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

Introduction

Despite significant advances in prenatal screening and care in recent years, the causes of some fetal deaths remain unclear. Fetal and neonatal autopsies play a crucial role in the development of perinatal medicine by assisting clinicians and radiologists to diagnose fetal diseases, determine causes of death, extend understanding of pathophysiologic mechanisms, and evaluate interventions and recurrence risk in subsequent pregnancies. Despite these benefits, the rate of fetal and neonatal pathologic autopsy continues to decrease worldwide due to various factors such as religious, cultural, and parental beliefs (1). Thus, a noninvasive technique with higher anatomical resolution urgently needs to be developed.

With advances in imaging, various virtual autopsy items have been developed, including postmortem conventional radiography, postmortem computed tomography (PMCT), postmortem magnetic resonance imaging (PMMRI), postmortem ultrasound, and minimally invasive autopsy (2,3). These imaging techniques preserve the integrity of the cadaver and have high diagnostic value. As a result, the use of fetal imaging autopsy has gradually increased due to greater family acceptance and growing clinical demand (4).

Unlike PMCT, which requires prolonged immersion in iodine-based contrast solutions to enhance soft-tissue contrast for radiological imaging, PMMRI provides superior intrinsic tissue contrast, facilitates non-invasive three-dimensional (3D) imaging, and prevents potential data loss due to inadequate staining (5). Since 1990 (6), PMMRI has been widely used in imaging autopsy. In some cases, it has also served as an alternative to conventional brain autopsy (7), demonstrating good consistency with traditional autopsy findings (8,9), strong correlation with neuropathological results, and the ability to avoid brain tissue atrophy and the damage associated with conventional brain autopsy (10).

High-field PMMRI also addresses technical limitations of conventional postmortem imaging (e.g., low spatial resolution and poor tissue contrast) and provides a new perspective for image autopsy. First, there has been a qualitative leap in spatial resolution, with isotropic resolution reaching 100 µm (11), approximately 15-fold higher than that of clinical 1.5 Tesla (1.5T) magnetic resonance imaging (MRI). Second, there has also been a significant improvement in tissue contrast, with the T2*-weighted imaging signal-to-noise ratio (SNR) reaching 216±34, significantly higher than that of routine autopsy tissue sections (SNR = 89±17, P<0.001) (12). A recent study reported that postmortem 7.0T MRI was more than twice as likely as 3.0T MRI to detect cortical lesions in patients with multiple sclerosis (13). A postmortem fetal study found that at all stages of layered brain development, 3.0T MRI was unable to observe the boundary between the subventricular and intermediate zones; however, 7.0T MRI was able to visualize this boundary in the occipital part of the brain at 16 and 20 weeks of gestation (14). Similarly, a virtual postmortem examination of traumatic brain injury demonstrated that 7.0T MRI provides higher resolution than 3.0T MRI, and has the ability to detect small lesions (15).

Because 7.0T MRI provides excellent SNR, spatial resolution, and tissue contrast, it is being widely applied in both nervous system imaging (16) and musculoskeletal system imaging (17). The comparison of postmortem MRI findings with gross and microscopic neuropathological data has a long history, Becher et al. (18) provided key insights into prenatal etiopathogenesis using earlier MRI technology. Building on this established foundation, high-field PMMRI can further address previous technical limitations. While earlier generations of MRI paved the way for virtual autopsy, 7.0T MRI represents a qualitative leap in spatial resolution, approaching the level of microscopic evaluation. Some studies have shown that the current MRI technology enables the investigation of the functional organization of the cerebral cortex at a submillimeter scale (19). The spatial resolution of MRI is steadily approaching that of microscopy. A recent study achieved a breakthrough in the linear and 3D measurement of fetal eyeballs using 7.0T PMMRI (20). Ultra-high-field (7.0T or higher) MRI technology improves the magnetic field strength and gradient performance, enabling micron-scale resolution at a high SNR, resulting in fetal image autopsy comparable to pathologic autopsy (21).

However, despite the clear benefits of 7.0T PMMRI in terms of spatial resolution, tissue contrast, and SNR over lower-field MRI, its application in fetal and neonatal autopsies still faces several challenges. First, as the transition of ultra-high-field 7.0T MRI from scientific research to broader clinical use is relatively recent, the number of reported cases and published studies on its postmortem application remains limited, and standardized protocols are lacking. In addition, the ultra-high-field environment imposes more stringent requirements on scanning parameters, gradient systems, and specimen preservation, thereby increasing its technical complexity. Further, the absence of uniform workflows across centers limits the reproducibility and comparability of results. Although 7.0T PMMRI has achieved near-microscopic resolution (22), the pathway for systematic integration with conventional pathology and multidisciplinary workflows has yet to be fully established. This study summarized and analyzed the potential factors involved in controlling each step of fetal and neonatal PMMRI based on existing literature and practical experience from 7.0T PMMRI scanning in 13 fetuses. The aim was to provide a more usable project management scheme for fetal and neonatal PMMRI that may be adopted by other institutions in the future.

