Aim: Quantitative measurement of spinopelvic parameters from radiographs is important for assessing spinal disorders but is limited by the subjectivity and inefficiency of manual techniques. Deep learning may enable automated measurement with accuracy rivaling human readers.
Methods: PubMed, Embase, Scopus, and Cochrane databases were searched for relevant studies. Eligible studies were published in English, used deep learning for automated spinopelvic measurement from radiographs, and reported performance against human raters. Mean absolute errors and correlation coefficients were pooled in a meta-analysis.
Results: Fifteen studies analyzing over 10,000 radiographs met the inclusion criteria, employing convolutional neural networks (CNNs) and other deep learning architectures. Pooled mean absolute errors were 4.3° [95% confidence interval (CI) 3.2-5.4] for Cobb angle, 3.9° (95%CI 2.7-5.1) for thoracic kyphosis, 3.6° (95%CI 2.8-4.4) for lumbar lordosis, 1.9° (95%CI 1.3-2.5) for pelvic tilt (PT), 4.1° (95%CI 2.7-5.5) for pelvic incidence (PI), and
Conclusion: Deep learning demonstrates promising accuracy for automated spinopelvic measurement, potentially rivaling experienced human readers. However, further optimization and rigorous multicenter validation are required before clinical implementation. These technologies may eventually improve the efficiency and reliability of quantitative spine image analysis.
Artificial intelligence (AI) is a powerful computational tool that is being utilized more frequently in healthcare. AI holds promise within surgical practice, including application in the care of challenging patient populations. Complex spine reconstruction requires thorough multi-variable preoperative analysis and then the precise enactment of a surgical plan. Spino-plastics employs vascularized bone grafts (VBGs) to augment spinal fusion in these high-risk patients. In this article, we discuss the great breadth of AI and the tremendous potential for advancing the field of spino-plastics: surgical candidacy and patient selection, imaging and virtual surgical planning (VSP), intraoperative utilization, and future implementation.
Aim: This review evaluates the use of artificial intelligence (AI), machine learning (ML), and natural language processing (NLP) technologies for enhancing current procedural terminology (CPT) coding accuracy and efficiency in plastic and reconstructive surgery and related disciplines to define a precedent for future implementation.
Methods: A systematic search of PubMed, Scopus, and Web of Science Core Collection was performed to identify studies that leveraged artificially intelligent technologies in coding related to surgical procedures commonly managed by plastic and reconstructive surgeons.
Results: 11 peer-reviewed articles, which encompassed more than 1,000 CPT codes across numerous surgical subspecialties with overlap in plastic and reconstructive surgery and model systems, were included. The key findings highlight that AI-driven models demonstrate high sensitivity, specificity, area under the receiver operating curve (AUROC), and accuracy. While performance metrics varied considerably depending on the specific AI model employed, these systems were found to be effective assistive technologies for medical coding. Studies underscored the advantages of integration, maximizing billing workflow and reducing administrative workload. However, studies of AI performance for billing and medical coding within plastic surgery settings specifically were sparse. Notably, these investigations emphasized the need to tailor models for targeted suitability.
Conclusions: This review highlights the potential of AI technologies to improve CPT coding by enabling time and resource management and ultimately combatting the mounting presence of surgeon burnout. The sparsity of plastic surgery-specific literature on this emerging topic and untested promise in the specialty calls for intentional plastic surgeon-driven initiatives in the development of such applications.
