Skip to main content

Table 6 A comparison on running speed and memory footprint for ResNet-18 (2D-CNN) and a 3D-CNN model

From: Predicting microvascular invasion in hepatocellular carcinoma: a deep learning model validated across hospitals

 

Input size

NVIDIA GTX 1650

Intel i7–9700

Intel i5–8265

Memory usage

3D-CNN in Jiang et al.

16x64x64/patient

0.0154 s

0.5009 s

0.6802 s

90.91 MB

2D-CNN in this study

16 slices/patient

0.0052 s

0.3890 s

0.5069 s

44.69 MB

6 slices /patient

(average)

0.0051 s

0.1546 s

0.2213 s

49 slices/patient

(max)

0.0082 s

1.1347 s

1.6178 s

  1. The running speed reported in the table is an averaged speed obtained from 100 runs using NVIDIA GeForce GTX 1650 GPU (with 4 GB dedicated video memory), Intel Core i7-9700 CPU, or Intel Core i5-8265 CPU devices.