Up to 40% of patients treated for rectal cancer suffer from therapy-related symptoms. Innervation injury is one of the suggested pathomechanisms of those symptoms hence the development of a valid, non-invasive tool for the assessment of neural systems is crucial. The aim of this work is to study the fractal properties of the surface electromyography signals obtained from patients suffering from rectal cancer. The anal sphincter activity was investigated for the group of 15 patients who underwent surgical treatment. Multifractal detrended fluctuation analysis was implemented to analyze the data, obtained at four different stages: one before treatment and three times after the surgery. The results from the standard detrended fluctuation analysis and empirical mode decomposition methods are presented and compared. The statistically significant differences between the stages of treatment were identified for the selected spectral parameters: width and maximum of the spectrum.