共找到 20 条结果
Artificial intelligence (AI) is rapidly transforming surgical practice with growing applications in colon and rectal surgery. This review explores perioperative AI tools that assist with operative planning, intraoperative guidance, and outcome optimization. Preoperative innovations include machine learning models that outperform traditional risk calculators for predicting complications and readmissions, as well as computer vision and radiomics for analyzing imaging in colorectal cancer and inflammatory bowel disease. The integration of molecular and multiomics data further enhances personalized, precision surgical planning. Intraoperatively, deep learning enables computational identification of critical anatomy, including vascular structures, ureters, and pelvic nerves, and supports the objective analysis of advanced imaging techniques such as indocyanine green fluorescence. In terms of surgical techniques, AI-driven video analysis facilitates surgical phase recognition and automated skill assessment, whereas emerging vision-language models and surgical foundation models promise improved documentation and context-aware guidance. Future directions include generative AI for simulation, AI-based coaching, and progress toward autonomous surgical robotics. Although research remains in the early stages and is not yet ready for widespread clinical implementation, ongoing work within the field of colorectal surgery underscores the potential of AI to augment decision-making and standardize surgical care.
Multidisciplinary team (MDT) discussions have become a cornerstone of colorectal cancer (CRC) management, integrating the expertise of surgeons, oncologists, radiologists, pathologists, and allied health professionals to facilitate personalized, evidence-based care. However, increasing complexity in treatment options, particularly with the rise of neoadjuvant strategies and immunotherapies, has rendered decision-making more challenging. Traditional tools-including clinical guidelines, risk calculators, and nomograms-offer structured decision support but lack flexibility and personalization. Artificial intelligence (AI), particularly through machine learning (ML), radiomics, and large language models (LLMs), is emerging as a transformative adjunct to clinical decision-making in CRC. Machine learning models have demonstrated strong predictive performance for treatment response, recurrence risk, and surgical complications, while radiomics and deep learning have improved diagnostic accuracy and treatment response prediction using imaging and endoscopy. LLMs such as ChatGPT have shown promising concordance with MDT recommendations in early studies, especially for standard clinical scenarios. However, limitations remain in handling complex, nuanced cases. Despite their growing capabilities, AI and LLMs are not yet integrated into routine MDT workflows due to concerns about interpretability, regulatory oversight, and ethical challenges. Future directions include developing real-time, multimodal AI-MDT platforms, improving explainability, ensuring equitable data representation, and integrating AI training into medical education. This review outlines current evidence on AI integration within CRC MDTs, highlighting both its clinical potential and the barriers that must be addressed to ensure safe, effective, and equitable implementation. Ultimately, AI is poised to augment-not replace-human expertise, enhancing the consistency, efficiency, and personalization of multidisciplinary CRC care.
From its origins in the late 1800s, colon and rectal surgery has evolved from itinerant practitioners to a highly specialized surgical discipline. We highlight the historical development and the current state of the colorectal surgery training programs. The American Proctologic Society (APS) was founded by Joseph M. Mathews in 1899 that catalyzed the formation of early training programs in University of Minnesota (1916) and Mayo Clinic (1919). The American Board of Colon and Rectal Surgery (ABCRS) was recognized as the eighteenth medical specialty board in 1949. Formal accreditation of the programs began in 1975, with the Accreditation Council of Graduate Medical Education (ACGME) gradually implementing standardized curricula and competency-based training milestones. The National Residency Match Program (NRMP) and Electronic Residency Application Service (ERAS) adopted in 1984 and 2003, respectively, further streamlined the application process. Training innovations like the laparoscopic education committees (1997) and robotic surgery programs (2010) bear testament to the dedication of the program directors to continually advance this field. With a structured curriculum, rigorous examinations, and a plethora of study resources, this specialty has evolved from informal apprenticeships to competency-based residency programs. Current challenges include grade inflation in evaluation, a 67% match rate (2024) despite program expansion, adapting to evolving trends in medical and surgical care, and mastering new surgical techniques in a limited amount of time. The collaboration between governing bodies ensures high-quality training standards for the colon and rectal surgeons of the future.