Methods

Study design and ethical approval

This single-center workflow-optimization study sought to summarize technical and organizational refinements for 7.0T PMMRI in fetal and neonatal cases. The clinical research and clinical project implementation were performed in accordance with the Declaration of Helsinki and its subsequent amendments, as well as relevant Chinese clinical research norms and regulations. This study was approved by the Ethics Committee of The First Affiliated Hospital of Army Medical University [No. (A) KY2023036]; when routine autopsy was required, standardized fee certification and registration were completed as per Health Commission requirements. Informed consent was obtained from the patients’ parents or legal guardians following structured counseling by an obstetrician.

In vivo 1.5T MRI acquisition

To establish a clinical baseline for comparison with postmortem findings, 13 of the 20 included cases underwent conventional in vivo 1.5T MRI (Magnetom Aera, Siemens Healthineers, Erlangen, Germany) prior to labor induction, as part of their routine clinical evaluation. These in vivo scans were performed using standard fetal imaging protocols, including T1-weighted and T2-weighted sequences acquired in multiple planes (the scanning parameters are provided in Table S1). The results of these prenatal scans were subsequently used to evaluate the incremental diagnostic value of 7.0T PMMRI compared with standard clinical imaging.

Cadaver preparation and 7.0T PMMRI

Pilot scans were initially performed using published fetal preparation protocols and MRI acquisition parameters as baseline references. Building on these preliminary trials, we applied a structured Plan-Do-Check-Act (PDCA) cycle, integrating systematic image quality control metrics—including the SNR, contrast-to-noise ratio, and artifact scoring—together with iterative multidisciplinary case discussions. Through this process of continuous refinement, preparation strategies and acquisition protocols were optimized to balance spatial resolution with scan duration, ultimately resulting in a best-practice protocol for fetal preparation and imaging parameters.

The fetus was sealed in a clean polyethylene bag and then positioned directly within a 32-channel skull coil. This setup not only prevented contamination but also helped maintain the initial 4 ℃ baseline temperature during the early stages of the scan. To mitigate the effects of continuous radiofrequency (RF) heating, the scanning protocol was strictly limited to a duration of 60 to 120 minutes. Additionally, temperature-sensitive sequences [e.g., diffusion-weighted imaging (DWI) and diffusion tensor imaging (DTI)] were prioritized early in the scanning sequence, while the baseline temperature of the cadaver (initially maintained at 4 ℃) remained relatively stable.

Image evaluation framework

The interpretation of MRI images was performed by two experienced radiologists, each with more than 10 years of experience in fetal in vivo MRI, and at least 3 years of experience in high-field MRI interpretation. The two radiologists independently reviewed all the cases using a structured checklist, and any differences were resolved through consensus. The pathological autopsy was performed by a perinatal pathologist with 15 years of clinical experience. Each case included the evaluation of 37 items across seven common anatomical systems in routine autopsy, including the nervous system, maxillofacial system, lungs, heart and great vessels, digestive and urinary systems, and musculoskeletal system. The evaluated anatomical structures are listed in Table S2. Discrepancies between prenatal imaging and postmortem findings were further investigated through a retrospective, non-blinded review by two senior radiologists to determine if the lesions were retrospectively visible or represented true diagnostic limitations of the 1.5T system.

Statistical analysis and quality improvement framework

When available, conventional autopsy served as the reference standard for the consistency analysis and clinical decision-making. The primary research outcomes included (I) optimization of image quality, and (II) diagnostic concordance with autopsy findings. The consistency among in vivo 1.5T MRI, 7.0T PMMRI, and pathological autopsy was assessed using Cohen’s κ coefficient. Missing data (e.g., absent prenatal 1.5T scans or a lack of autopsy results for specific cases) were handled using case-wise deletion for each correlation analysis, ensuring that only cases with complete paired data were included in the respective agreement calculations. Case 6, which lacked a matching reference standard (pathological autopsy), was excluded from the agreement analysis to ensure the accuracy of Cohen’s kappa. P values <0.05 were considered statistically significant. Through systematic review, organization, and multidisciplinary discussion of the entire 7.0T PMMRI workflow, a comprehensive protocol for scan optimization was developed.

Results

Summary of fetal cadaver preparation experience

Care was taken to avoid compression of the fetuses during labor induction, as deformation may destroy the fetal cadaver shape (Figure 1A). As fetuses exposed to formalin may exhibit autolysis and immersion-related changes (Figure 1B), 7.0T PMMRI was performed at the earliest possible time within 24 hours without formalin fixation or any other agents. The cadavers were stored and transported in a cold chamber at 4 ℃, as freezing produces a large number of ice crystals and damages the brain structure (Figure 1C,1D). Before removal from the storage chamber, the fetal surface was cleaned to remove blood clots and tissue fluid adherent to the fetal surface, as these can interfere with the uniformity of the MRI magnetic field and imaging quality.

