AU2021106628A4 - Image processing based covid-19 prediction using deep learning - Google Patents

Abstract

IMAGE PROCESSING BASED COVID-19 PREDICTION USING DEEP LEARNING ABSTRACT The present invention is related to image processing based COVID-19 prediction using deep learning . The objective of present invention is to solve the abnormalities presented in the prior art techniques related to diagnose of covid-19 disease using radiograph images. 26 DRAWINGS CLASSER Norm - Test InputI s mr ewfetn -1024ltae X-ray imaps Feature Extraction FIGURE 1 27

Description

DRAWINGS

CLASSER Norm

- Test InputI s mrewfetn - 1024ltae X-ray imaps Feature Extraction

FIGURE 1

IMAGE PROCESSING BASED COVID-19

PREDICTION USING DEEP LEARNING FIELD OF INVENTION

[001]. The present invention relates to the technical field of machine

learning based covid detection.

[002]. The present invention relates to the field of image processing of

radiograph.

[003]. Particularly, the present invention relates to the field of

detection of covid-19 using image processing of the 2D Xray image

of the chest of the patient.

[004]. More particularly, the present invention is related to image

processing based COVID-19 prediction using deep learning.

BACKGROUND & PRIOR ART

[005]. The subject matter discussed in the background section should

not be assumed to be prior art merely as a result of its mention in the

background section. Similarly, a problem mentioned in the

background section or associated with the subject matter of the

background section should not be assumed to have been previously

recognized in the prior art. The subject matter in the background

section merely represents different approaches, which in-and-of

themselves may also be inventions.

[006]. In today's world application and acceptance of machine learning has

increased since it shows some promising results regarding automatic

analysis in various medical fields. Machine learning is a subpart of

artificial intelligence (AI) and Deep learning (DL) is a widespread

exploration area of ML and it's a subset of machine learning. It is

responsible to provide end to end output using input dataset, deprived of

physical feature extraction.

[007]. these days, Deep Learning (DL) is a subfield of Al worried about

strategies motivated by neurons of the cerebrum. Today, DL is rapidly

turning into a significant innovation in picture/video order and location.

DL relies upon calculations for thinking process re-enactment also,

information mining, or for creating reflections. Shrouded deep layers on

DL maps inputs information to marks to break down concealed examples

in confounded information.

[008]. Other than their utilization in clinical x-beam acknowledgment, DL

models are likewise utilized in different regions in the utilization of

picture handling and PC vision in clinical. DL improves such a clinical

framework to acknowledge higher results, augment ailment scope, and

executing appropriate continuous clinical picture sickness location

frameworks.

[009]. Deep learning techniques have shown promising results in recent

years to accomplish radiological tasks automated analysis of multimedia

medical images.

[0010].One of the Deep convolution neural networks (DCNNs) strong deep

learning has been widely applied in architecture and such practical

applications pattern recognition and image classification in an intuitive

way. DCNN is able to use four picks to follow, 1) weight training on very

large available datasets; 2) Refine the grid Pre-trained DCNN weights

based on small datasets; 3) Observational pre-training is being applied to

get started Network weight before installing the DCNN model in the

application; and 4) Also called to use pre-trained DCNN being used as an off-shelf CNN feature extractor. In previous studies, we used DCNN classification of X-ray radiation images for the diagnosis of chronic diseases of the patient, equitable contract in cross-sectional photos and with that, the application of deep learning techniques to limit and the revelation of COVID-19 in the chest radiation is not limited to prediction

[0011]. Some of the work listed herewith:

[0012].. 10902955DETECTING COVID-19 USING SURROGATES US

26.01.2021Int.Class G16H 0/80Appl.No 16878433Applicant

GEORGETOWN UNIVERSITYInventor Howard Federoff A triage

system that determines whether a user is likely to have contracted a

disease based on sensor data received from a user device (e.g., a

smartphone or activity tracker). Each symptom is identified by comparing

sensor data to a predetermined baseline and comparing the difference to a

predetermined symptom threshold. Because direct measurement of

symptoms using the sensors available to the user may not be feasible or

sufficiently accurate, the triage system also uses surrogates the identify

certain symptoms. For example, a fever may be identified using heart

data, a cough or shortness of breath may be identified by analyzing

recorded sound, fatigue may be identified by analyzing the movement of

the user device, and loss of smell or taste may be identified by recording sound and using speech detection algorithms to identify phrases in the recorded sound indicative of loss of smell or taste.