Artificial intelligence (AI) is rapidly transforming surgical care, offering unprecedented capabilities in diagnostics, planning, and intraoperative guidance. However, its integration into clinical practice raises complex ethical challenges that must be addressed to ensure responsible and equitable use. This chapter aims to explore the ethical implications of AI in surgery through the lens of the four foundational principles of medical ethics: autonomy, beneficence, nonmaleficence, and justice. Respecting patient autonomy requires clinicians to disclose all relevant information and ensure understanding of a proposed intervention (or AI tool) for a patient to make an informed, voluntary decision. Numerous studies have shown that both patients and clinicians often lack sufficient understanding of AI tools, complicating efforts to explain their intended function, or obtain truly informed consent for their use in patient care. This has the potential to diminish trust in the patient-clinician relationship and must be considered when using AI tools. AI's potential to uphold beneficence is evident in its ability to enhance surgical precision and outcomes. Yet, its reliance on historical data introduces risks of bias and error, threatening the principle of nonmaleficence. In this chapter, we explore these topics further to highlight the need for robust oversight and clinician involvement to prevent harm to patients. When biased data are used to train an AI tool, it may lead to unequal care across patient populations-i.e., exacerbating existing disparities in health care. Additionally, the lack of clear accountability for AI-driven errors raises legal and ethical concerns about liability-whether it lies with clinicians, health care institutions, or developers. To ethically integrate AI into surgical practice, the chapter calls for comprehensive frameworks that ensure transparency, data integrity, clinician and patient education, and regulatory oversight. These measures are essential to safeguard patient welfare as AI continues to reshape the future of surgical care.
Colorectal cancer (CRC) is the third most commonly diagnosed malignancy worldwide. Prognosis is significantly worsened in patients with colorectal liver metastases (CRLM), whose management requires a multidisciplinary approach encompassing diagnosis, systemic therapy, surgery, and surveillance. Artificial intelligence (AI) offers the potential to improve treatment processes and outcomes across all aspects of CRLM care. This review summarizes current and future applications of AI throughout the CRLM treatment continuum. In diagnostics, radiomics-based AI models have demonstrated improved sensitivity in detecting small or ambiguous liver lesions, supporting radiologist interpretation, and improving efficiency. Similarly, AI models are increasingly employed to predict systemic treatment response, using deep learning (DL) to extract imaging-derived features that correlate with genomic and histopathologic profiles relevant to therapy selection. In surgical planning, AI tools can assist in preoperative preparation and optimization by measuring tumor volume and transection planes. Intraoperatively, computer vision and augmented reality are emerging to support tumor localization, margin assessment, and real-time anatomical navigation. Postoperatively, advanced AI models can integrate clinical, radiologic, and molecular data to stratify recurrence risk and inform individualized follow-up strategies. Despite its promise, clinical translation of AI in CRLM remains limited by the retrospective nature of many studies, challenges with external validation, and limitations in the interpretability of model decisions. Still, AI has the potential to be a transformative tool in the treatment of CRLM by supporting precision, standardization, and personalization across the treatment spectrum.
Effective patient-physician communication is a cornerstone of surgical care, yet increasing clinical complexity and time constraints often limit opportunities for shared decision-making. Artificial intelligence (AI) is emerging as a tool to supplement these interactions by improving perioperative risk communication, tailoring patient education, and reducing physician workload. Current applications include chatbots for preconsultation history gathering, AI-generated decision aids, and risk prediction models, which synthesize large datasets to provide individualized outcome predictions. Visualization platforms and large language model-driven documents further demonstrate potential to improve readability, completeness, and patient comprehension compared with traditional surgeon-generated materials. Research suggests that AI may enhance shared decision-making by increasing patient understanding, reducing decisional conflict, and supporting individualized risk-benefit discussions. Populations with complex needs-including patients with multiple comorbidities, limited English proficiency, or low health literacy-may experience particular benefit from AI-assisted communication. Indirect advantages are also evident; studies of ambient AI scribes report decreased documentation burden, reduced after-hours work, and improved patient engagement, suggesting downstream improvements in the patient-physician relationship. Despite these promising developments, limitations persist. AI models are prone to bias and lack transparency, raising concerns regarding fairness, accuracy, and trustworthiness. Overreliance on AI risks diminishing essential human elements of care, while errors in translation or misrepresentation of nuance may disproportionately affect vulnerable groups. Ensuring robust evaluation, representative training datasets, and clinician oversight is critical to safe implementation. AI tools show substantial promise in improving surgical communication, but more research is required to establish their effectiveness, address ethical challenges, and define best practices for integration into clinical care.
Surgical education in colorectal surgery is a multifaceted process that requires the acquisition of theoretical knowledge of basic sciences, disease recognition, and the command of evidence-based treatments. Technical skill building through simulation and supervised operative responsibility is also essential. The rise of minimally invasive techniques poses further challenges, requiring specialized training in new, advanced, and costly technologies. This paper aims to dissect the current state of colorectal surgery education globally, elaborating on quality indicators of the delivery of training. Significant disparities are noted internationally, with high-income countries offering structured training programs, specialized fellowships, and formal certification, while many low- and middle-income countries face challenges in accessing advanced training resources. Furthermore, cultural and societal factors significantly impact training outcomes across different health systems, including hierarchical structures, gender disparities, and limited diversity in surgical leadership. By acknowledging these inequalities and the factors that foster them, solutions can be explored, aiming to ensure equitable access to high-quality colorectal surgery education worldwide, ultimately improving patient outcomes.
