Skip to main content
EJNMMI Research

Table 2 Readers visual analyses, majority vote results, and mean interpretation duration

From: Relevance of 18F-DOPA visual and semi-quantitative PET metrics for the diagnostic of Parkinson disease in clinical practice: a machine learning-based inference study

 

Diagnostic performances

Mean interpretation duration (min) [95%CI]

Reader 1

TP = 30.0% (33/110)

FP = 3.6% (4/110)

1.09 [1.06–1.11]

 

FN = 13.6% (15/110)

TN = 52.7% (58/110)

 

Reader 2

TP = 29.1% (32/110)

FP = 3.6% (4/110)

0.99 [0.95–1.03]

 

FN = 14.5% (16/110)

TN = 52.7% (58/110)

 

Reader 3

TP = 28.2% (31/110)

FP = 3.6% (4/110)

1.44 [0.95–1.03]

 

FN = 15.5% (17/110)

TN = 52.7% (58/110)

 

Reader 4

TP = 28.2% (31/110)

FP = 10.9% (12/110)

1.97 [1.92–2.01]

 

FN = 15.5% (17/110)

TN = 45.5% (50/110)

 

Reader 5

TP = 20.0% (22/110)

FP = 2.7% (3/110)

0.7 [0.65–0.75]

 

FN = 23.6% (26/110)

TN = 53.6% (59/110)

 

Majority vote

TP = 29.1% (32/110)

FP = 3.6% (4/110)

1.24 [1.20–1.28]*

 

FN = 14.5% (16/110)

TN = 52.7% (58/110)

 
  1. Mean interpretation time was expressed with 95% confidence interval [95%CI]
  2. TP true positive, FP false positive, FN false negative, TN true negative
  3. *Pooled mean interpretation duration across the five readers
Back to article page