[0013].20200214649DETECTION OF COVID-19 US - 9.07.2020Int.Class

A61B 5/OOAppl.No 16824178Applicant HOMAS PAUL

COGLEYInventor THOMAS PAUL COGLEY The steps of the method

are: providing a generator adapted to produce specific wavelengths to

maximize light emissions; providing a scintillator in operative proximity

to the generator, the scintillator having an associated scintillator screen of

a specific phosphor or other excitive type sensitive to various

wavelengths, the scintillator being sensitive to a wavelength that

maximizes light emissions; positioning a patient to be diagnosed between

the generator and the scintillator; emitting a specific wavelength from the

generator to and through the patient onto the scintillator screen whereby

the associated scintillator screen will light up and sparkle to produce a

light image of the patient; providing a camera in operative proximity to

the scintillator screen to record the light image produced on the

scintillator screen; providing a computer with a computer screen and

software; capturing and analyzing the recorded light images from the

camera; and obtaining a diagnosis for treatment of the patient.

[0014].202110172010T BASED COVID DETECTION SYSTEM AND

METHOD THEREOF AU - 29.04.2021Int.Class A61B 5/OOAppl.No

02110172OApplicant hattacharya, RijuInventor A system for IOT based

covid detection, the system comprises of: an infrared temperature sensor

for monitoring a temperature of a user from a certain distance, a plurality

of camera for capturing a face image of the user at regular interval of

time in order to identify a mask on the face of the user, wherein an image

processing module for detecting the mask on the face of the user using a

neural network, a controlling module for comparing a real-time

temperature and distance obtained from the sensor with a threshold value,

and generate a plurality of command signal when the temperature and

distance exceeds a threshold value and the user is without mask, a drone

laser module for focusing a laser beam on the user detected without a

mask, and a communication module for transmitting a plurality of

information to a nearby hospital. FIGURE 3

[0015].111647690RT-RAA PRIMER PAIR AND DIAGNOSTIC KIT FOR

DETECTING COVID-19 VIRUS CN - 11.09.2020Int.Class C12Q

1/70Appl.No 02010578160.XApplicant HUAQIAO

UNIVERSITYInventor IU DANThe invention discloses an RT-RAA

primer pair for detecting COVID-19 virus. An upstream primer of the

RT-RAA primer pair comprises a nucleic acid sequence as shown in SEQ

No 1, and a downstream primerof the RT-RAA primer pair comprises a

nucleic acid sequence as shown in SEQ No 2. A detection method established by adopting the primer pair provided by the invention has the advantages of simplicityin operation, difficulty in cross infection, rapid reaction and accurate and reliable result, is suitable for rapid detection of

COVID-19, and can be used for early diagnosis and early isolation of

epidemics, reducing the infection rate and controlling the spread of the

epidemics.

[0016].111951964METHOD AND SYSTEM FOR RAPIDLY DETECTING

COVID-19 CN - 17.11.2020nt.Class G16H 50/30Appl.No

02010753816.7Applicant SHANDONG UNIVERSITYInventor DENG

WEIQIAO The invention provides a method and system for rapidly

detecting COVID-19, and belongs to the technical field of biomedicine

and data processing. According to the method, COVID-19 patients and

healthypeople are classified by taking the COVID-19 patients and the

healthy people as classification targets and taking exhaled NO

concentration and basic characteristics of the human body as

characteristic quantities at the same time, and the diagnosis correct rate is

up to 90% or above. The method is low in detection cost, high in speed

and high in accuracy, so that the method has good practical application

value.