Despite significant advances in colorectal surgery (CRS), postoperative venous thromboembolism (VTE) remains a critical issue that contributes to substantial morbidity and mortality. The incidence of VTE, including deep vein thrombosis, pulmonary embolism, and portomesenteric vein thrombosis, in the colorectal surgical population varies from 2 to 15%, with elevated risks in patients with colorectal cancer and inflammatory bowel disease. This review article examines the effects of VTE on postoperative outcomes and explores the efficacy of extended chemoprophylaxis (ePPx) for mitigating these risks. We will review the rates of morbidity and mortality associated with VTE, as well as the role of postdischarge ePPx in VTE prevention, while exploring how other specialties utilize ePPx strategies to decrease their postdischarge VTE rates, some of which may be translatable to CRS patients. Our analysis highlights the role of various prophylactic measures, including low-molecular-weight heparin (LMWH), aspirin, and direct oral anticoagulants (DOACs), comparing their effectiveness and cost implications as well as the use of thromboelastography to help guide ePPx management. Overall, findings suggest that VTE ePPx with LMWH significantly reduces the incidence of postoperative VTE and related complications, although patient compliance remains a challenge. While aspirin is a cost-effective alternative, its efficacy in patients with CRS requires further investigation. Emerging data on DOACs indicate their potential as viable options for ePPx, although their safety profile requires careful consideration. Tailored ePPx strategies, particularly with LMWH, appear to be crucial for reducing VTE in CRS patients. Further research is needed to refine the prophylactic approaches and establish standardized guidelines that incorporate new insights into VTE prevention and management in CRS.
Preoperative risk mitigation is vital for improving surgical outcomes and patient safety, particularly in colorectal cancer (CRC) surgeries. While traditional approaches have primarily focused on postoperative care, the preoperative period is a unique opportunity for intervention to enhance patients' physiological readiness for surgery and minimize complications. This narrative review examines the general principles of preoperative risk mitigation, identifies common complications in colorectal surgery, and explores the impact of patient comorbidities on surgical outcomes. Additionally, the review discusses the strategic management of modifiable risk factors. The integration and impact of prehabilitation protocols in colorectal surgery are also evaluated. Evidence indicates that addressing modifiable preoperative risk factors can significantly improve surgical outcomes. Obesity management, nutritional optimization, and enhancing functional capacity through prehabilitation have been shown to reduce postoperative complications. Multimodal prehabilitation benefits high-risk and frail patients, improving their postoperative recovery and reducing complication rates. The preoperative period is crucial for implementing risk mitigation strategies to enhance surgical outcomes in CRC patients. Interventions targeting modifiable risk factors and integrating prehabilitation protocols can complement traditional postoperative care, improving recovery and reducing complications. Despite promising findings, further research is necessary to fully understand the long-term benefits and optimize preoperative interventions to mitigate postoperative morbidities effectively.
Artificial intelligence (AI) and machine learning are poised to transform trauma care across the entire continuum, from prehospital triage to postoperative critical care. Trauma systems are uniquely suited for AI integration due to the time-sensitive, high-volume, and data-rich nature of care delivery. In this narrative review, we describe current and emerging AI applications across the trauma care spectrum, including triage, acute resuscitation, operative decision-making, intensive care, and detection of complications. We also examine AI's potential in nontraditional care environments, including prehospital, rural, and military settings, where resource constraints and variability in provider expertise pose significant challenges. Across multiple domains, AI models outperform conventional approaches in predicting injury severity, identifying patients in need of intervention, and detecting complications. Specific tools include AI-powered triage support, resuscitation sequencing systems, real-time imaging interpretation, and outcome prediction applications. Despite this promise, many AI applications remain investigational, and widespread adoption will require validation, transparency, and alignment with ethical and regulatory standards. Thoughtful implementation of AI in trauma care has the potential to enhance decision-making, improve patient outcomes, and address disparities in access to high-quality trauma care.
