Search

Artificial intelligence (AI) focuses on processing and interpreting complex information as well as identifying relationships and patterns among complex data. Artificial intelligence- and machine learning (ML)-driven predictions have shown promising potential in influencing real-time decisions and improving surgical outcomes by facilitating screening, diagnosis, risk assessment, preoperative planning, and shared decision-making. Fundamental understanding of the algorithms, as well as their development and interpretation, is essential for the evolution of AI in surgery. In this article, we provide surgeons with a fundamental understanding of AI-driven predictive models through an overview of common ML and deep learning algorithms, model development, performance metrics and interpretation. This would serve as a basis for understanding ML-based research, while fostering new ideas and innovations for furthering the reach of this emerging discipline.

Machine learning systems have become integrated into some of the most vital decision-making aspects of humanity, including hiring decisions, loan applications, and automobile safety, to name just a few. As applications increase in both gravity and complexity, the data quality and algorithmic interpretability of the systems must rise to meet those challenges. This is especially vital for navigating the nuances of health care, particularly among the high stakes of surgical operations. In addition to inherent ethical challenges of enabling a “black box” system to influence decision-making in patient care, the creation of biased datasets leads to biased algorithms with the power to perpetuate discrimination and reinforce disparities. Transparency and responsibility are paramount to the implementation of artificial intelligence in surgical decision-making and autonomous robotic surgery. Machine learning has been permeating health care across diverse clinical and surgical contexts but continues to face sizable obstacles, including apprehension from patients and providers alike. To integrate the technology fully while upholding standard of care and patient-provider trust, one must acknowledge and address the ethical, financial, and legal implications of using artificial intelligence for patient care.

Background: The updated Lymphedema Life Impact Scale (LLIS, version 2) has been widely used to evaluate the effect of lymphedema from the patient's perspective. We sought to assess its ability to accurately and efficiently measure lymphedema-related impact using modern psychometric techniques. Methods and Results: We collected a total of 1054 patient-reported outcome measure scores from 285 patients with upper extremity lymphedema and 65 patients with lower extremity lymphedema between 2016 and 2020. We first evaluated the relationship between the LLIS score, L-Dex score, and limb volume difference (LVD), and used classical test and item response theories to assess its psychometric performance. The LLIS score was only very weakly associated with LVD (r = 0.17, p

Background ChatGPT is an artificial intelligence language model developed and released by OpenAI (San Francisco, CA) in late 2022. Objectives The aim of this study was to evaluate the performance of ChatGPT on the Plastic Surgery In-Service Examination and to compare it to residents’ performance nationally. Methods The Plastic Surgery In-Service Examinations from 2018 to 2022 were used as a question source. For each question, the stem and all multiple-choice options were imported into ChatGPT. The 2022 examination was used to compare the performance of ChatGPT to plastic surgery residents nationally. Results In total, 1129 questions were included in the final analysis and ChatGPT answered 630 (55.8%) of these correctly. ChatGPT scored the highest on the 2021 exam (60.1%) and on the comprehensive section (58.7%). There were no significant differences regarding questions answered correctly among exam years or among the different exam sections. ChatGPT answered 57% of questions correctly on the 2022 exam. When compared to the performance of plastic surgery residents in 2022, ChatGPT would rank in the 49th percentile for first-year integrated plastic surgery residents, 13th percentile for second-year residents, 5th percentile for third- and fourth-year residents, and 0th percentile for fifth- and sixth-year residents. Conclusions ChatGPT performs at the level of a first-year resident on the Plastic Surgery In-Service Examination. However, it performed poorly when compared with residents in more advanced years of training. Although ChatGPT has many undeniable benefits and potential uses in the field of healthcare and medical education, it will require additional research to assess its efficacy.

The technology of artificial intelligence (AI) has made significant in-roads into the field of medicine over the last decade. With surgery being a discipline where repetition is the key to mastery, the scope of AI presents enormous potential for resident education through the analysis of technique and delivery of structured feedback for performance improvement. In an era marred by a raging pandemic that has decreased exposure and opportunity, AI offers an attractive solution towards improving operating room efficiency, safe patient care in the hands of supervised residents and can ultimately culminate in reduced health care costs. Through this article, we elucidate the current adoption of the artificial intelligence technology and its prospects for advancing surgical education.

