3. Presentation Outline
1. Background and Motivation
2. Aims and Objective
3. Problem Statement / Solution
4. Methodology
5. Relevant Work / Literature Review
6. Visuals
7. Future Work
8. Conclusion
9. Timeline
10. References
4. Background
The entire world has been impacted
by the global pandemic, and
unfortunately, some people have lost their
lives. Even now, in China, there is a
resurgence of COVID- 19 cases, which is
causing breathing problems and is
difficult detecting the type of infection.
Abnormal breathing sounds can be a sign
of serious respiratory illnesses like
Pneumonia, Bronchitis, COVID-19, or
Asthma, which can be fatal if not
detected and treated promptly.
5. Aims and objectives:
Aim of this project is to create a machine learning model that can
accurately identify and classify different type of respiratory sound . By
training the model on large dataset of respiratory sound we hope to
enable it to distinguish between various types of sound those produce by
healthy and affected lungs .
Objectives:
1. To evaluate the effectiveness of machine learning model based on
various factors such as size , computational complexity and classification
performance .
2. To compare the performance of different machine learning model and
determine which ones are most accurate and efficient at classifying
respiratory sound .
3. To Develop user friendly interface that allow patient to easily record
and upload their breathing sound for analysis
4. To enable heath care provider to remotely examine respiratory sound ,
thereby improving access to timely medical care especially in urgent cases
6. Motivation:
We decided to work on this project
Because:
One of the main reasons for our
motivation is the lack of access to
physicians in urgent cases.
Also, for various reasons, some
people are unable to visit a physician in
a timely manner. To address this issue,
people can use a digital stethoscope
that allows them to record their
respiratory sounds and upload them to
our model.
By analyzing these recordings, our
model can determine whether or not a
physician's visit is necessary. This can
help people receive timely medical
attention, even if they are unable to
visit a physician in person.
7. Problem Statement
Every professional is bound
according to its surrounding ,
man who is busy cannot go to the
physician on time due to its busy
schedule , and avoid its health
life .
Problem Solution
We are going to propose a
model which can record
breathing sound and predict the
disease also the severity of the
health state .
10. Future Work
We will offer model files that can be
loaded and used for prediction purpose.
There are numerous approaches in the
field of deep learning, including popular
architechtures such as ResNet, VGG. We
will thoroughly analyze the performance of
each of these algorithms in terms of both
accuracy and latency, in order to determine
the most suitable option for our specific
needs and goals.
12. FYP TIMELINE
Task Name Jan Feb March April May June July Aug Sept
Literature Review
Surveying/Permission Hospitals &
Clinics
Formulating Plans/Budgets For
data Collection
Data Collection For Sites
Data Annotation
Model Training/ Optimization
Thesis Writing