Skip to main content
. 2021 Jul 5;202:62–69. doi: 10.1016/j.ymeth.2021.07.001

Fig. 3.

Fig. 3

An example workflow explaining for our detecting architecture: the input array, a three-channel image composed with adjacent CT slices, is transported to the Spatio-temporal encoder in two optional homogenous forms. The key item of the input sequence is immutably located in the middle, and others aim to comprehend the spatial context if they existed.