Deep learning (DL) is a subset of machine learning that is rapidly gaining traction in surgical fields. Its tremendous capacity for powerful data-driven problem-solving has generated computational breakthroughs in many realms, with the fields of medicine and surgery becoming increasingly prominent avenues. Through its multi-layer architecture of interconnected neural networks, DL enables feature extraction and pattern recognition of highly complex and large-volume data. Across various surgical specialties, DL is being applied to optimize both preoperative planning and intraoperative performance in new and innovative ways. Surgeons are now able to integrate deep learning tools into their practice to improve patient safety and outcomes. Through this review, we explore the applications of deep learning in surgery and related subspecialties with an aim to shed light on the practical utilization of this technology in the present and near future.

Predatory journals have exploited the open access publishing model and are considered as a major threat to the integrity of scientific research. The goal of this study was to characterize predatory publishing practices in plastic surgery.To identify potentially predatory journals in the field of plastic surgery, the authors searched the Cabells' Predatory Reports and Beall's List using preidentified keywords. For presumed legitimate open access journals, the Directory of Open Access Journals (DOAJ) was queried. The characteristics of potentially predatory journals were compared to those of legitimate open access plastic surgery journals.The authors identified a total of 25 plastic surgery-focused journals. Out of the 25 potentially predatory journals, only 15 journals had articles published within the last 5 years, with a mean number of articles of 33 ± 39 (range, 2 to 159 articles). The mean number of predatory violations according to Cabells' criteria was 6.8 ± 1.4 (range, 3 to 9). Using the DOAJ database, the authors identified a total of 24 plastic surgery-related journals. Compared to potentially predatory journals, journals from the DOAJ were more likely to be indexed in PubMed (0 versus 50 percent, respectively, p0.0001). Time to publication was significantly higher in journals from the DOAJ (17 ± 7 versus 4 ± 1 weeks; p = 0.006). Despite higher article processing charges in the DOAJ group, this difference was not statically significant ($1425 ± $717 versus $1071 ± $1060; p = 0.13).Predatory journals are pervasive in the medical literature and plastic surgery is no exception. Plastic surgeons should practice due diligence when choosing a target journal for their articles. Journals with predatory practices should be distinguished from legitimate open access publication platforms.

Implant-based breast reconstruction infections often require implant explantation. Whereas some plastic surgeons pursue autologous reconstruction following the first implant-based breast reconstruction failure caused by infection, others argue that a second attempt is acceptable.The authors conducted a retrospective study of patients who underwent a second reconstruction attempt with implant-based or free flap breast reconstruction following explantation because of infection between 2006 and 2019. Surgical and patient-reported outcomes were compared between the two groups.A total of 6093 implant-based breast reconstructions were performed during the study period, of which 130 breasts met our inclusion criteria [implant-based, n = 86 (66 percent); free flap, n = 44 (34 percent)]. No significant differences in rates of overall (25 percent versus 36 percent; p = 0.2) or major (20 percent versus 21 percent; p = 0.95) complications were identified between the free flap and implant-based cohorts, respectively. Implant-based breast reconstruction patients were more likely to experience a second infection (27 percent versus 2 percent; p = 0.0007) and reconstruction failure (21 percent versus 5 percent; p = 0.019). Among irradiated patients, reconstruction failure was reported in 44 percent of the implant-based and 7 percent of the free flap cohorts (p = 0.02). Free flap patients reported significantly higher scores for Satisfaction with Breasts (73.7 ± 20.1 versus 48.5 ± 27.9; p = 0.0046).Following implant-based breast reconstruction explantation because of infection, implant-based and free flap breast reconstruction had similar rates of overall and major complications; however, implant-based breast reconstruction had considerably higher rates of infection and reconstructive failures and lower patient-reported scores for Satisfaction with Breasts. Given the high rates of implant-based breast reconstruction failure in patients with prior radiotherapy and infection-based failure, plastic surgeons should strongly consider autologous reconstruction in this patient population.Therapeutic, III.

