Search

To investigate the utility of T1ρ and T2 relaxations for assessing the severity of liver fibrosis (F stage) and necro-inflammation (A stage) in patients with chronic liver disease (CLD).We calculated T1ρ and T2 relaxations of the liver parenchyma in 82 patients who underwent liver surgery. F and A stages of enrolled patients were assessed by referring to surgically resected specimens. The relationships between T1ρ or T2 relaxation and F or A stage were assessed using one-way analysis of variance followed by Tukey's multiple comparison test, Spearman's rank correlation test and a receiver operating characteristic analysis.The T1ρ and T2 values of the liver parenchyma were significantly increased as the F and A stages progressed. The T1ρ and T2 values showed significant differences between F0 and F4, between F1 and F4, and between F2 and F4. In addition, T1ρ values showed a significant difference between F0 and F3 as well. The highest diagnostic ability for fibrosis was obtained when differentiating ≥F3 from ≤F2 using T1ρ: the sensitivity was 82.8%, the specificity 79.2% and the area under the curve (AUC) 0.87. The sensitivity and AUC of T1ρ relaxation (46.9% and 0.67) were significantly higher than those of T2 relaxation (29.7% and 0.60) for differentiating ≥A1 from A0.T1ρ and T2 relaxations have potential as a biochemical marker for assessing the severity of liver fibrosis and necro-inflammation. T1ρ relaxation may be slightly superior to T2 relaxation in terms of diagnostic ability for liver fibrosis and necro-inflammation.

Purpose Hierarchical clustering (HC), an unsupervised machine learning (ML) technique, was applied to multi-parametric MR (mp-MR) for prostate cancer (PCa). The aim of this study is to demonstrate HC can diagnose PCa in a straightforward interpretable way, in contrast to deep learning (DL) techniques. Methods HC was constructed using mp-MR including intravoxel incoherent motion, diffusion kurtosis imaging, and dynamic contrast-enhanced MRI from 40 tumor and normal tissues in peripheral zone (PZ) and 23 tumor and normal tissues in transition zone (TZ). HC model was optimized by assessing the combinations of several dissimilarity and linkage methods. Goodness of HC model was validated by internal methods. Results Accuracy for differentiating tumor and normal tissue by optimal HC model was 96.3% in PZ and 97.8% in TZ, comparable to current clinical standards. Relationship between input (DWI and permeability parameters) and output (tumor and normal tissue cluster) was shown by heat maps, consistent with literature. Conclusion HC can accurately differentiate PCa and normal tissue, comparable to state-of-the-art diffusion based parameters. Contrary to DL techniques, HC is an operator-independent ML technique producing results that can be interpreted such that the results can be knowledgeably judged.