Application of MRI 3 D reconstruction of anal fistula in the standardized training of anorectal residents

Yikun Li, Yibo Yao, Shijie Zhang, Fenghua Weng, Kai Ma, Chen WangResearch Square 2025 · DOI

TL;DR

This study demonstrates that integrating 3D MRI reconstruction into the training of anorectal residents significantly improves their ability to accurately and quickly interpret complex anal fistulas. By bridging the "spatial gap" between 2D cross-sectional slices and 3D anatomy, the technology reduced average diagnostic time by over 40% while markedly increasing the accuracy of anatomical classification.

What Problem Does This Paper Address?

Anal fistulas—abnormal tunnels connecting the anal canal to the skin—are notoriously difficult to treat because of their complex, branching paths through pelvic muscles. Magnetic Resonance Imaging (MRI) is the gold standard for diagnosis, but traditional 2D MRI slices require a high degree of "mental reconstruction."

For medical residents, learning to translate these flat images into a 3D surgical plan is a steep challenge. Misinterpreting the relationship between a fistula and the sphincter muscles can lead to surgical failure or, worse, fecal incontinence. The authors address the educational gap: can 3D visualization tools accelerate the learning curve and improve diagnostic precision for trainees?

Methods

The researchers conducted a randomized controlled trial involving 30 residents at Longhua Hospital (Shanghai University of Traditional Chinese Medicine).

  • The Split: Residents were divided into a control group (2D images only) and a 3D group (2D images plus a 3D reconstruction model).
  • The Task: Both groups interpreted the imaging data of 10 identical patient cases, filling out structured reports on fistula location, classification (using the Parks and AGA systems), and the presence of branches or abscesses.
  • The Tech: 3D models were generated using 3D Slicer 4.8.2, an open-source medical image processing platform. Radiologists manually outlined the fistula lesions and key anatomical structures (internal/external sphincters, levator ani muscle) in different colors to create a rotatable, 360-degree digital model.

Datasets

The study utilized a curated clinical dataset of 10 confirmed cases of anal fistula treated at Longhua Hospital between June 2023 and January 2024.

  • The cases were selected to represent a variety of complexities based on the Parks classification (4 cases of sphincter type 4, 3 of type 3, 1 of type 1, and 2 of type 2).
  • Cases involving inflammatory bowel disease, tuberculosis, or tumors were excluded to focus specifically on cryptoglandular fistulas.

Key Insights and Ideas

  • Reducing Cognitive Load: The most significant "aha moment" is the impact on spatial thinking. The 3D models provide "stereoscopic vision" and "dynamic exploration," allowing students to see the fistula's path relative to muscle layers without the mental gymnastics required by 2D slices.
  • Speed as a Proxy for Competence: The 3D group wasn't just more accurate; they were much faster. This suggests that 3D models help trainees reach a "gestalt" understanding of the pathology much earlier in the diagnostic process.
  • Targeted Assistance: The 3D models were most helpful for identifying the direction of the fistula, its classification, and branching tubes—the exact areas where 2D interpretation typically fails for novices.
  • Subjective Ease: Residents in the 3D group reported significantly lower levels of perceived difficulty, which is a critical factor in student engagement and the prevention of "diagnostic fatigue."

Key Results

The quantitative improvements were stark across nearly all metrics:

  • Efficiency: The 3D group averaged 239.89 seconds per case, compared to 409.28 seconds for the control group (p < 0.01).
  • Accuracy: The 3D group showed statistically significant improvements in identifying internal/external mouth positions, AGA/Parks classifications, and the presence of branch ducts (p < 0.01).
  • Student Feedback: 73.08% of the residents found the 3D models "very helpful," particularly for visualizing the spatial relationship between the fistula and the pelvic floor muscles.

Limitations and Future Work

  • Sample Size: The study was limited to 30 residents at a single institution. Larger, multi-center trials would be needed to generalize these findings.
  • Model Complexity: The current 3D models focused primarily on the fistula and major muscles. The authors note that adding more anatomical structures (like nerves and blood vessels) could further improve the models' utility for surgical planning.
  • Long-term Retention: The study measured immediate performance. It remains to be seen if training with 3D models helps residents eventually become better at reading 2D images alone, or if it creates a dependency on 3D visualization.
  • Automation: The 3D reconstructions were performed manually by specialists. Future work could explore AI-driven automated segmentation to make this technology more scalable for daily clinical and educational use.

Summary updated

Abstract

<title>Abstract</title> Objective: To explore the practical application effect of MRI 3D reconstruction technology of anal fistula in anorectal teaching. Methods: 30 doctors who received standardized resident training in the anorectal department in Longhua Hospital affiliated to Shanghai University of Traditional Chinese Medicine from January to June 2024 were selected as the research subjects. According to the order of entry, randomly divided into control group and 3 D group, 15 physicians in each group. The control group read only the two-dimensional images of the anal fistula MRI, and the 3 D group read both the two-dimensional MRI images and the 3 D reconstruction model. The two groups watched the imaging data of the same 10 patients with anal fistula, and compared the reading ability of doctors (the correct rate of fistula structured report, the time spent by reading), and the physician self-evaluation (the difficulty of reading, the help of 3 D reconstruction model). Results: In the report of fistula structure: internal and external mouth location of fistula, AGA classification, Parks classification, presence of branch duct, rectal scrotal sac / vaginal fistula, The accuracy of the 3 D group was higher than that of the control group (p &lt;0.01); 3 D group average reading time 239.89s, The 409.28s was significantly shorter compared with the control group (p &lt;0.01); The 3 D group had less self-rated of difficulty (p &lt;0.01), And 73.08% of the doctors thought that the 3 D reconstruction model was very helpful for the reading film (73.08% was helpful), The main manifestations in the anal fistula fistula direction, fistula classification and whether there is a branch tube is more obvious. Conclusion: The three-dimensional reconstruction technology of anal fistula MRI can significantly improve the MRI reading ability of resident doctors in the standardized training of anorectal department, shorten the reading time, cultivate the spatial thinking ability, and improve the clinical teaching effect of anorectal department.
BibTeX
@inproceedings{li2025application,
  author = {Yikun Li and Yibo Yao and Shijie Zhang and Fenghua Weng and Kai Ma and Chen Wang},
  title = {{Application of MRI 3 D reconstruction of anal fistula in the standardized training of anorectal residents}},
  booktitle = {Research Square},
  year = {2025},
  doi = {10.21203/rs.3.rs-5708146/v1},
}
Figure