7.0 Tesla MRI in fetal autopsy: a preliminary exploration of integrating image optimization with clinical practice in...

Fetal and neonatal autopsies are crucial for determining causes of death and guiding future pregnancies; however, due to cultural and personal factors, their clinical application rates are decreasing. Noninvasive postmortem magnetic resonance imaging (PMMRI) has emerged as an alternative to fetal and neonatal autopsies. Ultra-high-field 7.0 Tesla (7.0T) PMMRI provides near-microscopic spatial resolution, superior tissue contrast, and a high signal-to-noise ratio compared to lower-field magnetic resonance imaging (MRI). However, its broader clinical application is challenged by stringent specimen preservation requirements, complex scanning parameters, and a lack of standardized multidisciplinary workflows. Thus, this study aimed to establish a multidisciplinary protocol for 7.0T PMMRI in fetuses, optimize the technical parameters based on practical experience, and validate its diagnostic consistency with pathological autopsy findings.

This single-center workflow-optimization study included 20 fetuses delivered by induced labor (mean gestational age 26.4 weeks). Of the 20 fetuses, 19 underwent final pathological autopsy, and 13 had previously undergone in vivo 1.5T MRI evaluations. A Plan-Do-Check-Act (PDCA) cycle was systematically applied to refine cadaver preparation, imaging sequences, and multidisciplinary collaboration. Diagnostic consistency among in vivo 1.5T MRI, postmortem 7.0T PMMRI, and pathological autopsy was statistically analyzed using Cohen's kappa coefficient.

The 7.0T PMMRI workflow was optimized across five key aspects: body selection and preparation, 7.0T MRI setup, personnel readiness, data processing, and multidisciplinary death review. To ensure image quality and minimize issues related to freezing, compression, and autolysis, strict postmortem pretreatments were established, including that imaging should be conducted within 24 hours at 4 ℃. Scanning protocols were optimized to a duration of 60-120 minutes through the selection of appropriate magnetic resonance (MR) coils and sequences [e.g., T2-weighted imaging spin-echo (T2WI-SE) and susceptibility-weighted imaging (SWI)]. The consistency analysis revealed excellent diagnostic agreement between 7.0T PMMRI and pathological autopsy (Cohen's kappa =0.872, P<0.001), a value significantly higher than that between in vivo 1.5T MRI and autopsy (Cohen's kappa =0.22).

Integrating 7.0T PMMRI into a structured multidisciplinary workflow via the PDCA cycle effectively improved standardization, operational efficiency, and image quality. High-resolution 7.0T PMMRI exhibited high diagnostic consistency with traditional autopsy, supporting its role as a valuable and reliable noninvasive alternative for fetal postmortem examinations.