There are no U.S. Food and Drug Administration (FDA)-approved indications for robotic plastic surgery. This study was designed to demonstrate safety and efficacy of the robotic latissimus dorsi harvest for FDA approval.This prospective, single-arm study was conducted under an investigational device exemption through the FDA for obtaining 510(k) approval. The primary safety endpoint was adverse events attributable to harvest. Primary efficacy endpoints included muscle viability after harvest and conversion to open technique. Secondary endpoints included postoperative pain and upper extremity function.Fifteen patients enrolled, with a mean age of 50 ± 10 years and a mean body mass index of 25 ± 4 kg/m2. There were no adverse events, all muscles were viable after harvest, and there were zero conversions to open procedures. Visual analogue scale scores for pain progressively decreased postoperatively and returned to baseline, indicating resolution of postoperative pain. Physical therapy assessment demonstrated recovery of function in all range-of-movement domains by 36 weeks. There was no difference in function of the operated extremity between baseline and 36 weeks' follow-up for Quick Disabilities of the Arm, Shoulder and Hand (QuickDASH) questionnaire (p = 0.87); QuickDASH Work Module (p0.05); and QuickDASH Sports Module (p0.05).The use of the da Vinci Robotic Surgical System is safe based on zero adverse events attributable to harvest and efficacious based on 100 percent muscle viability after harvest and zero conversions to open technique. There appears to be little to no long-term functional deficit or pain from muscle harvest. Given these results and their own prestudy guidelines, the robotic latissimus dorsi qualifies for 510(k) submission by Intuitive Surgical and approval by the U.S. FDA.Therapeutic, IV.

Osseous reconstruction following total spondylectomy/vertebrectomy in the thoracic spine is indicated to restore spinal stability. To assist with bony reconstruction, vascularized bone flaps including free vascularized fibula flaps and pedicled rib flaps can be used. However, there are limited data comparing various techniques. The authors aimed to evaluate the outcomes of free vascularized fibula flaps and pedicled rib flaps used for thoracic spinal reconstruction.The authors reviewed 44 vascularized bone flaps [10 anterior pedicled rib flaps, 25 posterior pedicled rib flaps, and nine vascularized fibula flaps] for corpectomy defects spanning T1 to L1 between January of 1999 and June of 2018. Mean age and follow-up were 46 ±17 years and 74 ± 52 months, respectively.The union rate was 93 percent, with a similar mean time to union among the three groups: free vascularized fibula flaps, 9 ± 4 months; anterior pedicled rib flaps, 9 ± 6 months; and posterior pedicled rib flaps, 9 ± 5 months (p = 0.95). Surgical-site complications were found in 27 vascularized bone flaps (61 percent), and reoperations and revisions were performed in 14 (32 percent) and 10 (23 percent) vascularized bone flaps, respectively. No differences were identified among anterior pedicled rib flaps, posterior pedicled rib flaps, and vascularized fibula flaps with regard to complication, reoperation, and revision rates.Free vascularized fibula flaps and pedicled rib flaps provide durable reconstruction for thoracic spinal defects. Union, time to union, revision, and reoperation rates were similar among anterior and posterior pedicled rib flaps and vascularized fibula flaps. Given the variability in indication and defect types among the flap cohorts, the authors' results should be interpreted carefully as an insight into the outcomes of different vascularized bone flaps for the unique cases of thoracic spinal reconstruction rather than to compare the different flaps used.Therapeutic, III.

