UltrasondConfienceMap

orienting to ultrasound image, model ultrasound image quality and decrease detected false positive disease candidate in shadow area our code refer to the paper of Karamalis, Athanasios, et al., in this paper¹, author inspired by the random walker, so my implement refer the python code of random walker segmentation in scipy ¹可以效率的书写文档

2d randolm walks

3d random walks

2d confidence map

3d confidence map

3d condidence map decrease false positive tumor candidate

green label is false positive tumor candidate in the shadow of ultrasound image, red is positive tumor candidate.

Footnotes

1. Karamalis A, Wein W, Klein T, et al. Ultrasound confidence maps using random walks[J]. Medical image analysis, 2012, 16(6): 1101-1112. ↩ ↩²