[0017].111856004REAGENT FOR DETECTING COVID-19 THROUGH

CHEMILUMINESCENCE IMMUNOASSAY AND DETECTION

METHOD THEREOF CN - 30.10.2020Int.Class GOIN 33/569Appl.No

02010218736.1Applicant WEIFANG KANGHUA BIOTECH 0.,

LTD.Inventor YANG ZHITINGThe invention provides a eagent for

detecting COVID-19 through chemiluminescence immunoassay and a

detection method thereof. An indirect process is used, a COVID-19

antigen is coated on a magnetic microsphere, the anti-human IgG/IgM is

labeled with acridinium ester, and the acridinium ester is excited by the

excitation liquid to emit light so as to detect the COVID-19 antibody, and

the detectionmethod combining the acridinium ester or acridinium

sulfonamide compound labeling technology and the superparamagnetic

nano-microsphere labeling technology is adopted; the superparamagnetic

nano-microsphere labeling technology is a technology of coating nano

microspheres with antigens or antibodies, and the superparamagnetic

nano-microsphere labeling technology can enlarge the reaction surface

area by 10 to 50 times and accelerate the reaction efficiency; the reaction

process is suspended in a reaction solution and belongs to homogeneous

reaction, so that the reaction speed can be increased; connection belongs

to chemical connection of bioactive substances, so that the coating

efficiency is higher, and the coating is more stable in a liquid

environment. The detection reagent disclosed by the invention is high in

specificity, high in sensitivity, simple and convenient to separate, rapid, non-toxic, safe and stable, short in detection time by adopting the reagent, high in efficiency andgood in stability.

[0018].20210088469ELECTROCHEMICAL APPROACH FOR COVID-19

DETECTION US - 25.03.2021Int.Class ON 7/327Appl.No

17100903Applicant Mohammad bdolahadInventor Mohammad

Abdolahad A system for diagnosing COVID-19 infection. The system

includes an electrochemical probe with three needle-shaped electrodes

configured to be inserted into a sputum sample, an electrochemical

stimulator-analyzer electrically connected to the electrochemical probe, a

memory having processor-readable instructions stored therein, and a

processor configured to access the memory and execute the processor

readable instructions to perform a method. The method includes applying

a sweeping range of electrical potentials to the electrochemical probe

utilizing the electrochemical stimulator-analyzer, measuring a set of

generated electrical currents versus the applied sweeping range of

electrical potential utilizing the electrochemical stimulator-analyzer,

receiving the set of electrical currents from the electrochemical

stimulator-analyzer, measuring a level of reactive oxygen species (ROS)

in the sputum sample by measuring a current peak of the set of electrical

currents, and detecting a COVID-19 infection status based on the

measured level of ROS. Detecting the COVID-19 infection status includes detecting an infection with COVID-19 if the measured current peak is in a first range of current peaks and detecting a non-infection with

COVID-19 if the measured current peak is in a second range of current

peaks.

[0019].111187863KIT AND DETECTION METHOD FOR DETECTING

COVID-19 WITH DOUBLE-ENZYME METHOD BY ISOTHERMAL

AMPLIFICATION CN - 2.05.2020Int.Class C12Q 1/70Appl.No

02010207363.8Applicant GUANGZHOU DAZHENG BIO-TECH CO.,

LTD.Inventor YANG JUN The invention discloses a kit and a detection

method for detecting COVID-19 with a double-enzyme method by

isothermal amplification. The kit comprises a Forward Primer, a Reverse

Primer, an MNAzymePart A, an MNAzymePart B and Sub2-FB.

According to the detection method disclosed by the invention, on the

basis of an RPA recombinase isothermal amplification technology,

MNAzyme is combined, and specific non-uniform amplification is

adopted to generate sufficient linear DNA, so the MNAzyme can

continuously emit light; the working temperature of the MNAzyme is

extremely wide, and fluorescence can bedetected under the condition of a

room temperature (37 DEG C), so the working temperature of the

MNAzyme is matched with the working temperature (37 DEG C) of RPA

amplification; the MNAzyme has highlight-emitting specificity and catalytic specificity; compared with a traditional RPA isothermal amplification technology, the method provided by the invention does not need to synthesize a complex probe sequence, can set a plurality of

MNAzymes (2-3) to obtain a strong output fluorescence signal, and

avoids the situation that a target nucleic acid signal is weak and cannot be

detected; and the method provided by the invention is simple, convenient

and rapid, can be used together with various instruments, can realize

amplification at 42 DEG C and 37 DEG C, and does not generate noise.