Figure 1 Examples of fetuses not suitable for 7.0T PMMRI. (A,B) Images of a fetus at 17⁺² weeks of gestation following induction of labor due to maternal cervical cancer. (A) shows extrusion, deformation, and collapse of the head; and (B) the red arrows indicate severe tissue dissolution. (C,D) 7.0T PMMRI images of a fetus at 30⁺⁴ weeks of gestation following induction of labor. (C) The red arrow indicates incomplete freezing of the head; (D) the red arrow indicates brain damage after thawing. 7.0T, 7.0 Tesla; PMMRI, postmortem magnetic resonance imaging.

Scan parameter optimization

Extensive exploration of the nervous system images (Figure 2) and body part images (Figure 3) was conducted as part of the optimization process to establish an empirically derived optimized imaging sequence (Table 1).

Figure 2 7.0T PMMRI images of the nervous system of a fetus at 28+4 weeks of gestation following induction of labor. (A) Axial T2WI and (B) sagittal T2WI images of the nervous system obtained using the T2WI-SE sequence demonstrated better delineation of the cranial structure and organization. (C) ADC map of the skull. (D) Cranial SWI (small-vein display). (E) T1-MP2RAGE coronal imaging in which T1WI showed limited fetal anatomical detail. 7.0T, 7.0 Tesla; ADC, apparent diffusion coefficient; MP2RAGE, magnetization prepared 2 rapid acquisition gradient echoes; PMMRI, postmortem magnetic resonance imaging; SWI, susceptibility-weighted imaging; T1WI, T1-weighted imaging; T2WI, T2-weighted imaging; T2WI-SE, T2-weighted imaging spin-echo.

Figure 3 7.0T PMMRI images of the thoracoabdominal system of (A) a fetus at 34⁺³ weeks of gestation and (B-F) a fetus at 27⁺³ weeks of gestation following induction of labor. (A) Coronal T2WI showing clear visualization of the spine, joints, liver, kidney, and intestine. (B) Coronal T2WI showing bile duct dilatation. (A,B: T2WI-SE sequences showing the anatomical structure in the body scan). (C) SWI imaging of the body showing the skeletal structure of the spine. (D) Abdominal True-FISP 3D image showing more pronounced magnetic susceptibility artifacts. (E) Abdominal T2 SPACE image with poor T2 contrast; and (F) Abdominal 3D-DESS image. 3D-DESS, three-dimensional double-echo steady-state; 7.0T, 7.0 Tesla; PMMRI, postmortem magnetic resonance imaging; SPACE, sampling perfection with application-optimized contrasts using different flip-angle evolutions; SWI, susceptibility-weighted imaging; T2WI, T2-weighted imaging; T2WI-SE, T2-weighted imaging spin-echo; True-FISP, true fast imaging with steady-state precession.

Consistency analysis between in vivo 1.5T MRI, 7.0T PMMRI, and pathological autopsy

A total of 20 fetuses delivered by induced labor were examined by 7.0T PMMRI, with a mean gestational age of 26.4 weeks (range, 17–34 weeks). The indications for termination included structural anomalies (n=15), suspected genetic or chromosomal syndromes (n=2), and maternal-fetal complications (n=2). Of the fetuses, 19 underwent final pathological autopsy; one case (case 6) lacked pathological autopsy results due to parental refusal of pathological autopsy, with consent limited to imaging autopsy; and 13 had undergone 1.5T MRI before labor induction (Table S3). Figure 4 provides a representative comparison between in vivo 1.5T MRI and 7.0T PMMRI in one case.

Figure 4 Comparison of in vivo 1.5T MRI and 7.0T PMMRI images of Case 11. At 27 weeks of gestation, 1.5T MRI was performed to detect microphthalmia. Genetic testing revealed PTCH1 gene mutations in both the father and the fetus, after which the parents decided to induce labor. (A,C) The 1.5T MRI images of the fetus; (B,D) the corresponding 7.0T PMMRI images. (A) shows that the right eye of the fetus was small (indicated by the white arrow), while the left eye showed no definite abnormality. (B) also shows that the left retina was detached (indicated by the long black arrow). (C) Sagittal 1.5T MRI of the spine shows no definite abnormality; (D) 7.0T PMMRI shows the cervical half vertebra of the fetus (indicated by the white box). MRI, magnetic resonance imaging; PMMRI, postmortem magnetic resonance imaging; T, Tesla.

The in vivo 1.5T MRI findings were compared with those of the 7.0T PMMRI and pathological autopsy. The Cohen’s kappa coefficients for agreement between 1.5T MRI and 7.0T PMMRI, and between 1.5T MRI and pathological autopsy were 0.22 (P=0.043) and 0.22 (P=0.041), respectively. The Cohen’s kappa coefficient for agreement between 7.0T PMMRI and pathological autopsy reached 0.872 (95% confidence interval: 0.706–1.000) (P<0.001). The consistency results are set out in Table 2.

