Algorithm 1: Preprocessing techniques for the multimodal breast images

Input: Datasets (D)$({D_{\mathrm{MGI}}, D}_{\mathrm{USI}}$)

# Step 1: Image size standardization:

for each dataset (${D_{\mathrm{MGI}}, D}_{\mathrm{USI}}$) in D:

for each image ($I_i$) in the dataset:

I_i_prime = Resize ($I_i,widt{h}_{out},heigh{t}_{out}$)

# Step 2: Normalization using Box-Cox transformation

for each dataset ($D_{MGI},D_{USI}$) in D:

for each image ($I_i$) in the dataset:

transformed_image = apply_BoxCoxTransformation $\left(I_i\right)$

def apply_BoxCoxTransformation(image):

$\lambda$ = 0.5

if $\lambda$ ! = 0:

$\mathrm{transformed}\text{\_}\mathrm{image}={\displaystyle \frac{{\mathrm{image}}^{\mathsf{\lambda }}-1}{\mathsf{\lambda }}}$

else:

$\mathrm{transformed}\text{\_}\mathrm{image}=\log \left(\mathrm{image}\right)$

return transformed_image