Three-dimensional reconstruction of magnetic resonance images of the anal sphincter and correlation between sphincter volume and pressure

OBJECTIVE: The purpose of this study was to assess the correlation between internal and external anal sphincter volumes and manometric anal pressures. STUDY DESIGN: Ten healthy nulliparous women underwent anal sphincter magnetic resonance imaging and anal manometry measurement. A 3-dimensional reconstruction of magnetic resonance images was accomplished with the use of 3-dimensional slicer. Sphincter volumes were measured 3 times by the same observer for each of 10 patients. The intrarater reliability was measured with the use of the intraclass correlation coefficient (ICC = σ² _patients/(σ² _patients + σ² _error)) from a 2-way analysis of variance model with terms for patient and measurement trial. Measurements that were recorded on anal manometry included squeeze length, length of the high-pressure zone, and maximal resting and squeeze pressures. RESULTS: The mean volumes (± SD) were 18.77 ± 4.64 cm³, 13.82 ± 3.8 cm³, and 32.36 ± 8.37 cm³ for internal, external, and combined sphincters, respectively. Intrarater reliability was 98% for external sphincter volume (95% CI, 94%-99%), 98% for internal sphincter volume (95% CI, 94%-99%), and 99% for total volume (95% CI, 97%-100%). On the 3-dimensional images, the internal sphincter was found to be cylindric in shape, with an ellipse as a base. It is elongated in the anterior and posterior direction and flattened on the sides. The external sphincter was found to be funnel-shaped, being narrower caudad and widening in the cephalad direction. Similar to the internal sphincter, the external sphincter is elongated in the anteroposterior diameter. Volumes of the internal, external, and combined sphincters did not correlate with the maximum pressures at rest and squeeze. Correlations higher than r = 0.5 were observed for all 3 sphincter volume measurements versus high pressure zone at squeeze. The highest correlation, r = 0.66, was for internal sphincter volume versus high pressure zone at squeeze. CONCLUSION: Three-dimensional reconstruction of the rectal sphincter musculature can be performed easily with 3-dimensional software. Measurements of the sphincter volumes have excellent intrarater reliability. Sphincter volumes do not correlate with pressures at rest or squeeze, but the internal sphincter volume correlates with the length of the high pressure zone at squeeze. Contrary to current generalized concepts, it is possible that the internal sphincter may play some role in generating the squeeze pressure. More research is necessary in applying 3-dimensional magnetic resonance image reconstruction in patients with different parity and continence status. Reconstruction of magnetic resonance images of the rectal sphincter musculature may prove to be beneficial in planning the treatment of patients with fecal incontinence.