Optimized PDCA cycle framework development

Based on our study results and practical experience, the entire process of the 7.0T PMMRI PDCA cycle was optimized. This included early personnel training for the project, family informed consent, body selection and handling, transfer and transport, 7.0T MRI scanning, pathological autopsy (as necessary), data processing, and multidisciplinary discussion, thereby ensuring clearly defined functions for each step and close collaboration among members (Figure 5).

Figure 5 Diagram of 7.0T PMMRI project process management. 7.0T, 7.0 Tesla; MRI, magnetic resonance imaging; PDCA, plan-do-check-act; PMMRI, postmortem magnetic resonance imaging.

Discussion

This study presents a practical workflow for 7.0T PMMRI, highlighting the importance of early imaging within 24 hours after death, avoidance of freezing, preservation at 4 ℃, and the use of high-channel coils. By balancing spatial resolution with operational feasibility, an optimal scan duration of 60–120 minutes was established. Our preliminary results in this small exploratory cohort suggest that the optimized 7.0T PMMRI workflow facilitates the acquisition of high-resolution images and shows encouraging diagnostic agreement with conventional autopsy findings. Further, the proposed project management and implementation framework provides actionable guidance to improve image quality and diagnostic consistency. Collectively, these advances support the clinical translation of 7.0T PMMRI and underscore its potential for wider adoption in fetal and neonatal autopsy practice.

Based on our practical experience, we identified key issues in the 7.0T PMMRI scanning process and standardized the scanning process to ensure its smooth application in clinical practice. Our main findings are summarized below.

Fetal cadaver preparation

Labor induction and preservation methods significantly affect 7.0T PMMRI examination results. Rapid autolysis of brain tissue is common in fetal autopsy specimens, while the prolonged preservation of stillbirth specimens often results in varying degrees of tissue degradation, posing substantial challenges for both research and diagnostic interpretation. Based on our practical experience, fetuses younger than 14 weeks of gestation are generally not suitable for this workflow due to their small organ volumes and high susceptibility to autolysis, which makes proper positioning and diagnostic interpretation extremely challenging. To ensure optimal imaging outcomes and avoid the unnecessary use of medical resources, strict cadaver selection criteria are required. Three key considerations are particularly important: (I) cryopreservation should be avoided, as ice crystal formation can severely disrupt brain microstructure; (II) excessive compression during labor induction may lead to significant deformation of the fetal body, thereby limiting its diagnostic value for imaging assessment; and (III) obstetricians should apply a postmortem immersion scoring system (23) (grades 1–4, representing none, mild, moderate, or severe immersion, respectively) to evaluate suitability for imaging. Specimens demonstrating marked autolysis, a postmortem interval greater than 24 hours, prolonged intrauterine retention, or preterm abortion should be excluded from PMMRI evaluation.

Traditional brain autopsy also has inherent limitations. Because brain tissue is predominantly water, it tends to collapse once removed from the cerebrospinal fluid and cranial vault, thereby compromising anatomical integrity and producing morphological distortion (24). Previous studies have reported that approximately 20% of brain autopsies have limited diagnostic value due to autolysis (25). Although formalin fixation can slow tissue decomposition, it also alters microstructural architecture and water diffusion properties (26), leading to reduced T1- and T2-weighted signal intensity and diminished gray-white matter contrast. For these reasons, 7.0T PMMRI should ideally be performed within 24 hours after death, avoiding the use of fixatives such as formalin. If timely imaging is not possible within this window, alternative fixation strategies should be considered to preserve diagnostic value.

During preparation, the fetal surface should be thoroughly cleaned before imaging to minimize the effects of residual blood clots or tissue fluid on magnetic field homogeneity and image quality. A non-magnetic safety check is essential to confirm the absence of ferromagnetic materials or foreign objects. For storage and transport, the fetus should be sealed in a polyethylene bag, placed in a low-temperature transport container, and consistently maintained in a 4 ℃ refrigeration unit (7). Comprehensive documentation, detailing basic demographic information, the induction time, the transport method, and the personnel involved, should also be completed.

Based on practical experience, this study confirmed that early imaging within 24 hours after death, avoidance of freezing, and preservation at 4 ℃ represent the optimal conditions for maintaining image quality in 7.0T PMMRI.

7.0T MRI preparation

7.0T MRI is used in both clinical and scientific research. Given the time particularity of 7.0T PMMRI, the examination time should be pre-estimated, scanning should be scheduled accordingly, and a special cadaver transport channel should be established. Additionally, sufficient post-examination disinfection time is required before the device is switched back to clinical mode.

After each PMMRI examination, the MRI table and coils should be disinfected according to standardized procedures: the surfaces should be disinfected using quaternary ammonium salt wipes or 70% isopropyl alcohol wipes, and the examination room should then be disinfected with sufficient ultraviolet light exposure for 30 minutes (in an unoccupied state), and continuous indoor disinfection should be performed using a non-magnetic air-circulation fan disinfection machine.

