RESUMO
Despite the extensive literature on missing data theory and cautionary articles emphasizing the importance of realistic analysis for healthcare data, a critical gap persists in incorporating domain knowledge into the missing data methods. In this paper, we argue that the remedy is to identify the key scenarios that lead to data missingness and investigate their theoretical implications. Based on this proposal, we first introduce an analysis framework where we investigate how different observation agents, such as physicians, influence the data availability and then scrutinize each scenario with respect to the steps in the missing data analysis. We apply this framework to the case study of observational data in healthcare facilities. We identify ten fundamental missingness scenarios and show how they influence the identification step for missing data graphical models, inverse probability weighting estimation, and exponential tilting sensitivity analysis. To emphasize how domain-informed analysis can improve method reliability, we conduct simulation studies under the influence of various missingness scenarios. We compare the results of three common methods in medical data analysis: complete-case analysis, Missforest imputation, and inverse probability weighting estimation. The experiments are conducted for two objectives: variable mean estimation and classification accuracy. We advocate for our analysis approach as a reference for the observational health data analysis. Beyond that, we also posit that the proposed analysis framework is applicable to other medical domains.
RESUMO
PURPOSE: To develop and validate a new cardiac self-gating algorithm using blind source separation for 2D cine steady-state free precession (SSFP) imaging. METHODS: A standard cine SSFP sequence was modified so that the center point of k-space was sampled with each excitation. The center points of k-space were processed by 4 blind source separation methods, and used to detect heartbeats and assign k-space data to appropriate time points in the cardiac cycle. The proposed self-gating technique was prospectively validated in 8 patients against the standard electrocardiogram (ECG)-gating method by comparing the cardiac cycle lengths, image quality metrics, and ventricular volume measurements. RESULTS: There was close agreement between the cardiac cycle length using the ECG- and self-gating methods (bias 0.0 bpm, 95% limits of agreement ±2.1 bpm). The image quality metrics were not significantly different between the ECG- and self-gated images. The ventricular volumes, stroke volumes, and mass measured from self-gated images were all comparable with those from ECG-gated images (all biases <5%). CONCLUSION: The self-gating method yielded comparable cardiac cycle length, image quality, and ventricular measurements compared with standard ECG-gated cine imaging. It may simplify patient preparation, be more robust when there is arrhythmia, and allow cardiac gating at higher field strengths.