Pandemic Helper using SingleStore

mpoxsinglestore-arch

Inspiration World Health Organization announced The Financial Intermediary Fund (FIF) - link for Pandemic Prevention, Preparedness and Response finances critical investments to strengthen pandemic prevention, preparedness, and response capacities at national, regional, and global levels, with a focus on low- and middle-income countries. The devastating human, economic, and social cost of COVID-19 has highlighted the urgent need for coordinated action to build stronger health systems and mobilize additional resources for pandemic prevention, preparedness, and response. While there are many institutions and financing mechanisms that support pandemic prevention, preparedness, and response activities, none of them is focused solely on it.COVID-19 has highlighted the pressing need for action to build stronger health systems. “Investing now will save lives and resources for the years to come.

Inspiration of this project is to learn single store ,ingest different types of data (Relational, Full Text, Time Series ,Geospatial etc..) and implement all the capabilities in single product.

1. RELATIONAL 2. TIME SERIES 3. FULL Text Search 4. GEOSPATIAL

What it does

The main goal of this project is to build a single product that will help public to prepare for pandemic - Monkey Pox.

Mpox-EXPLORE

Explore helps the users to understand trends of how cases are increasing over time and symptoms users are experiencing for Monkey Pox Confirmed(Positive), suspected cases. Then ,Extract the medical entities from Monkey Pox related Tweets, Reddits and News using AWS Comprehend medical and explore the symptom trends on various media data.

MEDICAL ENTITIES:

ANATOMY: Detects references to the parts of the body or body systems and the locations of those parts or systems. MEDICAL_CONDITION: Detects the signs, symptoms, and diagnosis of medical conditions. MEDICATION: Detects medication and dosage information for the patient. PROTECTED_HEALTH_INFORMATION: Detects the patient's personal information. TEST_TREATMENT_PROCEDURE: Detects the procedures that are used to determine a medical condition. IME_EXPRESSION: Detects entities related to time when they are associated with a detected entity. Monkey Pox Cases:

Data : globaldothealth provides updated case count - https://github.com/globaldothealth/monkeypox

MONKEY POX CASES Total Number Of Cases each day to understand the spike in cases . Global Cases Count Map to understand in cases volume per each country

When User Clicks on specific country ,He can further identify the specific symptoms people are facing and their time series trends. Also, Dataset includes Source Column - URL of news story or government source where this case was confirmed . We further extract the text from source URL and then utilize AWS Comprehend Medical to extract the medical entities.

Sentiment extracted using AWS Comprehend from Reddits User comments and stacked bar for each sentiments to identify how sentiments have changed over time Medical Conditions and other entities are extracted using AWS Comprehend Medical to understand symptoms, signs over time. Key Phrases are extracted using AWS comprehend and word cloud to display top key phrases. Entities (Organization, Person and Location) is extracted using AWS Comprehend and donut chart is displayed .

User can click on Sentiment bar and understand the symptoms for negative sentiments etc...

MonkeyPox News

Twitter

Pandemic Situational Reports https://www.who.int/publications/m/item/multi-country-outbreak-of-monkeypox--external-situation-report--7---5-october-2022#

Mpox-DETECT: The confirmatory Polymerase Chain Reaction (PCR) tests and other biochemical assays are not readily available in sufficient quantities.

Mpox-Detect helps the user to Quickly detect the Monkey Pox disease and contact the hospital .

1.Firstly, User can Input the General Features like Country he is located ,Gender, Travel History and Symptoms he is facing like Rash, Lesions etc. and then we get the understand the stats from monkey pox cases confirmed data.

2 . Once user understands trends based on symptoms and general features, User can then upload the skin images and get the probability that user might have the disease based on skin images computer-aided monkeypox identification from skin lesion images .Monkeypox Skin Lesion Dataset (MSLD)" is created by collecting and processing images from different means of web-scrapping i.e., from news portals, websites and publicly accessible case reports.

RESCUE:

This will help the user who is requesting for help on platforms like Subreddit - https://www.reddit.com/r/monkeypoxpositive/ where there is specific flair like Medical Help etc... We then extract the Medical entities and entities from these Reddits using AWS Comprehend and Comprehend Medical and further User can drill down from each flairs (ex: Help- Medical) and identify the medical entities (Symptoms etc..) .

MonkeyPox Scientific Literature Dataset contains all available scientific knowledge published on the topic of Monkeypox scraped from PubMed, starting from as early as 1974 to 2022, up to the abstract level.

We then extract the medical entities from the scientific Literature using AWS comprehend medical. Also, Search is done using Single Store Full Text Search Capabilities.

MATCH: MATCH(Abstract) AGAINST ({searchvalue}' HIGHLIGHT:

Then, Users can ask the questions related to issues they are facing and get the related answers from scientific literature.

Also ,Emergency Management can monitor the issues people are facing ,get relevant answers from scientific literature for complex questions and respond to the patients.

How we built it Researched about the all the required relevant data and Built Dashboard using Plotly Dash

Challenges we ran into cannot connect singlestoredb sql server on Heroku Faced Issues storing the image vectors. Time Series

What we learned

Multi Modal Database - https://medium.com/@VeryFatBoy/singlestoredb-the-rise-and-rise-of-multi-model-database-systems-c9abdf54e093

Full Text Search - Different features in Full text search https://medium.com/@VeryFatBoy/using-singlestore-for-full-text-index-and-search-bba1ec96df2b

Geospatial Data-https://medium.com/@VeryFatBoy/using-singlestore-as-a-geospatial-database-28ddf92684af

Time Series Data - https://medium.com/@VeryFatBoy/using-singlestore-as-a-time-series-database-6517a1a36a4e

Learnt a lot about Single Store Capabilities

What's next for Pandemic Helper - Single Store

Learning about Mindsb in memory machine learning and implementing - https://medium.com/@VeryFatBoy/quickstart-to-using-singlestore-db-mindsdb-and-deepnote-for-data-science-1f607f24843d Images Data - https://medium.com/@VeryFatBoy/image-classification-using-singlestoredb-keras-and-tensorflow-b6c3877d7571

Scaling it to handle real time big data