In 7.0T PMMRI, maximum-fit channels such as the 32-channel skull coil should be selected, and a segmented scan method for the head and body should be adopted. The high RF energy of 7.0T MRI imposes strict specific absorption rate (SAR) limits for living participants. However, for cadavers, SAR limits are meaningless and may hamper image acquisition. Thus, RF and gradient modes should be adjusted to maximize the SAR limit range.

Previous fetal and neonatal 7.0T PMMRI studies often required several to tens of hours to collect MRI sequences for ultra-high imaging data (27). However, such extended scanning durations, often with multiple imaging doctors, are unsustainable. To facilitate clinical translation, imaging physicians and MRI technicians should work together to design standardized scanning sequences of 60–120 minutes to obtain optimal images within a limited time frame.

7.0T PMMRI scan sequence

When acquiring nervous system sequences, the T2-weighted imaging spin-echo (T2WI-SE) sequence provides a remarkable comparison of brain structure and organization. A spatial resolution of 0.1×0.1×1 mm³ can be achieved within 10 minutes and 12 seconds. T1-weighted imaging (T1WI) provides fetal anatomical detail, due to the shorter echo time (TE) characteristics at 7T, whereas susceptibility-weighted imaging (SWI) provides superior vascular and hemorrhagic contrast. However, while SWI can identify hemorrhagic sites, it is often difficult to determine whether these hemorrhages occurred prenatally or were induced during labor. A previous study (28) on 3.0T PMMRI have reported marked improvements in DWI and DTI, even in the presence of autolysis. DTI may also provide crucial information, including time of death. The apparent diffusion coefficient (ADC) is a measurement index of particular research significance in 7.0T MRI. While SAR constraints may be lifted for cadavers to optimize acquisition, scanning for 60 to 120 minutes may still result in RF-related heating, introducing a temperature gradient from start to finish. While macroscopic temperature artifacts are minimal, this thermal shift inherently affects the quantitative precision of ADC values. Future quantitative diffusion studies at 7.0T should be conducted to account for these temperature-dependent effects.

The anatomy of the body sequence was also displayed using the T2WI-SE sequence with 0.1×0.1×1 mm³ resolution. SWI was used to highlight the fetal bone structure. 3D sequences have high imaging requirements due to the complex structure of the abdomen; however, due to magnetic susceptibility effects at 7.0T, the True-FISP (true fast imaging with steady-state precession) sequence exhibits more susceptibility artifacts. The variable flip-angle fast spin-echo sequence [sampling perfection with application-optimized contrasts using different flip-angle evolutions (SPACE)] was poorly organized at 7.0T. The three-dimensional double-echo steady-state (3D-DESS) sequence is an ideal 3D sequence; however, care should be taken to eliminate artifacts during acquisition.

To mitigate the effects of continuous RF heating associated with bypassing SAR limits, the scanning protocols were strictly limited to 60–120 minutes. Further, temperature-sensitive sequences such as DWI and DTI were prioritized early in the scanning order, while the cadaver’s baseline temperature (initially maintained at 4 ℃) remained relatively stable. Although the SAR restrictions were relaxed in the cadaveric imaging for optimization purposes, the RF heating generated during the 60- to 120-minute scan process may create a temperature gradient from the beginning to the end of the examination. Although the macroscopic temperature artifacts observed in this study were minimal, such thermal changes may affect the quantitative accuracy of the ADC values. Currently, due to the lack of effective real-time temperature monitoring methods, this remains one of the limitations of the study. To further improve the accuracy of quantitative research, it is recommended that medical centers with appropriate resources use MR-compatible temperature probes (e.g., positioned in the fetal axillary region) for continuous real-time monitoring throughout the scan. This will provide key data for calibrating temperature-dependent changes in ADC values and improve the accuracy of quantitative diffusion assessment. Future 7.0T quantitative diffusion studies should take these temperature-related variables into account and explore more rigorous heat monitoring schemes.

Safety training for MRI personnel

Due to superconductivity and magnetic field uniformity requirements, high-field 7.0T MRI systems maintain a constant magnetic field even when not actively scanning. At 3 m from the magnet along the main magnetic field (Z-axis), the field intensity is 40 mT, nearly four times that of the 3.0T MRI system, resulting in a stronger projectile effect on ferromagnetic metals entering the 7.0T magnetic field region, and thereby posing safety risks. Thus, all personnel must complete 7.0T magnetic field safety training and may only work in the magnetic field area under the supervision of MRI technicians.

Multidisciplinary collaboration

Multidisciplinary collaboration is essential for the accurate diagnosis of fetal malformations and for providing informed counseling regarding subsequent pregnancies, while also offering critical guidance for refining the 7.0T PMMRI workflow. Pathologists, obstetricians, neonatologists, and geneticists should each contribute to the refinement of their respective diagnostic reports. Establishing a dedicated fetal and neonatal autopsy task force would further strengthen this approach by enabling regular case-by-case review of causes of death through structured online or in-person meetings. Such collaboration facilitates the development of standardized mortality reports and promotes consistent, high-quality diagnostic practices.

