Formative Quiz on Presentation by F.San and Jyri

Last updated about 2 years ago

10 questions

Note from the author:

Prepared for ELEC-E5531

Instructions

Please answer the questions below.

Required

Library

Formative Quiz on Presentation by F.San and Jyri

Last updated about 2 years ago

10 questions

Note from the author:

Prepared for ELEC-E5531

Instructions

Please answer the questions below.

Required

Question 1

Which of the following is NOT a common symptom of ataxia?

Slurred speech

Difficulty with balance and coordination

Loss of vision

Tremors in the arms and legs

Required

Question 2

Why did the researchers of this paper use low-pass Chebishev filter for pre-processing?

To detect speech boundaries

To reduce all the noise above 1 kHz

To fine tune the threshold short-term energy

Required

Question 3

One of the features chosen was “spectral centroid”. What does that “spectral centroid” refer to?

The measure of spectral density.

The point where most of signal energy is contained.

The highest point in a frequency spectrum.

The range of frequencies covered by a signal.

Required

Question 4

Familiarize yourself with the speech of an ataxia person by listening to this 1 minute video.

I've watched it.

I didn't watch it.

Required

Question 5

Based on this video, how long did it take for Tallulah to be diagnosed with ataxia?

3 months

1 year 

3 years

10 years

Required

Question 6

What was a limitation of this study?

Lack of variability of speech disturbance score

Small number of available subjects

Sometimes same clinical scores were used for different classes (e.g. healthy & low severity)

All of the above

Required

Question 7

Why do we need a AI-powered method to diagnose ataxia?

For some patients, it takes so long to be diagnosed with traditional methods.

Expert listener might sometimes be wrong.

Every expert listener rates the parameters differently (rater variability).

For monitoring, patients need to go to clinic frequently, which is inconvenient.

In countries where free healthcare is not provided, frequent visits may be costly.

Ataxia can only be diagnosed with AI.

Required

Question 8

Which was not one of the performance metrics used in the study?

Recall

R2 value

Precision

F1 score

Accuracy 

Required

Question 9

Which one of the options is wrong for the model described in the paper?

It can sometimes confuse low severity patients with high severity patients. 

It can sometimes confuse healthy patients with low severity patients.

It can sometimes confuse healthy patients with high severity patients. 

Required

Question 10

10.

Listen this ataxia person's speech. Then compare your own speech with it. Write down 3 differences.

Question 1

Which of the following is NOT a common symptom of ataxia?

Slurred speech

Difficulty with balance and coordination

Loss of vision

Tremors in the arms and legs

Question 2

Why did the researchers of this paper use low-pass Chebishev filter for pre-processing?

To detect speech boundaries

To reduce all the noise above 1 kHz

To fine tune the threshold short-term energy

Question 3

One of the features chosen was “spectral centroid”. What does that “spectral centroid” refer to?

The measure of spectral density.

The point where most of signal energy is contained.

The highest point in a frequency spectrum.

The range of frequencies covered by a signal.

Question 4

Familiarize yourself with the speech of an ataxia person by listening to this 1 minute video.

I've watched it.

I didn't watch it.

Question 5

Based on this video, how long did it take for Tallulah to be diagnosed with ataxia?

3 months

1 year 

3 years

10 years

Question 6

What was a limitation of this study?

Lack of variability of speech disturbance score

Small number of available subjects

Sometimes same clinical scores were used for different classes (e.g. healthy & low severity)

All of the above

Question 7

Why do we need a AI-powered method to diagnose ataxia?

For some patients, it takes so long to be diagnosed with traditional methods.

Expert listener might sometimes be wrong.

Every expert listener rates the parameters differently (rater variability).

For monitoring, patients need to go to clinic frequently, which is inconvenient.

In countries where free healthcare is not provided, frequent visits may be costly.

Ataxia can only be diagnosed with AI.

Question 8

Which was not one of the performance metrics used in the study?

Recall

R2 value

Precision

F1 score

Accuracy 

Question 9

Which one of the options is wrong for the model described in the paper?

It can sometimes confuse low severity patients with high severity patients. 

It can sometimes confuse healthy patients with low severity patients.

It can sometimes confuse healthy patients with high severity patients. 

Question 10

10.