospective study of 52 operations | |
| McQUAY et al. | Types of death in acute myocardial infarction | |
| Marini et al. | Pictorial review of the pulmonary vasculature: from arteries to veins | |
| Revel et al. | Automated lobar quantification of emphysema in patients with severe COPD | |
| Catellier et al. | Calcific deposits in the heart | |
| Soo et al. | Idiopathic pulmonary fibrosis: current and future directions | |
| Buckert et al. | Myocardial strain characteristics and outcomes after transcatheter aortic valve replacement | |
| Galushka et al. | Case-based tissue classification for monitoring leg ulcer healing | |
| Kawabata et al. | Postoperative subdural air collection is a risk factor for chronic subdural hematoma after surgical clipping of cerebral aneurysms | |
| Valente et al. | Acquired peripheral pulmonary artery aneurysms: morphological spectrum of disease and multidetector computed tomography angiography findings—cases series and literature review | |
| Raoof et al. | Lung imaging in COPD Part 2: emerging concepts | |
| Hara et al. | Impact of 64 multi-detector computed tomography for the evaluation of aortic paraprosthetic regurgitation | |
| Wu et al. | A report of mesothelial/monocytic incidental cardiac excrescences and a literature review | |
| Yu et al. | Prevalence of coronary artery disease in symptomatic patients with zero coronary artery calcium score in different age population | |
| Tongxin et al. | Application research of three-dimensional printing technology and three-dimensional computed tomography in segmentectomy | |
| Dine et al. | The Effect of the Emergency Physicians' Clinical Decision of Targeted Ultrasonography Application in Non‐Traumatic Shock Patients | |
| Shostak et al. | A diagnostic conundrum: progressive tubular lung mass in asymptomatic young woman | |
| Korraa et al. | Outcome of patients with spontaneous pneumothorax admitted in Abbasia Chest Hospital | |
| Aydın et al. | Surgical treatment of congenital lobar emphysema: a report of nine patients | |
| Korol et al. | IP Palii, OO Melnychenko Ukrainian Military Medical Academy, Kyiv, 1Poltava State Medical University, Poltava CLINICAL AND INSTRUMENTAL FEATURES OF DIAGNOSTICS OF COMBAT SURGICAL CHEST INJURY WITH TISSUE DEFECTS |
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 |