[0020]..202021054173AN INTELLIGENT SYSTEM TO DETECT COVID

19 PROTOCOLS IN - 15.01.2021 Int.Class H04W 16/00Appl.No

202021054173Applicant Ms. Pooja GuptaInventor Ms. Pooja Gupta The

application discloses An Intelligent System to Detect COVID-19

Protocols; this system is design to detect covid-19 suspicious person

using Image Processing system. The system contains a more than one

camera, thermal camera which capture image and send to processing unit,

processing unit check the body temperature, facial mask. Processing unit

check body temperature with a normal human range and face mask found

on person then it moves for further processing by keeping on sanitizing

sparkle for 1-20 sec, otherwise processing unit alarm for notifying covid

19 suspicious person.

[0021].20200340945ELECTROCHEMICAL APPROACH FOR COVID-19

DETECTION US - 29.10.2020 Int.Class GOiN 27/327Appl.No

16924718Applicant Mohammad AbdolahadInventor Mohammad

Abdolahad A method for diagnosing COVID-19 infection of a person.

The method includes acquiring a sputum sample of a person, measuring a

level of ROS in the sputum sample, and detecting a COVID-19 infection

status of the person based on the measured level of ROS. Measuring the

level of ROS in the sputum sample includes recording a cyclic

voltammetry (CV) pattern from the sputum sample and measuring a

current peak of the recorded CV pattern. Detecting the COVID-19

infection status of the person includes detecting COVID-19 infection of

the person responsive to the measured current peak being in a first range

of more than 230 pA and detecting COVID-19 non-infection of the

person responsive to the measured current peak being in a second range

of less than 190 A...

[0022]. Groupings of alternative elements or embodiments of the invention

disclosed herein are not to be construed as limitations. Each group member

can be referred to and claimed individually or in any combination with

other members of the group or other elements found herein. One or more

members of a group can be included in, or deleted from, a group for

reasons of convenience and/or patentability. When any such inclusion or deletion occurs, the specification is herein deemed to contain the group as modified thus fulfilling the written description of all Markus groups used in the appended claims.

[0023]. As used in the description herein and throughout the claims that

follow, the meaning of "a," "an," and "the" includes plural reference unless

the context clearly dictates otherwise. Also, as used in the description herein,

the meaning of "in" includes "in" and "on" unless the context clearly

dictates otherwise.

[0024]. The recitation of ranges of values herein is merely intended to serve as

a shorthand method of referring individually to each separate value falling

within the range. Unless otherwise indicated herein, each individual value is

incorporated into the specification as if it were individually recited herein.

All methods described herein can be performed in any suitable order unless

otherwise indicated herein or otherwise clearly contradicted by context.

[0025]. The use of any and all examples, or exemplary language (e.g. "such

as") provided with respect to certain embodiments herein is intended merely

to better illuminate the invention and does not pose a limitation on the scope

of the invention otherwise claimed. No language in the specification should

be construed as indicating any non-claimed element essential to the practice

of the invention.

[0026]. The above information disclosed in this Background section is

only for enhancement of understanding of the background of the

invention and therefore it may contain information that does not form

the prior art that is already known in this country to a person of

ordinary skill in the art.

SUMMARY

[0027]. The present invention mainly cures and solves the technical

problems existing in the prior art. In response to these problems, the

present invention provides image processing based COVID-19

prediction using deep learning.

[0028]. As one aspect of the present invention relates to a A

computer implemented method for the processing of X-rays image

of chest for diagnose covid-19 disease, wherein the computer

implemented method comprising steps of Gathering X-beam

pictures in one dataset and stacked for scaling at a fixed size of 224

X 224 pixels to be reasonable for additional handling inside a deep

learning pipeline; Applying A One-hot encoding on the names of

picture information to show the instance of positive COVID-19 or

"not" for each picture in the dataset, using a training dataset of

machine learning module, wherein tuned one of eight deep learning

models are used , the pre-processed dataset is 70-30 split in

standard way, That implies 30% of picture information will be

utilized for testing stage, Once more, parting 70% information will

be utilized for building equivalent preparing and approval sets,

wherein the Subsample irregular determinations of preparing

picture information for the deep learning classifier, and afterward

apply assessment measurements to show the recorded exhibition on

the approval set; and Taking in records the testing information of

to the tuned deep learning classifier to arrange all the picture

patches into one of two cases: affirmed positive COVID-19 or

ordinary case (negative COVID-19).

