Research Article
Constructing Benchmark Databases and Protocols for Medical Image Analysis: Diabetic Retinopathy
Table 1
Summary of the current state of the reference image databases in terms of the key questions addressed in Section
2.1.
| Key questions | STARE (vessel) | STARE (disc) | DRIVE | MESSIDOR | CMIF | ROC | REVIEW |
| C2: “Is there a data set for which the correct answers are known?” | x | | x | x | | x | x | C3: “Are there data sets in common use?” | x | x | x | x | x | x | x | C4: “Are there experiments which show algorithms are stable and work as expected?” | x | | x | | | x | | C5: “Are there any strawman algorithms?” | x | x | x | | | | | C6.1: “What code is available?” | | | | | | x | | C6.2: “What data is available?” | x | x | x | x | x | x | x | C7: “Is there a quantitative methodology for the design of algorithms?” | | | | | | | | C8.1: “What should we be measuring to quantify performance?” | x | x | x | | | x | x | C8.2: “What metrics are used?” | | x | x | | | x | x |
| | 6 | 5 | 7 | 3 | 2 | 7 | 5 |
|
|