The distribution pattern of esophageal cancer lymph node metastases and consequently the lymph node stations that should be resected are currently being investigated in the international distribution of lymph node metastases in esophageal carcinoma (TIGER) study. This observational study has no specific entry criteria, resulting in significant variation among participating centers regarding the extent of lymphatic tissue clearance during surgery. To ensure reliable interpretation of the TIGER study results, this study aims to develop an intraoperative metric (surgical quality assessment tool) to examine the extent of lymphadenectomy. This multicenter prospective study assesses the extent of lymphadenectomy during esophagectomies performed within the TIGER study. After consensus on the applicability of surgical quality assessment in observational studies (phase 1), a short photo/video of the thoracic, abdominal, and [if applicable] cervical area will be captured after lymphadenectomy (phase 2). Those images will be rated by expert surgeons using a structured assessment tool (phase 3), and an automatic artificial intelligence (AI)-based quality assessment will be developed (phase 4). The variability in lymphadenectomy will be adjusted when analyzing TIGER study outcomes. Technique standardization by surgical quality assurance (SQA) reduces variation in trial outcomes. This cannot be applied to the TIGER study because of its observational nature, although surgical heterogeneity might influence TIGER study outcomes. To monitor surgical performance within observational studies, surgical quality assessment can be applied, although this has not been done yet. This study will ensure a reliable interpretation of TIGER study results and could be used as a benchmark for quality assessment in future surgical studies.
Natural language processing (NLP) is the study of systems that allow machines to understand, interpret, and generate human language. With the advent of large language models (LLMs), non-technical industries can also harness the power of NLP. This includes healthcare, specifically surgical care and plastic surgery. This manuscript is an introductory review for plastic surgeons to understand the current state and future potential of NLP in patient consultations. The integration of NLP into plastic surgery patient consultations can transform both documentation and communication. These applications include information extraction, patient chart summarization, ambient transcription, coding, enhancing patient understanding, translation, and a patient-facing chatbot. We discuss the current progress toward building these applications and highlight their challenges. NLP has the potential to personalize care, enhance patient satisfaction, and improve workflows for plastic surgeons. Altogether, NLP can radically transform our current model of consultation into one that is more patient-centered.
Healthcare applications of artificial intelligence (AI) and machine learning (ML) are currently in a stage of exponential growth; however, their adoption into clinical practice across clinical specialties remains uneven. In spine surgery, the presence of challenging clinical problems, advanced intraoperative technologies, and large multi-center datasets positions the field well for the integration of these technologies into the clinic and operating room (OR). Here, we review recent advances in AI/ML applications in several key domains of spine surgery, identify methodological challenges shared by many approaches, and suggest solutions that may lead to these approaches becoming validated, commercialized tools that can reach clinical practice. Ultimately, we aim for this narrative review to help catalyze further progress in the development and commercialization of AI/ML to benefit future spine patients.
This editorial announces the need for a consensus conference on definitions of surgical nomenclature, which is evolving due to the introduction of non-human (hardware, software) devices and applications, including robotics. A recently created entity, the Artificial Intelligence Organization for Next Generation Surgeons (AIONS), comprised primarily of AIS editorial board members, has proposed updated definitions on the following terms: surgery, endoluminal surgery, percutaneous surgery, robots, robotic-assisted surgery (RAS), remote surgery, artificial intelligence surgery (AIS), robotic surgery, surgomics, non-invasive surgery. These definitions will be discussed during a Consensus Conference on Definitions of Artificial Intelligence, Surgery, Surgomics and Robotics, which is scheduled for February 12, 2025.
From diagnostics and treatments to surgical techniques and postoperative outcomes, the field of spine surgery is advancing at a historically unprecedented rate. Given the widespread integration of artificial intelligence (AI) in various industries, its implementation in the medical field is not a question of if, but when it will happen. AI’s ability to sort, analyze, and summarize vast quantities of data demonstrates great potential in assisting surgical professionals in all levels of training. Virtual reality (VR) enables users to explore and interact in a three-dimensional, computer-generated environment, and its application in the field of spine surgery can include bringing awareness and exposure of the field to medical students, surgical training and repetition of residents and fellows, and surgical planning for attendings. Augmented reality (AR) has significant potential through its versatile applications, offering benefits in medical education and training. While there are costs associated with the implementation of AI and VR in training curriculums for spine professionals, the long-term benefits and savings to various stakeholders outweigh the initial investment. This paper intends to offer a focused summary of the impact of AI and VR tools in spine education and training.
Aim: A growing body of literature reports on prediction models for patient-reported outcomes of spine surgery, carrying broad implications for use in value-based care and decision making. This review assesses the performance and transparency of reporting of these models.
