The Value of Bowel Sounds in Evaluating Postoperative Intestinal Function Recovery in Colorectal Cancer Patients (BRIEF)

Colorectal cancer (CRC) is the third most common malignancy worldwide and the second leading cause of cancer-related mortality. In recent years, both the incidence and mortality of colorectal cancer in China have shown a significant upward trend. With population aging and lifestyle changes, the disease is increasingly affecting younger individuals, with the proportion of patients under 40 years of age steadily rising. Treatment strategies vary according to disease stage; however, for most cases amenable to primary resection, surgical intervention remains the cornerstone of therapy.

With the widespread implementation of Enhanced Recovery After Surgery (ERAS) protocols, postoperative rehabilitation has been substantially improved through optimization of preoperative, intraoperative, and postoperative management strategies, such as shortened preoperative fasting, the use of laparoscopic techniques, and early postoperative oral intake, thereby facilitating accelerated recovery and earlier discharge. Following curative surgery for colorectal cancer, patients commonly experience reduced or transiently absent intestinal motility due to surgical manipulation and the effects of anesthesia, with gradual return to normal function over several days. Early postoperative feeding has been shown to promote gastrointestinal functional recovery.

However, a subset of patients may exhibit delayed recovery of bowel function, characterized by diminished bowel sound activity, prolonged time to first flatus, and often accompanied by abdominal distension. Postoperative ileus (POI) is one of the most common causes of delayed postoperative gastrointestinal recovery and typically presents as transient suppression of gastrointestinal motility. When the duration exceeds the expected physiological range, potential pathological conditions, such as early postoperative small bowel obstruction (EPSBO), should be considered. In such circumstances, oral intake may exacerbate the ileus and even precipitate intestinal obstruction. Therefore, timely differentiation and accurate diagnosis of the patient's postoperative gastrointestinal status are of critical importance.

Abdominal auscultation, as a noninvasive diagnostic technique, has a history spanning several centuries. In 1905, Cannon was the first to report systematic investigations of abdominal auscultation, proposing that regular bowel sounds are generated by intestinal peristaltic activity. Subsequent studies have demonstrated that bowel sounds vary by anatomical location: sounds originating from the small intestine are typically high-pitched, high-amplitude, and frequent, whereas colonic sounds are lower in frequency, weaker in amplitude, and less frequent. Postoperatively, the recovery of gastrointestinal motility differs among intestinal segments, with the small intestine recovering first, followed by the stomach and the colon.

Abdominal auscultation has been widely used as an adjunctive tool in the evaluation of conditions such as intestinal obstruction and ileus owing to its immediacy and convenience and is commonly applied in patients with delayed postoperative gastrointestinal recovery. However, the reliability of auscultation-based assessment remains controversial because of the inherent subjectivity in clinicians' descriptions of bowel sounds. A prospective study by Felder et al. demonstrated a low sensitivity of 22%-32% for manual auscultation in distinguishing healthy from diseased individuals, along with low positive predictive values of 23%-44%. Similarly, Thomas et al. conducted a prospective study examining the relationship between auscultated bowel sounds and postoperative oral tolerance and flatus, finding no significant correlation between bowel sounds and postoperative abdominal distension, flatus, defecation, or tolerance of oral intake.

Therefore, digital bowel sound monitoring devices, which objectively detect and analyze bowel sounds, have emerged as important adjunctive tools for postoperative bowel sound assessment. These systems enable precise quantification of acoustic features such as frequency, amplitude, and energy. Currently, both domestic and international researchers have begun to explore the use of bowel sound monitoring systems in the diagnosis and management of gastrointestinal disorders. Preliminary small-sample studies suggest that such systems hold substantial clinical potential in the precision management of intestinal obstruction, the diagnosis of functional gastrointestinal disorders, and the prediction of acute gastrointestinal injury in critically ill patients.

Nevertheless, progress in bowel sound monitoring technology has been relatively slow, primarily because of several challenges: (1) the weak and highly stochastic nature of bowel sound signals; (2) substantial interference from environmental and internal physiological noise; and (3) dynamic fluctuations in bowel sounds related to individual physiological states, with marked interindividual variability. To address these challenges, our center has independently developed a highly intelligent digital bowel sound monitoring system based on a distributed, multichannel, real-time acquisition framework. This system enables simultaneous collection of bowel sound signals from different abdominal regions while concurrently recording **cardiac sounds** and ambient noise as reference signals for noise suppression, followed by centralized classification and recognition of bowel sounds. Through this approach, bowel sound signals can be accurately extracted in complex acoustic environments, significantly improving the accuracy of feature detection, parameter computation, and classification algorithms while also assisting clinicians with synchronized auscultation.

At present, our center has conducted multiple iterations of the bowel sound monitoring device, completing optimized denoising and detection algorithms that enable accurate extraction of bowel sound signals in complex environments. Furthermore, we have implemented a bowel sound recognition algorithm that integrates an attention-based Broadcasting Residual Neural Network (Broadcasting ResNet) with a Dual-Path Recurrent Neural Network (Dual-Path RNN). This algorithm is capable of extracting time-frequency features of bowel sounds, thereby supporting the auxiliary diagnosis of various intestinal disorders.

In summary, there is an urgent need for a high-quality, prospective clinical trial to investigate the predictive value of digital bowel sound monitoring for postoperative gastrointestinal functional recovery. Such a study would facilitate the clinical translation of this technology and assist clinicians in the diagnosis and early intervention of postoperative intestinal dysfunction.