DCE-MRI and DWI Integration for Breast Lesions Assessment and Heterogeneity Quantification
In order to better predict and follow treatment responses in cancer patients, there is growing interest in noninvasively characterizing tumor heterogeneity based on MR images possessing different contrast and quantitative information. This requires...
more
|
Full text:
|
|
|
Link for citation:
|
|
|
In order to better predict and follow treatment responses in cancer patients, there is growing interest in noninvasively characterizing tumor heterogeneity based on MR images possessing different contrast and quantitative information. This requires mechanisms for integrating such data and reducing the data dimensionality to levels amenable to interpretation by human readers. Here we propose a two-step pipeline for integrating diffusion and perfusion MRI that we demonstrate in the quantification of breast lesion heterogeneity. First, the images acquired with the two modalities are aligned using an intermodal registration. Dissimilarity-based clustering is then performed exploiting the information coming from both modalities. To this end an ad hoc distance metric is developed and tested for tuning the weighting for the two modalities. The distributions of the diffusion parameter values in subregions identified by the algorithm are extracted and compared through nonparametric testing for posterior evaluation of the tissue heterogeneity. Results show that the joint exploitation of the information brought by DCE and DWI leads to consistent results accounting for both perfusion and microstructural information yielding a greater refinement of the segmentation than the separate processing of the two modalities, consistent with that drawn manually by a radiologist with access to the same data.
|
Informative and Reliable Tract Segmentation for Preoperative Planning
Identifying white matter (WM) tracts to locate eloquent areas for preoperative surgical planning is a challenging task. Manual WM tract annotations are often used but they are time-consuming, suffer from inter- and intra-rater variability, and noise...
more
|
Full text:
|
|
|
Link for citation:
|
|
|
Identifying white matter (WM) tracts to locate eloquent areas for preoperative surgical planning is a challenging task. Manual WM tract annotations are often used but they are time-consuming, suffer from inter- and intra-rater variability, and noise intrinsic to diffusion MRI may make manual interpretation difficult. As a result, in clinical practice direct electrical stimulation is necessary to precisely locate WM tracts during surgery. A measure of WM tract segmentation unreliability could be important to guide surgical planning and operations. In this study, we use deep learning to perform reliable tract segmentation in combination with uncertainty quantification to measure segmentation unreliability. We use a 3D U-Net to segment white matter tracts. We then estimate model and data uncertainty using test time dropout and test time augmentation, respectively. We use a volume-based calibration approach to compute representative predicted probabilities from the estimated uncertainties. In our findings, we obtain a Dice of ≈0.82 which is comparable to the state-of-the-art for multi-label segmentation and Hausdorff distance <10mm. We demonstrate a high positive correlation between volume variance and segmentation errors, which indicates a good measure of reliability for tract segmentation ad uncertainty estimation. Finally, we show that calibrated predicted volumes are more likely to encompass the ground truth segmentation volume than uncalibrated predicted volumes. This study is a step toward more informed and reliable WM tract segmentation for clinical decision-making.
|
