Visual Perception

Question 1

X and y axis of psychometric function of light discrimination

Answer 1

Y= % of light detected by an individual
X= Light intensity

Question 2

Threshold

Answer 2

Brightness of the light at which the patient correctly detects the light 50% of the time.

Question 3

Methods to determine a patient’s threshold

Answer 3

1. Method of ascending limits (how we measure dark adaptation)
2. Method of descending limits (how we usually determine acuity- greatest drawback is anticipation)
3. Method of constant stimuli (examiner varies the intensity of light randomly. Time consuming)
4. Stairstep method (gradually increase light 1 increment at a time until the pt detects light, then decrease until they no longer see it. HVF)
5. Adjustment method (subject adjusts the light stimuli themselves)

Question 4

Individuals with strict vs lax criteria

Answer 4

Strict: They will not press the button unless they know 100% that they see the light. Will lead to false negatives.

Lax: They will press the button a lot. Click happy. Want to finish the test. Will lead to false positives.

Question 5

Forced choice method

Answer 5

The experimenter forces the patient to choose between two alternatives that are presented at the same time. Ex: TAC

Question 6

Psychometric function for 2 alternative forced choice test

Answer 6

Ranges from 50% to 100%
This is because the patient has a 50% chance of guessing correctly between the two alternatives that are presented.

Threshold is located at 75%

Ex: 4 alternative faced choice ranges from 25-100 and the threshold will be at 62.5%

Question 7

Signal Detection Theory

Answer 7

Noise is random and corrupts the signal.
The visual system does not receive the pure signal. They receive a combo of noise + signal, which is shaped like a bell curve. The visual system must separate S from S+N

If signal is weak, there is little separation from S+N
If signal is strong, there is greater separation.
^The distance between S+N and S is called detectability.

Question 8

The total number of people with the disease in the population (Total incidence) add what two values

Answer 8

TP and FN

Question 9

The total number of people without the disease in the population

Answer 9

TN and FP

Question 10

Sensitivity

and false negative rate

Answer 10

The probability the test will accurately detect those patients with the disease. Think TP! Tp is sensitive.

TP/ TP + FN

False negative rate = 1- sensitivity

Question 11

Specificity

Answer 11

The probability the test will accurately detect those patents without the disease.

TN/TN + FP

False positive = 1- specificity

Question 12

Positive predictive value

Answer 12

The probability the patient has the disease if the test is positive

PPV = TP / TP + FP

Question 13

Negative predictive value

Answer 13

The probability the patient does not have the disease if the test is negative

NPV = TN/ TN + FN

Question 14

What does sensitivity and specificity determine. Do we use it often?

Answer 14

Determine how well a test will detect whether the disease is present or absent. Clinically, we are more interested in the probability that our patient has the disease if the test is positive (PPV) or negative (NPV)

Question 15

ROC

Receiver operating characteristic curve X and Y

Answer 15

Y axis is true positives

X axis is false positives

Question 16

ROC determines what

Answer 16

How a patient’s criteria affects detectability of a signal

Patients with lax criteria have high true positives and high false positives. They fall on the upper right part of the curve

Patients with strict criteria have low true positives and low false positives. They fall on the lower left part of the curve.