COVID-19 DETECTION USING DEEP LEARNING AND CNN WITH DATA ANALYSIS, OUTBREAK PREDICTION AND VISUALISATION.

Features Implemented:-

GOALS OF THE PROJECT :-

Successful detection of Covid-19 using Chest X-Rays by building a Convolutional Neural Network (CNN) and visualising the world data using Covid-19 Trends.

CODE EDITOR (ANY ONE AMONGST THESE)

• Pycharm
• Google Colab
• Jupyter Notebooks
• Anaconda Navigator

PACKAGES USED =>

• NUMPY
• PANDAS
• MATPLOTLIB
• FOLIUM
• PLOTLY
• KERAS (DENSE, CONV2D, MAXPOOL2D, DROPOUT, FLATTEN, MODELS, SEQUENTIAL,etc).

DATASET AVAILABILITY =>

• COVID - POSITIVE CASES => https://www.kaggle.com/datasets/paultimothymooney/chest-xray-pneumonia
• COVID - NEGATIVE CASES => Scraped Data from :-
  1> https://radiopaedia.org/
  2> https://www.sirm.org/category/senza-categoria/covid-19/
  3> https://www.eurorad.org/
  4>https://coronacases.org/
  

For chest X-Ray dcm, jpg, or png are preferred. Not only X-Rays but CT-Scans can also be used as a dataset for the project.

AIM AND GOALS :-

PART-1 (SARS-COV-2 DETECTION)=>

• Blood tests are costly (not affordable by all sections of the society).
• Blood tests take time to conduct (approx 5 to 6 hours per patient).
• Extent of The Spread In the Body Can be detected using Deep Learning Models And CNN.
• Classify using Image Classification Models And Segmentation Techniques and prediction of COVID Positive or Negative Verdict.

PART-2 (DATA TRENDS ANALYSIS)=>

Visualising the preprocessed data set under the following heads :-
• WorldWide COVID-19 Cases.
• Cases Density Animation On World Map Along With Time Lapses.
• Cases over the Time With Area Plot.
• Using Folium Maps for presenting Confirmed, Active, Recovered, Death, New Cases, Population, Cases/Million with Choropleth Maps.
• Deaths And Recoveries Per 100 Cases.
• Top 15-20 Countries Data Analysis, Scatter Plots, Line Plots, Pyplots, Bar Plots, Tree Map Analysis And Growth Rates ( x days from 100+ cases, x days from 1000+ cases, x days from 100,000+ cases ).
• Confirmed Cases Country And Day-Wise Visualisation.
• COVID-19 Pandemic Comparison With Other Similar Epidemics (SARS, EBOLA, MERS, H1N1).

NOTE :- All of the plots and visualisations are completely interactive and customised (Zoom-In, Zoom-Out, Transcend, etc, Features shall be made available).

PART-3 => (COVID-19 SPREAD AND OUTBREAK ANALYSIS IN INDIA USING MACHINE LEARNING).

Reading files by creating dataframes using Pandas.

Using Coordinates of Indian States And Union-Territories and grouping day by day data of countries like India, Italy, Korea, and Wuhan, etc.

Progression Of the Cases in our country India as compared to other countries.

Visualisation Inference amongst the various states such as => (Kerala crossing Maharashtra in terms of highest number of confirmed cases, Haryana and telangana having highest count of confirmed Foreign National Count, top 3 states with maximum number of confirmed cases).

Affected States/Union Territories in a short time span thereby, extending it to a larger dataset.

Visualise the spread in our country graphically for analysing the present condition.

How are the CoronaVirus cases rising? (Linear, Exponential, Logarithmic, Parabolic, etc).

PART-4

• Connecting the Deep Learning Model to A Web-Based Application wherein the user shall be able to make choices which data to be rendered or visualised on the screen.
• Deploying Tech Stack => Heroku Or Netlify.
Technologies Or Tools To Be Used :-
• ReactJS(Maybe for enhanced GUI)
• HTML
• CSS
• Github, etc.

Visualiser Website Deployed At :- https://covid-visualisation-2019.netlify.app/

Complete Integrated Website Deployed At :- https://taupe-arithmetic-480e9c.netlify.app/