Methods: We queried four studies reporting the development and/or validation of prediction models for patient-reported outcome measures (PROMs) following elective spine surgery with performance metrics such as the area under the receiver operating curve (AUC) scores. Adherence to transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD-AI) guidelines was assessed. One representative model was selected from each study.
Results: Of 4,471 screened studies, 35 were included, with nine development, 24 development and evaluation, and two evaluation studies. Sixteen machine learning models and 19 traditional prediction models were represented. Oswestry disability index (ODI) and modified Japanese Orthopaedic Association (mJOA) scores were most commonly used. Among 29 categorical outcome prediction models, the median [interquartile range (IQR)] AUC was 0.79 [0.73, 0.84]. The median [IQR] AUC was 0.825 [0.76, 0.84] among machine learning models and 0.74 [0.71, 0.81] among traditional models. Adherence to TRIPOD-AI guidelines was inconsistent, with no studies commenting on healthcare inequalities in the sample population, model fairness, or disclosure of study protocols or registration.
Conclusion: We found considerable variation between studies, not only in chosen patient populations and outcome measures, but also in their manner of evaluation and reporting. Agreement about outcome definitions, more frequent external validation, and improved completeness of reporting may facilitate the effective use and interpretation of these models.
Aim: This paper aims to advance autonomous surgical operations through imitation learning from video demonstrations.
Methods: To address this objective, we propose two main contributions: (1) We introduce a new dataset of virtual kidney tumor environments to train our model on. The dataset is composed of video demonstrations of tumor removal from the kidney, executed in a virtual environment, and kinematic data of the robot tools; (2) We employed an imitation learning architecture composed of vision transformers (ViT) to handle the frames extracted from the videos and of a long short-term memory (LSTM) structure to process surgical motion sequences with a sliding window mechanism. This model processes video frames and prior poses to predict the poses for both robotic arms. A self-generating sequence approach was implemented, where each predicted pose served as the latest element in the sequence, subsequently used as input for the next prediction together with the current frame of the video. The choice of architecture and methodology was guided by the need to effectively model the sequential nature of surgical operations.
Results: The model achieved promising results, exhibiting an average position error of 0.5 cm. The model was able to execute correctly 70% of the test tasks. This highlights the sequence-based approach’s efficacy in capturing and predicting surgical trajectories.
Conclusion: Our study supports imitation learning’s viability for acquiring task execution policies in surgical robotics. The sequence-based model, combining ViT and LSTM architectures, successfully handles surgical trajectories.
Aim: Large language models (LLMs) like ChatGPT have transformed access to health information. For transgender individuals considering gender-affirming surgery (GAS), accurate and reliable information is essential for informed decision making. This study aimed to quantitatively assess the use of ChatGPT among individuals considering GAS and its impact on their decision-making process.
Methods: A cross-sectional survey was conducted in January 2024 on Prolific. Participants included English-speaking U.S. users over 18 whose current gender differed from their assigned gender at birth. The survey collected demographic information, evaluated interest in GAS, and examined interactions with ChatGPT. Descriptive statistics were used for analysis.
Results: The study included 207 participants (average age 30.2 years), primarily identifying as non-binary (40.6%), transgender men (29.5%), and transgender women (13%). Most expressed interest in GAS (89%). Primary information sources for GAS were online forums (24.6%), medical websites (21.3%), and social media (17.4%). While many had used ChatGPT (73%), few utilized it for GAS information (6.7%). Among those who did, the majority (70%) rated its usefulness as moderate to slight, with some reporting a positive influence on their decision making (40%). Trust in ChatGPT’s information was moderate to highly rated by 80% of participants.
Conclusions: In our cohort, ChatGPT is less commonly used for GAS information than online forums and medical websites. This suggests that patients prefer platforms that offer visual content, human interaction, and relatability. These findings highlight the importance of guiding patients toward reliable health information sources, such as healthcare providers, reputable medical websites, and academic literature, to support informed decision making.
