Skip to main content
Fig. 3 | BMC Medical Imaging

Fig. 3

From: Dimensionality reduction-based fusion approaches for imaging and non-imaging biomedical data: concepts, workflow, and use-cases

Fig. 3

Sample predictive heatmaps for detection of prostate cancer in vivo through combining MRI and MRS data. a shows a T2w MRI section with the MRS grid overlaid in white. The expert annotation of cancer presence is also shown with a red outline around those voxels that were assessed as cancerous. Corresponding automated classification results are shown for using: b T2w MRI texture features alone, c MRS peak area metabolite ratios, d DFS-ES, e DFS-EC, f DFS-KC. These are visualized in the form of heatmaps, where red corresponds to higher probability of CaP presence. The expert annotation of CaP presence is also superposed via a red outline in each image

Back to article page