Obstetricians should collect maternal and fetal genetic, hematological, and clinical data, and provide specific examination directions for imaging departments based on PMMRI image targeting requirements for further clinical analysis. Radiologists should be familiar with embryonic and fetal development imaging features and participate in image quality control. MRI technicians should have extensive experience in fetal imaging, understand the technical limitations of MRI, and optimize scan sequences for data processing and quality control. Pathologists should perform detailed postmortem examinations and systematically correlate their findings with imaging results. Imaging autopsy reports should comprehensively describe normal developmental markers, pathophysiologic alterations, and the effects of postmortem decomposition on imaging, while extracting maximal diagnostic information to advance perinatal medicine and guide future pregnancies. To ensure continuous improvement, autopsy and imaging data should be regularly summarized and analyzed in large-scale studies, establishing a positive feedback loop in which clinical data inform scanning protocols, imaging findings enhance clinical understanding, 7.0T PMMRI guides autopsy procedures, and autopsy results in turn validate and refine imaging workflows.

Further, the interpretation of 7.0T PMMRI findings should ideally be integrated into a holistic evaluation of the maternal-placental-fetal triad. Beyond anatomical resolution, 7.0T PMMRI facilitates an integrated scientific approach to understanding complex disease pathways. By providing detailed findings across multiple organ systems, this technology enables the identification of systemic patterns. Such integrated imaging evaluations assist in distinguishing between genetic syndromes and acquired disease pathways. This holistic perspective is essential for advancing fetal-neonatal neurology and offers families a deeper understanding of adverse perinatal outcomes.

Beyond the immediate postmortem period, these interdisciplinary fetal-neonatal neurology collaborations strengthen efforts to offer proactive and reactive interventions across prenatal, neonatal, and childhood periods. Establishing a dedicated task force enables a continuous feedback loop in which imaging findings enhance clinical understanding and guide future medical interventions, thereby advancing the field of perinatal medicine.

PDCA cycle

The PDCA cycle is a management tool for the continuous improvement of an enterprise product or process (29). For teams newly implementing PMMRI, establishing a 7.0T workflow has important exploratory value. In the initial phase, it is essential to define the team composition and clearly delineate the core responsibilities of obstetrics, neonatology, radiology, and pathology. Using a PDCA management model as a framework, each step of the process—including informed consent, cadaver screening and preparation, information transfer, transport, 7.0T scanning, pathological examination (when required), data analysis, and multidisciplinary mortality review—should be systematically evaluated to ensure standardized procedures, clear accountability, and effective collaboration. By continuously identifying gaps between planned and actual outcomes and incorporating lessons learned, teams can progressively reduce cadaver waiting times, enhance imaging quality, improve diagnostic efficiency, and further refine the overall PMMRI workflow.

Specifically, the iterative PDCA cycles led to the optimization of the scan duration (reducing it to a sustainable 60–120 minutes), while maintaining a high SNR and significantly reducing artifacts caused by inadequate specimen cleaning or suboptimal storage conditions. The PDCA cycles also improved inter-departmental efficiency, including a reduction in the cadaver waiting time from death to imaging, through the establishment of the dedicated multidisciplinary task force.

Limitations

This study had several limitations. First, the relatively small sample size and single-center design may limit the robustness and generalizability of the findings. However, the primary objective of the study was to establish a preliminary and optimized process for fetal 7.0T PMMRI. Thus, this represents a preliminary and exploratory study. Future validation should be conducted in larger, multicenter cohorts to confirm its diagnostic accuracy and broader clinical applicability. Second, although most cases underwent premortem 1.5T MRI and all cases underwent 7.0T PMMRI with pathological autopsy, the lack of 3.0T MRI data remains a notable limitation. Future studies should expand the sample size and incorporate 3.0T MRI as a comparator to more comprehensively evaluate PMMRI performance across different field strengths, while further refining the 7.0T PMMRI workflow and enhancing its clinical efficiency.

Conclusions

High-resolution imaging data from 7.0T PMMRI provide valuable insights for the postmortem diagnosis of fetuses following induced labor. The implementation of a PDCA cycle further enhances the standardization and efficiency of the PMMRI process. By integrating evidence from the literature with clinical practice, this study improved process standardization and initially increased the success rate and image quality of 7.0T PMMRI examinations, providing useful exploratory findings for other institutions conducting fetal and neonatal postmortem imaging examinations.

Acknowledgments

The authors thank the parents and family members who provided specimens for their help in this study.