OBJECTIVE OF THE INVENTION

[0029]. The principal objective of the present invention is to provide

image processing based COVID-19 prediction using deep learning.

BRIEF DESCRIPTION OF DRAWINGS

[0030]. Further clarify various aspects of some example embodiments of

the present invention, a more particular description of the invention

will be rendered by reference to specific embodiments thereof which

are illustrated in the appended drawings. It is appreciated that these

drawings depict only illustrated embodiments of the invention and are

therefore not to be considered limiting of its scope. The invention will

be described and explained with additional specificity and detail

through the use of the accompanying drawings.

[0031]. In order that the advantages of the present invention will be

easily understood, a detailed description of the invention is discussed

below in conjunction with the appended drawings, which, however,

should not be considered to limit the scope of the invention to the

accompanying drawings, in which:

[0032]. Figure 1 shows an exemplary representation of system for

product evaluation through comments analysis using machine

learning, according to the present invention

[0033]. Figure 2 shows Overview of Present Deep Learning framework

for categorising the X-Ray image data and result of the present

invention..

DETAIL DESCRIPTION

[0034]. The present invention discloses image processing based

COVID-19 prediction using deep learning.

[0035]. Figure 1 shows the exemplary representation of image

processing based COVID-19 prediction using deep learning

according to the present invention.

[0036]. Although the present disclosure has been described with the

purpose of two smart frameworks for providing privacy and protection

in block chain based private transactions using cloud computing

approach, it should be appreciated that the same has been done merely

to illustrate the invention in an exemplary manner and to highlight any

other purpose or function for which explained structures or

configurations could be used and is covered within the scope of the

present disclosure.

[0037]. An image processing based COVID-19 prediction using deep

learning is disclosed.

[0038]. The computer implemented method for the processing of X

rays image of chest for diagnose covid-19 disease is disclosed

herewith.

[0039]. The computer implemented method comprising steps of

Gathering X-beam pictures in one dataset and stacked for scaling at a

fixed size of 224 X 224 pixels to be reasonable for additional handling

inside a deep learning pipeline;

[0040]. A One-hot encoding on the names of picture information is

applied to show the instance of positive COVID-19 or "not" for each

picture in the dataset, using a training dataset of machine learning

module, wherein tuned one of eight deep learning models are used.

[0041]. The pre-processed dataset is 70-30 split in standard way, That

implies 30% of picture information will be utilized for testing stage,

Once more, parting 70% information will be utilized for building

equivalent preparing and approval sets.

[0042]. The Subsample irregular determinations of preparing picture

information for the deep learning classifier, and afterward apply

assessment measurements to show the recorded exhibition on the

approval set; and

[0043]. The testing information of to the tuned deep learning classifier is

put on records to arrange all the picture patches into one of two cases:

affirmed positive COVID-19 or ordinary case (negative COVID-19).

[0044]. The Deep COVID-Net structure including deep learning

classifiers.

[0045]. The Deep transfer learning approach is applied for the

identification of COVID-19 using limited dataset, uses a 5-fold cross

validation is used for better testing and validation..

[0046]. The Graph Convolution Network (GCN)is used to build a large

and heterogeneous comment word graph which contain word nodes

and comment nodes so that global word co-occurrence can be

explicitly modeled and graph convolution can be easily adapted.

[0047]. As another embodiment of present invention a method for

Product or service Evaluation Using Comment Analysis in online

ecommerce website is disclosed in this disclosure.

[0048]. The method comprising steps of extracting a plurality of

comments viewpoint in user comment section of ecommerce website

using a data extraction module.

[0049]. A data set module is prepared using a plurality of comment

viewpoint is input in sentiment classification model.

[0050]. The semantic similarity is calculated between each comment

viewpoint in each emotion class comment using a deep learning

module using the deep learning module using data processing

module.

[0051]. An index on product is assigned according to the result of the

data processing module.

[0052]. The present model used in the present invention is

utilizing eight deep CNNs models that were chosen in this research

for predict the accurate result. The main objective of this invention

is to propose an efficient deep transfer learning algorithm for

Covid-19 recognition in the chest using 2D x-ray.

[0053]. To make the model predictable, huge numbers of chest X

beam images collected for experiment purpose. Frameworks based

on deep convolution neural network to support and diagnose

COVID-19 primarily as classification task.

