Table 6 The effect of sliding window tile stride on slide-level tasks using the EfficientNet-B3 architecture.
From: Overcoming the limitations of patch-based learning to detect cancer in whole slide images
Stride | Dice (% overlap) | Error in \(\sqrt{{d}_{1}{d}_{2}}\) | Cfs. mtx |
|---|---|---|---|
512 | 48 | 9.38 | \(\left(\begin{array}{cc}1& 7\\ 0& 12\end{array}\right)\) |
256 | 71 | 1.52 | \(\left(\begin{array}{cc}6& 2\\ 0& 12\end{array}\right)\) |
128 | 62 | 2.45 | \(\left(\begin{array}{cc}5& 3\\ 0& 12\end{array}\right)\) |
64 | 66 | 1.65 | \(\left(\begin{array}{cc}4& 4\\ 0& 12\end{array}\right)\) |
32 | 64 | 1.63 | \(\left(\begin{array}{cc}3& 5\\ 0& 12\end{array}\right)\) |
16 | 61 | 1.66 | \(\left(\begin{array}{cc}3& 5\\ 0& 12\end{array}\right)\) |