As artificial intelligence (AI) technologies evolve in sophistication, they offer the potential to benefit various aspects of plastic and reconstructive surgery practice. From enhancing surgical precision within the operating room to streamlining administrative tasks and supporting the diagnosis and treatment of patients, AI may grow into an invaluable tool that redefines standards of care within plastic surgery. Given the nascent and largely theoretical role of AI in plastic surgery, numerous questions arise regarding its safety, actual utility, ethical considerations, and policies needed to regulate its use. This manuscript aims to provide commentary on AI in healthcare and to discuss an alternative viewpoint of its use in plastic surgery. Americans remain hesitant about healthcare providers leveraging AI in their care. Ongoing scrutiny is required to protect patients from unintended sequelae, safeguard their privacy, mitigate bias, and reduce harm. Early legislation by the United States federal government has aimed to define a role for AI in healthcare, yet more explicit guidance is required. Uncertainty in medico-legal implications begs the question of where liability would fall if AI use causes adverse outcomes. If applied appropriately, AI may ultimately improve patient outcomes and satisfaction with their plastic surgery care. With less energy dedicated toward automatable tasks and tools that push the envelope of human performance, plastic surgeons may be better equipped to care for their patients. We advocate for a cautiously optimistic approach to AI’s incorporation within plastic and reconstructive surgery.
Reliable planning, execution, and postoperative monitoring in microvascular free flap reconstruction are essential to optimize clinical outcomes. Artificial intelligence has demonstrated value in several applications to clinical medicine and surgery, including image analysis and simulation, outcomes modeling, and evaluation of large datasets. Within microvascular reconstruction, artificial intelligence has been increasingly applied to preoperative planning, intraoperative decision making, and postoperative monitoring. The present paper aims to review salient applications to each. The authors conclude by suggesting areas suitable for future analysis.
Clinical integration of artificial intelligence (AI) in spinal surgery is still in its early stages, with spinal imaging being the most prominent. We present a review of recent literature on the topic. The reporting of traditional spinal imaging has been slow due to overburdened staff and unreliable in some patients. AI applications have shown promising results in improving the speed and quality of imaging while reducing costs and radiation exposure. Specific examples of clinical implementation include osteoporosis screening, diagnosing degenerative spine diseases and differentiating tuberculous and pyogenic spondylitis, helping in preoperative measurements and surgical planning. Other tools have demonstrated the ability to help clinicians in real time to reduce rates of missed fractures and to rule out cord impingement in emergency settings. Novel variants of magnetic resonance imaging (MRI) and synthetic computed tomography (sCT) scans, without ionizing radiation, have been successful in reducing the resource burden and scan time, while maintaining clinical utility. At its current stage, AI has the potential to improve significantly and is expected to tremendously enhance the efficiency and accuracy of radiologists and spine care providers. However, clinical validation studies are still required before the widespread integration of AI in direct patient care.
Breast reconstruction is a critical component of breast cancer treatment. With the rapid integration of Artificial Intelligence (AI) into healthcare, its potential to revolutionize breast reconstruction has become increasingly evident. This narrative review examines the latest AI developments across the preoperative, intraoperative, and postoperative phases of breast reconstruction. In preoperative consultations, AI and augmented reality (AR)-driven simulations help both the surgeons and the patients visualize reconstruction outcomes. Imaging analysis and predictive modeling enhance the precision and efficiency of autologous procedures such as deep inferior epigastric artery perforator flap-based reconstruction. Within the operating room, AI applications such as real-time perforator mapping and AR modeling offer plastic surgeons improved control and visualization, which helps to reduce postoperative complications. Furthermore, AI models help surgeons design and deliver more personalized and value-based postoperative care, thereby improving patient satisfaction and overall cost-effectiveness. While AI applications demonstrate promising utility, challenges such as high costs, reliability, and the need for extensive clinical validation remain. Ongoing research and large-scale clinical trials are crucial to fully harness AI’s potential in improving breast reconstruction outcomes.