[0054]. These deep learning models are used to classified deep

features. Deep transfer learning approach is applied for the

identification of COVID-19 using limited dataset. Also in the

present invention 5-fold cross validation is used for better testing

and validation.

[0055]. All X-beam pictures have been gathered in one dataset and

stacked for scaling at a fixed size of 224 X 224 pixels to be

reasonable for additional handling inside the deep learning

pipeline. One-hot encoding is then applied on the names of picture

information to show the instance of positive COVID-19 or "not"

for each picture in the dataset. So as to begin the preparation period

of chose and additionally tuned one of eight deep learning models,

the pre-processed dataset is 70-30 split in standard way. That

implies 30% of picture information will be utilized for testing

stage. Once more, parting 70% information will be utilized for

building equivalent preparing and approval sets. Subsample

irregular determinations of preparing picture information for the

deep learning classifier, and afterward apply assessment

measurements to show the recorded exhibition on the approval set.

[0056]. In the last advance of the present structure, the testing

information is taken care of to the tuned deep learning classifier to

arrange all the picture patches into one of two cases: affirmed

positive COVID-19 or ordinary case (negative COVID-19), as

appeared in Figure 2.

[0057]. Toward the finish of the work process, the general

execution investigation for every deep learning classifier will be

assessed dependent on the measurements depicted in the

accompanying segment. The Deep COVID-Net structure including

deep learning classifiers have been actualized utilizing Python and

the Keras bundle with TensorFlow2 over a Google Collaboratory

(CoLab) Linux server with Ubuntu 18.04 working framework

utilizing K80 GPU layout card. What's more, the analyses were

executed utilizing the graphical preparing unit (GPU) and RAM

with 12 GB and circle with 107 GB, individually..

[0058]. The figures and the foregoing description give examples of

embodiments. Those skilled in the art will appreciate that one or

more of the described elements may well be combined into a single

functional element. Alternatively, certain elements may be split into

multiple functional elements. Elements from one embodiment may

be added to another embodiment.

[0059]. For example, order of processes described herein may be

changed and are not limited to the manner described herein.

Moreover, the actions of any block diagram need not be

implemented in the order shown; nor do all of the acts need to be

necessarily performed. Also, those acts that are not dependent on

other acts may be performed in parallel with the other acts. The

scope of embodiments is by no means limited by these specific

examples.

[0060]. Although implementations of the invention have been described

in a language specific to structural features and/or methods, it is to

be understood that the appended claims are not necessarily limited to

the specific features or methods described. Rather, the specific

features and methods are disclosed as examples of implementations

of the invention.

Claims (3)

CLAIMS I/We claim:

1. A computer implemented method for the processing of X-rays

image of chest for diagnose covid-19 disease, wherein the

computer implemented method comprising steps of:

Gathering X-beam pictures in one dataset and stacked for scaling

at a fixed size of 224 X 224 pixels to be reasonable for additional

handling inside a deep learning pipeline;

Applying A One-hot encoding on the names of picture

information to show the instance of positive COVID-19 or "not"

for each picture in the dataset, using a training dataset of machine

learning module, wherein tuned one of eight deep learning models

are used , the pre-processed dataset is 70-30 split in standard way,

That implies 30% of picture information will be utilized for

testing stage, Once more, parting 70% information will be

utilized for building equivalent preparing and approval sets,

wherein the Subsample irregular determinations of preparing picture information for the deep learning classifier, and afterward apply assessment measurements to show the recorded exhibition on the approval set; and

Taking in records the testing information of to the tuned deep

learning classifier to arrange all the picture patches into one of

two cases: affirmed positive COVID-19 or ordinary case

(negative COVID-19).

2. The computer implemented method for the processing of X

rays image of chest for diagnose covid-19 disease as claimed in

claim 1, the Deep COVID-Net structure including deep learning

classifiers.

3. The computer implemented method for the processing of X

rays image of chest for diagnose covid-19 disease as claimed in

claim 1, wherein the Deep transfer learning approach is applied

for the identification of COVID-19 using limited dataset, uses a 5

fold cross validation is used for better testing and validation..

EDITORIAL NOTE 23 Aug 2021

2021106628

There are 2 pages of drawings only.