Footnote

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2079/dss

Funding: This study was supported by the 7.0T Magnetic Resonance Translational Medicine Research Center Special Fund (Project No. 424Z2Q31).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-2079/coif). W.C. reports that as an employee of Siemens Healthineers Ltd., he offered technical assistance in MRI sequence development for this study. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The clinical research and clinical project implementation were performed in accordance with the Declaration of Helsinki and its subsequent amendments; and relevant Chinese clinical research norms and regulations. This study was approved by the Ethics Committee of The First Affiliated Hospital of Army Medical University [No. (A) KY2023036]. Before undertaking this project, informed consent was obtained from patients’ parents or legal guardians.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Lewis C, Riddington M, Hill M, Arthurs OJ, Hutchinson JC, Chitty LS, Bevan C, Fisher J, Ward J, Sebire NJ. Availability of less invasive prenatal, perinatal and paediatric autopsy will improve uptake rates: a mixed-methods study with bereaved parents. BJOG 2019;126:745-53. [Crossref] [PubMed]
2. Kang X, Carlin A, Cannie MM, Sanchez TC, Jani JC. Fetal postmortem imaging: an overview of current techniques and future perspectives. Am J Obstet Gynecol 2020;223:493-515. [Crossref] [PubMed]
3. Blokker BM, Weustink AC, Wagensveld IM, von der Thüsen JH, Pezzato A, Dammers R, Bakker J, Renken NS, den Bakker MA, van Kemenade FJ, Krestin GP, Hunink MGM, Oosterhuis JW. Conventional Autopsy versus Minimally Invasive Autopsy with Postmortem MRI, CT, and CT-guided Biopsy: Comparison of Diagnostic Performance. Radiology 2018;289:658-67. [Crossref] [PubMed]
4. Cannie M, Votino C, Moerman P, Vanheste R, Segers V, Van Berkel K, Hanssens M, Kang X, Cos T, Kir M, Balepa L, Divano L, Foulon W, De Mey J, Jani J. Acceptance, reliability and confidence of diagnosis of fetal and neonatal virtuopsy compared with conventional autopsy: a prospective study. Ultrasound Obstet Gynecol 2012;39:659-65. [Crossref] [PubMed]
5. Arthurs OJ, Guy A, Thayyil S, Wade A, Jones R, Norman W, Scott R, Robertson NJ, Jacques TS, Chong WK, Gunny R, Saunders D, Olsen OE, Owens CM, Offiah AC, Chitty LS, Taylor AM, Sebire NJMagnetic Resonance Imaging Autopsy Study. (MaRIAS) Collaborative Group. Comparison of diagnostic performance for perinatal and paediatric post-mortem imaging: CT versus MRI. Eur Radiol 2016;26:2327-36. [Crossref] [PubMed]
6. Ros PR, Li KC, Vo P, Baer H, Staab EV. Preautopsy magnetic resonance imaging: initial experience. Magn Reson Imaging 1990;8:303-8. [Crossref] [PubMed]
7. Izzo G, Talenti G, Falanga G, Moscatelli M, Conte G, Scola E, Doneda C, Parazzini C, Rustico M, Triulzi F, Righini A. Intrauterine fetal MR versus postmortem MR imaging after therapeutic termination of pregnancy: evaluation of the concordance in the detection of brain abnormalities at early gestational stage. Eur Radiol 2019;29:2740-50. [Crossref] [PubMed]
8. Thayyil S, Sebire NJ, Chitty LS, Wade A, Chong W, Olsen O, Gunny RS, Offiah AC, Owens CM, Saunders DE, Scott RJ, Jones R, Norman W, Addison S, Bainbridge A, Cady EB, Vita ED, Robertson NJ, Taylor AMMARIAS collaborative group. Post-mortem MRI versus conventional autopsy in fetuses and children: a prospective validation study. Lancet 2013;382:223-33. [Crossref] [PubMed]
9. Ashwin C, Hutchinson JC, Kang X, Langan D, Jones R, Norman W, Cannie M, Jani J, Sebire NJ, Arthurs OJ. Learning effect on perinatal post-mortem magnetic resonance imaging reporting: single reporter diagnostic accuracy of 200 cases. Prenat Diagn 2017;37:566-74. [Crossref] [PubMed]
10. Lequin MH, Huisman TA. Postmortem MR imaging in the fetal and neonatal period. Magn Reson Imaging Clin N Am 2012;20:129-43. [Crossref] [PubMed]
11. Moeremans M, Avni FE, d'Haene N, Lam NM, Metens T, D'Hondt A. Combined prenatal US and post-mortem fetal MRI: can they replace conventional autopsy for fetal body abnormalities? Eur Radiol 2024;34:632-42. [Crossref] [PubMed]