Quality modern colorectal surgery education demands more than simply teaching technical skills; it requires a deep understanding of how surgeons learn. This article explores foundational pedagogical theories that support effective instruction across all stages of surgical training. Cognitive load theory, behaviorist models, and constructivist approaches such as experiential and social learning are examined for their relevance to surgical education. Deliberate practice and mastery learning frameworks offer structured methods for developing procedural expertise. Teaching strategies such as simulation and feedback models are discussed to promote competence and autonomy. Additionally, the article considers emotional and motivational components of learning, highlighting the roles of self-efficacy, feedback, and reflective practice. By aligning educational practice with established theory, colorectal surgeons can more effectively train future colleagues and elevate standards of care, professionalism, and continuous learning in a dynamic healthcare landscape.
Artificial intelligence (AI) offers a promising solution to the long-standing challenge of accurately predicting treatment response in rectal cancer. In this narrative review, we summarize current AI-driven approaches to predicting pathologic complete response in rectal cancer. We also outline key barriers to clinical translation, including lack of standardization, small and geographically skewed training cohorts, domain shift across scanners and institutions, and broader ethical, regulatory, and medicolegal concerns. Finally, we highlight future directions, including federated learning to enable privacy-preserving multicenter model training, and emerging concepts such as virtual and digital twins that may support real-time adaptive therapy. These advances suggest that AI-based prediction of response in rectal cancer could be extremely valuable, but will require methodologically rigorous, multi-institutional efforts to be safely and equitably implemented.
The rigorous demands of colon and rectal surgery necessitate comprehensive assessment and continuous professional development. Historically, surgical training relied solely on time-based progression including a non-specific but holistic declaration of competency by program directors, but a global shift toward competency-based medical education (CBME) now prioritizes skill acquisition. For residents, current assessment integrates ACGME milestones, detailed ACGME case logs, and the CARSITE examination, moving beyond traditional subjective evaluations. Although initial board certification retains its written and oral examination structure, maintenance of certification for practicing surgeons has evolved to CertLink, a continuous longitudinal assessment. Furthermore, the advent of ASCRS U and on-demand content has revolutionized access to continuing medical education (CME). This manuscript provides a comprehensive overview of modern assessment strategies for both trainees and board-certified colorectal surgeons, detailing current practices, emerging trends, and future considerations. The commitment to robust assessment ensures high standards of practice and optimal patient outcomes in this dynamic specialty.
The growth of the digital age has corresponded with decreased operative experience and concern for low readiness for practice among surgical trainees, allowing for rapidly advancing technology to attempt to fill this educational need, often through independent study. Online platforms provide an accessible, convenient space for trainees and faculty to obtain e-Learning materials, including operative videos, educational podcasts, recorded lectures, and interactive content. Social media continues to grow as a space for dissemination of such materials and for live and ongoing discussion along the continuum from medical students to expert surgeons. More recently, artificial intelligence-based tools are being studied and implemented as methods for self-assessment for surgical trainees for clinical acumen, board examination preparation, and automated review of intraoperative video. Simulation remains an integral component of the independent development of technical skills with ongoing advancement in physical models and the integration of artificial intelligence and extended reality tools. Surgical education will continue to evolve and benefit from the integration of these technologies into traditional learning methods.
Rectal prolapse is the intussusception of the rectum, resulting in its full-thickness protrusion out of the anus. Approximately 0.5% of the general population is affected by this condition, with a higher occurrence in women and the elderly. While benign, rectal prolapse can be debilitating, as it can cause pain, bleeding, mucus discharge, and fecal incontinence. The earliest documented records of rectal prolapse date back to Ancient Egypt (1500-1200 BC), describing laxatives and topical therapeutics for the treatment of an anus turned inside out. Many techniques were devised to hold the reduced prolapsed rectum in place, ranging from cords to the use of bandages. Surgical cauterization of the anal sphincter to prevent recurrent prolapse began as early as 6 BC in India, a practice that continued into the 1800s. Advances in fundamental understanding of the anatomy of the colon, rectum, and anus in the 19th and early 20th centuries paved the way for modern surgical approaches. The 20th century saw the development of procedures proposed by surgeons such as Delorme, Moschowitz, Ripstein, and Altemeier. The emergence of the laparoscopic rectopexy in 1992 and subsequent use of robotic-assisted techniques in the early 2000s marked the transition to modern rectal prolapse surgeries, improving both precision and outcomes. The overall management of rectal prolapse has evolved significantly from ancient remedies to the surgeries known today, yet some fundamental similarities remain consistent between today's outlook and what was documented centuries ago. Understanding the historical evolution of the diagnosis and treatment of rectal prolapse provides insight into the contemporary management of the disorder.
