Lung Cancer Classification in Histopathology Images Using Multiresolution Efficient Nets

Table 4

Comparison with state-of-the-artwork.

Architectures	Parameters	Input size	Classification	Accuracy result (%)
EfficientNetB4 [29]	17 million	Not specified	Binary and multiclass for COVID-19 diagnosis	96
DCNN [3]	60 million	256 × 256	Three classes of lung cancer subtypes	71.1
Residual neural network [8]	0.27 million	50 × 50	Lung cancer type from CT scan images	85.71
Inception-v3 [13]	23 million	512 × 512	Gastric and colonic from histopathological	96
SC-CNN [30]	Not specified	27 × 27	Nuclei in colon cancer histology images	68
EfficientNetB2 (our approach)	9.2 million	260 × 260	For histopathology images of lung and colon cancer (five classes)	97.24