A Sparse-modeled ROI for GLAM Construction in Image Classification Problems—A Case Study of Breast Cancer - Computational Intelligence Techniques for Comparative Genomics - page 51

Fig. 5 Observed timings of 100 images

Fig. 6 Average times of two
models

region. For example, assume that the observed minimum and maximum values of
x and y coordinates of segmented portion in sequence are 10, 20, 140 and 220.
Then, the

220. The time taken by the 100
mammograms for the construction of GLAM is reported in Table 1 . After exper-
imentation, in average, sparse model completes the construction of GLAM in
113.911 s whereas time taken by the choice of

xed window is from 10 × 20 to 140 ×

xed window for ROI is 280.86 s. In

speci c, the average time taken by sparse-modeled ROI for the selected images in

Fig. 1 is 10.5, 434, 072 and 975 s whereas the existing rectangular-shaped ROI is
328.2470765 s. Figure 5 represents the timings of the two models for 100 mam-
mograms, and Fig. 6 depicts the mean time taken by the two models for 100
mammograms. Results have shown that existing

xed window model takes more

than double the time than proposed sparse model.