Colorectal cancer incidence and mortality have declined over time, due in part to high-quality screening and surveillance colonoscopy. Nevertheless, postcolonoscopy colorectal cancer (PCCRC) occurs in up to 7% of cases and is inversely related to examination quality. Artificial intelligence-assisted colonoscopy aims to improve performance metrics and, ultimately, patient outcomes. Multiple randomized trials show that computer-aided polyp detection (CADe) increases adenoma detection, predominantly for diminutive lesions (≤5 mm). Computer-aided polyp characterization (CADx) enables real-time optical diagnosis, potentially shifting management of diminutive polyps by supporting resect-and-discard and diagnose-and-leave in situ strategies. Computer-aided quality assessment (CAQ) systems monitor key metrics-including cecal intubation rate, withdrawal time, speed, and mucosal exposure. Whether CADe alone leads to a reduction in PCCRC or cancer-related mortality remains to be determined; in the near term, a combined approach using CADe, CADx, and CAQ is most likely to deliver the greatest improvements in patient outcomes.
Artificial intelligence (AI) is poised to become transformational in many aspects of modern society and has attracted significant interest within the field of medicine. This review outlines the foundational components of modern AI such as including machine learning, deep learning, natural language processing, computer vision, and generative modeling, and examines their emerging applications within surgery. In the preoperative domain, AI-driven risk stratification models inform patient selection and resource allocation, while parallel advances in deep learning-enabled anatomic segmentation and three-dimensional reconstruction have the potential to streamline surgical planning by automating labor-intensive imaging workflows. Intraoperatively, maturing capabilities in phase recognition, anatomic identification, augmented reality overlay, and real-time decision support demonstrate the possibility for improved safety, workflow efficiency, and early recognition of surgical and physiologic challenges. And although the first fully autonomous AI-driven surgical robot in humans is likely still far off, the recent advances in robotic surgery suggest this may no longer be the purview of science fiction. For all its promise, significant challenges still persist for the robust implementation of AI into surgical workflows regarding data governance, algorithmic transparency, regulatory oversight, model generalizability, and, especially, many philosophical and ethical questions that remain unanswered.
Barriers to anal dyplasia screening are numerous and multifactorial. It is a rare disease, comprising only 2.5% of all gastrointestinal malignancies. Of the populations at high risk, men who have sex with other men (MSM), especially those living with HIV, represent the most studied high-risk population. Although knowledge of human papillomavirus (HPV) is relatively high in this population, its link to anal cancer and self-perception of risk of anal cancer in this population remains low. There are little, if any, data on provider knowledge or awareness of anal dysplasia or the high-risk populations that would benefit from screening. Given the recent advent of comprehensive anal cancer screening guidelines by the International Anal Neoplasia Society, this gap in knowledge is not suprising. Even among providers aware of these guidelines, other barriers exist, including possible discomfort discussing anal health, sexual practices, or discomfort in the performance of the screening test. Providers may also face challenges interpreting results of screening tests or finding specialists in high-resolution anoscopy for referral in the instance of a positive screening test. Finally, MSM and persons living with HIV have historically experienced stigmatization within the health care system. These historical underpinnings further complicate efforts to implement screening programs in these populations.
Sexual health is a key component of well-being and quality of life. Following colorectal surgery, many patients experience sexual dysfunction in the form of difficulty with libido, arousal, pain, and fertility. Autonomic nerve damage can explain many of these symptoms and may be a result of surgical trauma, tumor invasion, chemotherapy and radiotherapy, or pelvic inflammation and infection. An understanding of pelvic neuroanatomy can help prevent direct nerve damage during surgical dissection, but the effects of postsurgical inflammation and ischemia may be unavoidable. Despite the availability of tools for assessing patient-reported sexual health, data to support an informed consent discussion including risks to sexual health, and a multidisciplinary team of sexual health providers who can be involved in management, sexual dysfunction in colorectal surgery patients remains underdiagnosed and inadequately addressed in clinical practice. This represents a key area for quality improvement in delivering holistic and empathic care to the colorectal patient population.
The Accreditation Council for Graduate Medical Education (ACGME) recently conducted their 10-year specialty-specific revision of the colon and rectal surgery program requirements. During this process, leaders in our field began to hypothesize what training in our specialty may look like in the future. Here we identify the potential of competency-based assessment through entrustable professional activities (EPAs) in subspecialty surgical education. We also recognize new technology, including AI and machine learning, and its application to forward education in the ever-evolving field of surgery. The key to the future of colorectal residency is harnessing these advancements to create a residency driven by "precision education": a personalized and targeted education for the individual trainee. Given the growing sources of educational assessment tools and competency data, this opportunity will only exponentially increase over the coming years and provide a backbone for the development of training in the future.