12. Tang Y, Blümcke I, Su TY, Choi JY, Krishnan B, Murakami H, Alexopoulos AV, Najm IM, Jones SE, Wang ZI. Black Line Sign in Focal Cortical Dysplasia IIB: A 7T MRI and Electroclinicopathologic Study. Neurology 2022;99:e616-26. [Crossref] [PubMed]
13. Kilsdonk ID, Jonkman LE, Klaver R, van Veluw SJ, Zwanenburg JJ, Kuijer JP, Pouwels PJ, Twisk JW, Wattjes MP, Luijten PR, Barkhof F, Geurts JJ. Increased cortical grey matter lesion detection in multiple sclerosis with 7 T MRI: a post-mortem verification study. Brain 2016;139:1472-81. [Crossref] [PubMed]
14. Zhang Z, Liu S, Lin X, Teng G, Yu T, Fang F, Zang F. Development of laminar organization of the fetal cerebrum at 3.0T and 7.0T: a postmortem MRI study. Neuroradiology 2011;53:177-84. [Crossref] [PubMed]
15. Gascho D, Zoelch N, Sommer S, Tappero C, Thali MJ, Deininger-Czermak E. 7-T MRI for brain virtual autopsy: a proof of concept in comparison to 3-T MRI and CT. Eur Radiol Exp 2021;5:3. [Crossref] [PubMed]
16. Polimeni JR, Uludağ K. Neuroimaging with ultra-high field MRI: Present and future. Neuroimage 2018;168:1-6. [Crossref] [PubMed]
17. Aringhieri G, Zampa V, Tosetti M. Musculoskeletal MRI at 7 T: do we need more or is it more than enough? Eur Radiol Exp 2020;4:48. [Crossref] [PubMed]
18. Becher JC, Bell JE, Keeling JW, McIntosh N, Wyatt B. The Scottish perinatal neuropathology study: clinicopathological correlation in early neonatal deaths. Arch Dis Child Fetal Neonatal Ed 2004;89:F399-407. [Crossref] [PubMed]
19. De Martino F, Yacoub E, Kemper V, Moerel M, Uludağ K, De Weerd P, Ugurbil K, Goebel R, Formisano E. The impact of ultra-high field MRI on cognitive and computational neuroimaging. Neuroimage 2018;168:366-82. [Crossref] [PubMed]
20. Zhang Z, Lin X, Yu Q, Teng G, Zang F, Wang X, Liu S, Hou Z. Fetal ocular development in the second trimester of pregnancy documented by 7.0 T postmortem Magnetic Resonance Imaging. PLoS One 2019;14:e0214939. [Crossref] [PubMed]
21. Thayyil S, Cleary JO, Sebire NJ, Scott RJ, Chong K, Gunny R, Owens CM, Olsen OE, Offiah AC, Parks HG, Chitty LS, Price AN, Yousry TA, Robertson NJ, Lythgoe MF, Taylor AM. Post-mortem examination of human fetuses: a comparison of whole-body high-field MRI at 9.4 T with conventional MRI and invasive autopsy. Lancet 2009;374:467-75. [Crossref] [PubMed]
22. Staicu A, Albu C, Popa-Stanila R, Bondor C, Chiriac L, Eniu D, Goidescu I, Florian AR, Surcel M, Cruciat G, Muresan D, Rotar I. Whole-body non-forensic fetal virtopsy using postmortem magnetic resonance imaging at 7 Tesla vs classical autopsy. Ultrasound Obstet Gynecol 2024;64:661-8. [Crossref] [PubMed]
23. Genest DR, Williams MA, Greene MF. Estimating the time of death in stillborn fetuses: I. Histologic evaluation of fetal organs; an autopsy study of 150 stillborns. Obstet Gynecol 1992;80:575-84. [PubMed]
24. Brookes JS, Hagmann C. MRI in fetal necropsy. J Magn Reson Imaging 2006;24:1221-8. [Crossref] [PubMed]
25. Thayyil S, Sebire NJ, Chitty LS, Wade A, Olsen O, Gunny RS, Offiah A, Saunders DE, Owens CM, Chong WK, Robertson NJ, Taylor AM. Post mortem magnetic resonance imaging in the fetus, infant and child: a comparative study with conventional autopsy (MaRIAS Protocol). BMC Pediatr 2011;11:120. [Crossref] [PubMed]
26. Pfefferbaum A, Sullivan EV, Adalsteinsson E, Garrick T, Harper C. Postmortem MR imaging of formalin-fixed human brain. Neuroimage 2004;21:1585-95. [Crossref] [PubMed]
27. Dawood Y, Strijkers GJ, Limpens J, Oostra RJ, de Bakker BS. Novel imaging techniques to study postmortem human fetal anatomy: a systematic review on microfocus-CT and ultra-high-field MRI. Eur Radiol 2020;30:2280-92. [Crossref] [PubMed]
28. Arthurs OJ, Thayyil S, Pauliah SS, Jacques TS, Chong WK, Gunny R, Saunders D, Addison S, Lally P, Cady E, Jones R, Norman W, Scott R, Robertson NJ, Wade A, Chitty L, Taylor AM, Sebire NJMagnetic Resonance Imaging Autopsy Study. (MaRIAS) Collaborative Group. Diagnostic accuracy and limitations of post-mortem MRI for neurological abnormalities in fetuses and children. Clin Radiol 2015;70:872-80. [Crossref] [PubMed]
29. Xu Y, Hu Y, Xu J, Liu M. Study on the application of tracking methodology and PDCA circulation management in quality control of hospital infection management. Panminerva Med 2021; Epub ahead of print. [Crossref] [PubMed]