This study evaluates clinical outcomes of vascularized lymph node transplantation (VLNT) from the lateral thoracic region and technical modifications.Consecutive patients that underwent lateral thoracic VLNT to treat extremity lymphedema were included. Demographic and treatment data were recorded, and outcomes data including limb volume, LDex score, and Lymphedema Life Impact Scale (LLIS), QuickDASH, and LEFS questionnaires, were collected prospectively. Consecutive patients that underwent single-photon emission computed tomography (SPECT/CT) lymphoscintigraphy axillary reverse lymphatic mapping (RLM) were analyzed to characterize the physiological drainage of the normal upper extremity.A consecutive series of 32 flaps were included. At 24 months postoperatively mean reduction in limb volume excess was 47.2% (±11.6; p = 0.0085), LDex score was 63.1% (±8.5; p 0.001), and LLIS score was 65.1% (±7.4; p 0.001). Preoperatively 14/31 patients (45.2%) reported cellulitis, and postoperatively there were no episodes at up to 24 months (p 0.001). No patient developed donor extremity lymphedema at mean 18.6 (±8.3) months follow-up. SPECT/CT-RLM of 182 normal axillae demonstrated that the sentinel lymph node(s) of the upper extremity was consistently anatomically located in the upper outer quadrant of the axilla (97%).VLNT from the lateral thoracic region is effective and versatile for the treatment of lymphedema with a low donor site complication rate.

PURPOSE Financial burden caused by cancer treatment is associated with material loss, distress, and poorer outcomes. Financial resources exist to support patients but identification of need is difficult. We sought to develop and test a tool to accurately predict an individual's risk of financial toxicity based on clinical, demographic, and patient-reported data prior to initiation of breast cancer treatment. PATIENTS AND METHODS We surveyed 611 patients undergoing breast cancer therapy at MD Anderson Cancer Center. We collected data using the validated COmprehensive Score for financial Toxicity (COST) patient-reported outcome measure alongside other financial indicators (credit score, income, and insurance status). We also collected clinical and perioperative data. We trained and tested an ensemble of machine learning (ML) algorithms (neural network, regularized linear model, support vector machines, and a classification tree) to predict financial toxicity. Data were randomly partitioned into training and test samples (2:1 ratio). Predictive performance was assessed using area-under-the-receiver-operating-characteristics-curve (AUROC), accuracy, sensitivity, and specificity. RESULTS In our test sample (N = 203), 48 of 203 women (23.6%) reported significant financial burden. The algorithm ensemble performed well to predict financial burden with an AUROC of 0.85, accuracy of 0.82, sensitivity of 0.85, and specificity of 0.81. Key clinical predictors of financial burden from the linear model were neoadjuvant therapy (βregularized, .11) and autologous, rather than implant-based, reconstruction (βregularized, .06). Notably, radiation and clinical tumor stage had no effect on financial burden. CONCLUSION ML models accurately predicted financial toxicity related to breast cancer treatment. These predictions may inform decision making and care planning to avoid financial distress during cancer treatment or enable targeted financial support. Further research is warranted to validate this tool and assess applicability for other types of cancer.

Background Patients with contaminated/dirty-infected defects are at high risk for postoperative complications after abdominal wall reconstruction (AWR). We evaluated outcomes of AWR using acellular dermal matrix (ADM) for mesh reinforcement and identified predictors of hernia recurrence (HR), surgical site occurrences (SSOs), and surgical site infections (SSIs). Study Design We conducted a retrospective cohort study of patients who underwent AWR using ADM, from March 2005 to June 2019. Outcomes were compared between Centers for Disease Control and Prevention (CDC) wound classifications. The primary outcome measure was HR. Secondary outcomes were SSOs and SSIs. Results We identified 725 AWRs using ADM that met the study criteria. Participants had a mean age of 60 ± 11.5 years, mean BMI of 31 ± 7 kg/m2, and mean follow-up time of 42 ± 29 months. Three hundred two patients (41.6%) had clean defects, 322 patients (44.4%) had clean-contaminated defects, and 101 patients (13.9%) had contaminated/dirty-infected defects. Patients with contaminated/dirty-infected defects had an HR rate of 20.8%, SSO rate of 54.5%, and SSI rate of 23.8%. Multivariate logistic regression found that contaminated/dirty-infected defects were independent predictors of SSOs (OR 2.99; 95% CI 1.72–5.18; p Conclusions SSIs and SSOs increase as contamination levels rise, but the risk of HR does not. AWR with ADM provides safe and durable outcomes, even with increasing levels of contamination.