Home
Prepared by
Group Name: Healers (Team 10)
Upkar Singh
109965602
[email protected]
Pranav Arora
110663117
[email protected]
Ben Corrales
11068948.
[email protected]
Vraj Mistry
109951458
[email protected]
Angel Valdez
109388402
[email protected]
Instructor:
Dr. Alex Lui
Course:
CSCI 150
Lab Section:
Section 2 - Thursday (3:30 - 4:20PM)
Teaching Assistant:
Joshly Johnson
Contents
CONTENTS II
REVISIONS II
1 [Introduction]
1.1 Document Purpose 1
1.2 Product Scope 1
1.3 Intended Audience and Document Overview
1.4 Definitions, Acronyms, and Abbreviations
1.5 Document Conventions
1.6 References and Acknowledgments
2 OVERALL DESCRIPTION
2.1 Product Overview
2.2 Product Functionality
2.3 Design and Implementation Constraints
2.4 Assumptions and Dependencies
3 SPECIFIC REQUIREMENTS
3.1 External Interface Requirements
3.2 Functional Requirements
3.3 Use Case Model
4 OTHER NON-FUNCTIONAL REQUIREMENTS
4.1 Performance Requirements
4.2 Safety and Security Requirements
4.3 Software Quality Attributes
5 OTHER REQUIREMENTS
APPENDIX A – DATA DICTIONARY
APPENDIX B - GROUP LOG
Wellness Support Center (WSC) is a web application that allows its users to run features containing medical information resources
Many individuals initially experiencing changes in health are not always equipped with the knowledge of treatment pathways. The Wellness Support Center is a hub for health information resources. By utilizing the power of data science the WSC benefits individuals looking to facilitate some of the legwork within health care. It provides users with the functionality to diagnose symptoms and provides the resources to efficiently continue investigating health issues.
Each user has the ability to log in and is provided access to supervised learning models. The functionality of providing queries to our symptom classifier helps users identify unknown health issues. Our health clinic recommendation system allows for a pertinent professional to now be contacted for further consultation. Thus far these features should not be considered and are not substitute for professional advice. The data sets were acquired from public sources; for pill classification the pill image data came from the national library of medicine https://lhncbc.nlm.nih.gov/news/pill-image-collection-available-open-access-download, For the diagnosing feature we acquired a symptom-disease data set from: https://www.kaggle.com/data/55703
This wiki works to cover the main perspective of the document i.e documentation. This documentation is focused on breaking the section down to better illustrate the purpose and functioning of the application. The requirements are considered in the beginning and then the steps are broken down to achieve the required application.
References and Acknowledgments
https://www.perforce.com/blog/alm/how-write-software-requirements-specification-srs-document
https://krazytech.com/projects/sample-software-requirements-specificationsrs-report-airline-database
A unique approach to the interface between statistical classifiers, the founding idea of the Wellness Support Center is closely paralleled to the likes of WebMD. It stands to be a hub for various health-related information resources. This application however is self-contained and corresponds only to itself. The entirety of WSC as a web application consists of just three main components; a client, server, and database.
The client acts in the form of a dynamic web page. User experience and input are made more interactive through standard styling techniques. Once logged in, a user is given a selection of features in order to submit queries on existing supervised learning agents.
Interactions with a user are the largest task within WSC, for the purpose of dealing with serialized learning agents, the micro-framework Flask was chosen. Flask is responsible for handling all incoming requests between the client and the database and just to the client alone. This unique framework allowed the inclusion of many helpful libraries to facilitate workflow.
The current database utilizes the SQLite library for a lightweight yet effective SQL database. This is where user accounts are stored. The accompanying result vectors from model queries are also stored within different tables of the database with a user identifier as a key amongst tables.
The Wellness Support Center is a website that lets user:
#Find nearby hospitals.
#Check their symptoms for potential disease.
#Use ML model to identify any medicines once the user uploads the picture.
Login/Sign-up page, feature selection page, feature input pages, result pages.
Initial login/sign-up page:
Description: A user arriving at the hosted landing page will be prompted to sign up or use an existing log-in.
Rationale: For the allowed use of model queries.
Dependency: None
Main page:
Description: Given a user has successfully logged in, they will be prompted to review previously administered outputs as well as select a feature to query.
Rationale: For the user to select application functionalities.
Dependency: UIR1
Model page:
Description: Given a user selects a specific model to use, this provides them the appropriate input means and acts to display model output.
Rationale: In order for the user to interact with models.
Dependency: UIR2
Clinician Recommendation page:
Description: Given a user has recently or at one point successfully used the diagnosing feature, a doctor's office recommendation is displayed after.
Rationale: For the user to have a start point in locating the appropriate clinician.
Dependency: UIR3
Overview page:
Description: A summary of past recommendations and model results for user to reference
Rationale: In order to retrieve past executed episodes
Dependency: UIR4
User Registration
Description: Given a user has navigated to the landing page for the first time, they must create a user log-in.
Rationale: For access to site login.
Dependency: None
User Log-in
Description: Given a user has created a log-in, they must be provided input fields to be validated by database values
Rationale: To receive access to application feature functionalities
Dependency: FR1
Diagnosing Feature
Description: Users select from a range of possible symptoms. This information is then used to query a decision tree model.
Rationale: In order for a user to receive medical information of possible disease
Dependency: FR2
Pill Classification Feature
Description: Users may submit personal images of their own or found pills in order to receive information on their use and side effects.
Rationale: In order for the user to identify medication
Dependency: FR2
Clinician Recommendation
Description: Given users have successfully completed the Diagnosing Feature it will prompt a search for clinicians with the most skill relative to the prognosis given.
Rationale: To find the next possible treatment pathway of a user
Dependency: FR3
This is a web application, so the requirement is a browser, preferably Google Chrome, because all the testing of the front-end was done on it. The IDE we used is Visual Studio Code. Software requirements are:
- Anaconda Navigator
- Python 3.7
- Visual studio code
- SQLlite
Safety and Security Requirements
NFR 2: Password must have at least 8 characters long with at least one uppercase letter, one lowercase letter, one number, and one unique symbol
Symptom list presented as a scroll capable checkbox form
Machine Learning Model
Google Map API