NT3 - Early experience of a 3D interactive model that demonstrates anal fistula anatomy generated from Magnetic Resonance Imaging.

Purpose/Background: Magnetic Resonance Imaging (MRI) has become the gold-standard for the pre-operative radiological assessment of anal fistulas. High resolution, multi-planar images accurately demonstrate fistula anatomy and its relationship to perianal structures. However, in the absence of a standardised approach, communication of MRI images between radiologists and surgeons can be hampered by the quality or individual interpretation of written radiology reports. We present our early applied experience of using a 3D model that accurately demonstrates each fistula’s unique anatomy, in an accessible format to both experts and trainees.

Methods/Interventions: MRI images are obtained from patients and reported by a specialist radiologist. A standard written report is provided and accompanies the 3D model that is generated from the images. This is a two-step process undertaken by a radiologist with specialist software. A template pelvic floor model is fitted to the patient’s pelvic floor dimensions, following which the fistula is highlighted (brushed) on multiple coronal images. This maintains specialist interpretation of the images when constructing each model that, on average, can be completed in 5 minutes. The software then translates this data into a 3D interactive model that can be viewed by surgeons.

Results/Outcomes: We present three examples of the 3D fistula model to demonstrate its use. Images are presented in 2D for the purpose of this abstract. Patient 1 (Figure 1) has a trans-sphincteric tract (green) with two internal openings in the upper anal canal (green and yellow) demonstrated on a posterior, longitudinal view. Patient 2 has a complicated fistula and shows how the model can include the levator plate (purple) with an inter-sphincteric fistula in the left posterior quadrant and a trans-sphincteric fistula that opens in the right posterior quadrant. The tracts cross high in the ischioanal fossa and extend above puborectalis. This can be seen in Figure 2 (posterior longitudinal view), Figure 3 (axial view) and in Figure 4 (posterior longitudinal view) where the rectum (red) has been added to demonstrate its relationship. Patient 3 (Figure 5) has an internal opening that gives rise to an anterior midline large volume trans-sphincteric tract. There is a secondary tract extending cranially abutting the levator plate that is in close proximity to the urethra. This case demonstrates how other anatomy can be included in the model as here the bladder and urethra are brushed blue.
Initial feedback from a cohort of UK and US surgeons was obtained through a webinar and online survey. This suggested that, for many surgeons, the fistula model accompanying an MRI report may be preferable to a written report alone to understand the anatomy of a fistula, counsel patients and train junior colleagues.

Conclusion/Discussion: We have demonstrated the application and utility of this new technology that enhances MRI scan interpretation. Perceived benefits are that it allows a better understanding of fistula anatomy, facilitating pre-operative planning, predicting sphincterotomy and improving patient counselling. Future work is planned to formally assess these benefits and explore other clinical applications.