Pediatrics
META ANALYSIS
● Meta-analysis
Post-mortem MRI shows 96% specificity, 76% sensitivity for cardiac anomalies in fetal/perinatal deaths
Journal of perinatal medicine
Published March 30, 2026
Palo Seetu, Mangla Mishu, Mishra Monica, Kanikaram Poojitha Kalyani, Srirambhatla Annapurna
PubMed ↗
DOI ↗
This systematic review and meta-analysis evaluated the diagnostic accuracy of post-mortem magnetic resonance imaging (pmMRI) for detecting cardiac anomalies in fetal and perinatal deaths, comparing findings to conventional autopsy. Following PRISMA 2020 guidelines, researchers conducted a systematic search of electronic databases and grey literature. Sixteen studies were included in the systematic review, with 12 studies (n=1810 fetuses) included in the meta-analysis. Quality assessment was performed using the QUADAS-2 tool, and meta-analysis used a bivariate random-effects model. Results showed pooled sensitivity of 76% (95% CI: 71-80%) and pooled specificity of 96% (95% CI: 95-97%) for pmMRI in detecting cardiac anomalies. The diagnostic odds ratio was 55.35 (95% CI: 22.73-134.79), and the area under the summary receiver operating characteristic (SROC) curve was 0.89, indicating excellent diagnostic performance. Subgroup analyses revealed comparable sensitivity between 1.5 and 3 Tesla scanners, though specificity was slightly higher for 1.5 Tesla. Diagnostic accuracy was generally better in larger fetuses and when higher field strength MRI (9.4 Tesla) was used. The authors conclude that pmMRI demonstrates high specificity and moderate sensitivity for detecting cardiac anomalies in fetal and perinatal deaths and may serve as a valuable non-invasive adjunct to conventional autopsy. They recommend standardization of imaging protocols and further research into high-field pmMRI integration to optimize diagnostic reliability.
When a baby dies before or shortly after birth, parents are often faced with a heartbreaking choice: consent to a full autopsy to find out why, or decline and live with unanswered questions. A new review of existing research looked at whether a special MRI scan performed after death could be a less invasive alternative for finding heart problems. The analysis combined data from 12 studies involving over 1,800 babies. The results show this post-mortem MRI is very good at correctly identifying babies who do not have a heart problem, with 96% accuracy. It's moderately good at finding those who do have a problem, catching about 76% of cases. This means it's a reliable tool for ruling out heart issues, but it might miss some. The scan worked slightly better for larger babies and when using more powerful MRI machines. The researchers conclude that while this scan isn't perfect, it could be a valuable, less invasive option to offer grieving families alongside or instead of a traditional autopsy, especially when looking for heart defects. They call for more standardized procedures to make it more reliable.
What this means for you: A special MRI scan after a baby's death shows high accuracy for detecting heart problems, offering a less invasive option for grieving families.
View Original Abstract ↓
OBJECTIVES: Post-mortem magnetic resonance imaging (pmMRI) has emerged as a promising, non-invasive alternative to conventional autopsy for detecting cardiac anomalies in fetal and perinatal deaths and we aimed to systematically evaluate the diagnostic accuracy of pmMRI in detecting cardiac anomalies in fetal and perinatal deaths.
METHODS: A systematic search of electronic databases and grey literature was conducted following PRISMA 2020 guidelines to analyse studies comparing pmMRI findings to conventional autopsy. Data extraction and quality assessment were independently performed by two reviewers using the QUADAS-2 tool. Meta-analysis was conducted using a bivariate random-effects model to calculate pooled sensitivity, specificity, and diagnostic odds ratios (DOR).
RESULTS: Sixteen studies were included in the systematic review, and 12 studies (n=1810 fetuses) were meta-analyzed. The pooled sensitivity and specificity of pmMRI for detecting cardiac anomalies were 76 % (95 % CI: 71-80 %) and 96 % (95 % CI: 95-97 %), respectively. The diagnostic odds ratio was 55.35 (95 % CI: 22.73-134.79), with an area under the SROC curve of 0.89, indicating excellent diagnostic performance. Subgroup analyses showed comparable sensitivity between 1.5 and 3 T scanners, though specificity was slightly higher for 1.5 T. Diagnostic accuracy was generally better in larger fetuses and when higher field strength MRI (9.4 T) was used.
CONCLUSIONS: pmMRI demonstrates high specificity and moderate sensitivity for detecting cardiac anomalies in fetal and perinatal deaths and may serve as a valuable non-invasive adjunct to conventional autopsy. Standardization of imaging protocols and further research into high-field pmMRI integration are recommended to optimize diagnostic reliability.