Postmortem Magnetic Resonance Imaging of In Utero Fetuses and Its Relationship With Postmortem Interval: A Multi-Organ Observational Study on Reduced Fetuses of Complicated Multiple Pregnancies.

BACKGROUND: Despite technological advancements in perinatal imaging, autopsy examination is still regarded as the reference standard to determine the time and reason of the fetal death. PURPOSE: This study was conducted to identify the intrauterine postmortem magnetic resonance imaging (PMMR) findings of fetuses, who underwent radiofrequency ablation (RFA). STUDY TYPE: Prospective. POPULATION: Fifty-three twin/triplet complicated pregnancies scheduled for selective reduction of one of the fetuses by RFA were included. FIELD STRENGTH/SEQUENCE: The imaging methods used are T1-weighted gradient-echo imaging (T1 GRE), T2 half-Fourier acquisition single-shot turbo spin-echo (SSTSE), and diffusion-weighted imaging (DWI) sequences. ASSESSMENT: The MRIs were interpreted by three radiologists. STATISTICAL TESTS: Data were analyzed using the software package SPSS Statistics Version 22.0. The used tests included one-way analysis of variance (ANOVA) and Duncan tests (significance level: P value <0.05). This analysis was performed with scikit-learn library (version 1.1.1) in Python version 3.9. RESULTS: Average PMMR scores of orbit, brain, and abdomen showed significant differences among different PM interval subgroups. The brain apparent diffusion coefficient (ADC) numbers of reduced and living fetuses were significantly different at any PM interval. To determine which findings are closely associated with the timing of fetal death, five different methods of feature selection were employed. The top eight selected features achieved the highest area under the curve (AUC) of 78.19%. DATA CONCLUSION: In utero, PMMR findings may be associated with the time of fetal death. Among different fetal organs evaluated, particularly PMMR top eight features specifically scores of orbits were associated with PM intrauterine time after death. LEVEL OF EVIDENCE: 2 TECHNICAL EFFICACY: Stage 2.

A Bayesian Dynamical Approach for Human Action Recognition.

We introduce a generative Bayesian switching dynamical model for action recognition in 3D skeletal data. Our model encodes highly correlated skeletal data into a few sets of low-dimensional switching temporal processes and from there decodes to the motion data and their associated action labels. We parameterize these temporal processes with regard to a switching deep autoregressive prior to accommodate both multimodal and higher-order nonlinear inter-dependencies. This results in a dynamical deep generative latent model that parses meaningful intrinsic states in skeletal dynamics and enables action recognition. These sequences of states provide visual and quantitative interpretations about motion primitives that gave rise to each action class, which have not been explored previously. In contrast to previous works, which often overlook temporal dynamics, our method explicitly model temporal transitions and is generative. Our experiments on two large-scale 3D skeletal datasets substantiate the superior performance of our model in comparison with the state-of-the-art methods. Specifically, our method achieved 6.3% higher action classification accuracy (by incorporating a dynamical generative framework), and 3.5% better predictive error (by employing a nonlinear second-order dynamical transition model) when compared with the best-performing competitors.