Retrieval of Brain Tumors with Region-Specific Bag-of-Visual-Words Representations in Contrast-Enhanced MRI Images

<table>Visual vocabularies constructed in the margin region with a patch size of <svg height="10.925" id="M35" style="vertical-align:-0.1638pt" version="1.1" viewbox="0 0 33.974998 10.925" width="33.974998" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g transform="matrix(.017,-0,0,-.017,.062,10.65)"><path d="M447 623l8 -12l-283 -613l-74 -10l-7 11q174 283 297 551h-216q-48 0 -62.5 -12t-33.5 -63h-29q10 60 18 148h382z" id="x37"></path></g><g transform="matrix(.017,-0,0,-.017,11.996,10.65)"><path d="M528 54l-36 -38l-198 201l-198 -201l-36 38l197 200l-197 201l36 38l198 -202l198 202l36 -38l-197 -201z" id="xD7"></path></g><g transform="matrix(.017,-0,0,-.017,25.748,10.65)"><path d="M447 623l8 -12l-283 -613l-74 -10l-7 11q174 283 297 551h-216q-48 0 -62.5 -12t-33.5 -63h-29q10 60 18 148h382z" id="x37"></path></g>
</svg> and a vocabulary size of 200. (a) Patches with normalization and whitening. (b) Patches without normalization and whitening.</table>

Computational and Mathematical Methods in Medicine

fig6

Figure 6

Figure 6: Retrieval of Brain Tumors with Region-Specific Bag-of-Visual-Words Representations in Contrast-Enhanced MRI Images 