Spatial Compounding of 3-D Fetal Brain Ultrasound Using Probabilistic Maps

摘要: A new method to address the problem of shadowing in fetal brain ultrasound volumes is presented. The proposed approach based on spatial composition multiple 3-D head projections using weighted Euclidean norm as an operator. support vector machine, which trained with optimal textural features, was used assign weighting according posterior probabilities tissue and shadows. Both phantom real were compounded previously reported operators compared validate it. quantitative evaluations revealed increases signal-to-noise ratio ≤35% contrast-to-noise ≤135% data. Qualitative comparisons made by obstetricians indicated that this novel adequately recovers improves visibility main cerebral structures. This may prove useful both for monitoring diagnosis defects. Overall outperforms methods reported.