Objective: To analyse the results of anal-sphincter repair; to examine the feasibility of motor latency measurements of the anal sphincter after magnetic sacral stimulation in healthy subjects, patients with a spinal cord injury and patients with faecal incontinence, using a new recording technique; to study, in the same 3 groups, the effects ofphasic magnetic sacral root stimulation on the anal sphincter and rectum; to describe normal, undisturbed anatomy of the anal canal and perianal structures in both men and nulliparous women using high-resolution phased array magnetic resonance imaging; to study pelvic floor movements in healthy volunteers of both sexes using a new instrument.

Methods: Latencies were recorded with an intraanal, bipolar sponge electrode and an intrarectal ground electrode. Rectal volume changes were measured with a barostat. Highresolution magnetic resonance images were obtained without an endoanal coil. Pelvic floor movements where measured with the subject seated, using a magnet attached to a rectal balloon.

Results: After a median period of 40 months, 31/55 female patients rated the result of analsphincter repair as excellent or good. Age>50 years, and post-operative urgency and loose stools were associated with poor outcome. Eight patients became fully continent for stool. 17% of the latency measurements failed. There were no significant differences between leftand right-sided stimulation. Faecal incontinence patients had prolonged pudendal nerve terminal motor latencies and prolonged latencies after left-sided magnetic stimulation. Phasic magnetic stimulation increased anal pressure in 100% of the 14 healthy subjects, 86% of the 14 spinal cord injury patients and 73% of the 18 faecal incontinence patients. A decrease in rectal volume was provoked in respectively 72 %, 79 % and 50 %. In all 33 volunteers, anal and perianal structures could be well defined by magnetic resonance imaging. The mid-anal canal was significantly longer than its anterior and posterior part. The female anterior sphincter was shorter than the male and occupied 30 % of the anal canal length. The female perineal body was thicker and easier to define than the male. The median pelvic floor lift and descent measured in 28 healthy volunteers, were 2 cm and 1.8 cm respectively. Day-to-day and inter-observer reproducibility were good. 20/28 subjects were able to expel the rectal balloon.

Conclusions: Anal-sphincter repair does not restore complete continence but leads to a satisfactory result in more than half of the patients. Additional bowel symptoms are common at follow-up. Latency measurements after magnetic stimulation are minimally invasive and have a low failure rate. They may be used to test the integrity of the distal motor pathway in patients who may benefit from continuous sacral root stimulation. Magnetic sacral root stimulation produces an increase in anal and rectal pressure and a decrease in rectal volume. Phased array magnetic resonance imaging is non-invasive and allows an accurate description of the normal anatomy of the anal canal and perianal structures. The new developed instrument measures cranial and caudal movement of the pelvic floor with minimal discomfort and good reproducibility.