ENTAgents: AI Agents for Complex Knowledge Otolaryngology

ENTAgents: AI Agents for Complex Knowledge Otolaryngology | medRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-P4HH5NV'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search ENTAgents: AI Agents for Complex Knowledge Otolaryngology Tsz Kin Chan , View ORCID Profile Ngoc-Duy Dinh doi: https://doi.org/10.1101/2025.01.01.25319863 Tsz Kin Chan 1 Department of Biomedical Engineering, The Chinese University of Hong Kong , Shatin, N.T., Hong Kong Find this author on Google Scholar Find this author on PubMed Search for this author on this site Ngoc-Duy Dinh 1 Department of Biomedical Engineering, The Chinese University of Hong Kong , Shatin, N.T., Hong Kong Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Ngoc-Duy Dinh For correspondence: ngocduydinh{at}cuhk.edu.hk dinhngocduy{at}u.nus.edu Abstract Full Text Info/History Metrics Data/Code Preview PDF Abstract Various healthcare applications based on large language models (LLMs) have emerged as LLMs show improved efficiency and error reduction. Recently, retrieval augmented generation (RAG) has been adopted frequently for LLM applications to solve the problem of hallucinations. Despite the success of RAG, it has its drawbacks, including incomplete semantic meanings, and large-scale dataset requirements. AI Agents have shown great potential in medicine and healthcare applications by leveraging their rich background knowledge and reasoning capabilities. In this paper, we introduce ENTAgents, a framework that utilizes both the RAG and multi-agent systems. To achieve better decision-making and enhanced explainability, we use the verbal reinforcement learning framework, Reflexion, as a reference for the task planning of ENTAgents. With the verbal feedback received from the last agent node, the agentic system can decide the best agent to utilize to improve the response of ENTAgents. We tested ENTAgents with various types of questions, including short questions, essay questions, and multiple-choice questions. ENTAgents improve accuracy by over 11.3% in handling short questions, with 2.78 folds in the length of the text and explaining clearly. ENTAgents shows higher accuracy and is more comprehensive in its answer compared to other LLM models. ENTAgents also demonstrates its ability to refine its answer with additional information to make it more thorough and educational. It also presents the capability to change its response to the correct answer in multiple-choice questions according to its self-reflection. Overall, ENTAgents is easy to use, can self-correct, and can provide detailed information in complex knowledge to users. We look forward to integrating ENTAgents into various healthcare scenarios, such as medical education and clinical decision support, particularly otolaryngology. 1. Introduction Large language models (LLMs) have been in the spotlight in different industries since the release of ChatGPT, with the healthcare sector being one of the potential industries that can unlock the power of LLMs and new opportunities in medical applications. Numerous utilizations of LLMs have been proposed in the healthcare system, including health record management [ 1 – 2 ], medical image analysis [ 3 – 4 ], and medical research [ 5 – 6 ]. These applications have made significant impacts on the healthcare systems by minimizing human errors, enhancing efficiency, and providing medical research support [ 6 – 7 ]. Nevertheless, there are still concerns in LLM-based medical applications. Since LLMs often generate hallucinations, which is a response that is either factually incorrect or nonsensical. The hallucination rate of top LLM models is from 1.3% to 4.3% [ 8 ]. Although these values are not significant, concerns could be raised when an inaccurate response is generated by the LLM application, especially when it comes to clinical decision support, as reliability and trustfulness of a system are the most important principles in the healthcare sector. To eliminate hallucinations in LLM, several techniques have been proposed in which the retrieval augmented generation (RAG) is one of the easiest and most cost-effective methods to deal with hallucinations. RAG is a technique that provides related accurate data from the external knowledge base to the LLMs. With the advancements of RAG techniques, RAG has shown promising results in alleviating hallucinations and a wide range of applications in healthcare [ 9 – 14 ]. Long et al introduced ChatENT, which is an AI framework specialized in otolaryngology, with the use of ChatGPT 4.0 and Retrieval-Augmented Language Modeling (RALM) [ 14 ], demonstrated a good capability of answering otolaryngology questions. Despite the success of the RAG systems, information loss, the need of large-scale dataset and time-consuming are the issues when building a RAG system. Often RAG fails to deliver the complete significant information to the LLM as the context is separated by the chunk size but not the contextual meaning, and a large volume of manipulated dataset is required to cover all the knowledge in one specialty [ 14 , 15 ]. It has evolved very fast in the medical field, it is not feasible to update the knowledge database frequently for RAG, especially for comprehensive healthcare applications. Therefore, there is a need to produce an LLM system that can be reliable and handle different tasks in the healthcare industry. Agentic systems have emerged as a popular approach in different applications due to their scalability and robustness, and these advantages are crucial to address the dynamic healthcare environments. Several agentic systems have been proposed to the healthcare industry, offering medical knowledge, diagnosis, clinical record management and more [ 16 – 20 ]. Nevertheless, the existing agentic systems use information from their database or repository to handle clinical diagnosis and treatment. They lack the ability to generate insights into specialized medications and unseen diseases [ 17 , 21 – 23 ]. In addition, the current agentic system is not capable of examining their answer or understanding the question, including irrelevant contexts as a result [ 24 ]. Here, we introduce an innovative LLM agentic framework, ENTAgents, which is a system addressing the medical specialty in otolaryngology and utilizing RAG and multiple agents to optimize its response. We take Reflexion, the verbal reinforcement learning framework, as a reference for the task planning of our agents [ 25 ]. ENTAgents demonstrate the ability to self-correction their answer based on the question, with the aid of information found by the agents. We show that ENTAgents can utilize different agents successfully and perform excellently in different types of questions. It is a cost-efficient, ease-of-use, time-saving approach to handling medical specialty questions without any need for fine-tuning and has great potential for medical education and clinical support for all medical specialties. 2. Methods 2.1 Overview Fig. 1 shows the pipeline of our proposed ENTAgents framework. The first node, “Drafter”, generates an initial answer based on the question by LLM. After that, the LLM model examines its answer and generates critiques in our designated format for reflection, followed by the search queries according to the reflection critiques. Subsequently, RAG is performed to extract relevant information from our knowledge base to search queries in the second agent node. For the third node, we call a “Revisor,” which revises the drafted answer based on the reflection critiques and extracted information. Then, the LLM will provide new reflection critiques and search queries if they decide the revised answer is not the best. The “Supervisor” node will examine the question and the revised answer to choose the best agent to acquire new information or finish the execution. The agent searches for information or data with their tool and summarizes it. They are then passed back to the Revisor, and hence, a loop is created until the Supervisor decides that the answer does not need to be adjusted. Llama 3.1 70B model from Meta is used for the Drafter and Revisor, and Llama-3-Groq-70b-Tool-Use from Groq is used for the Supervisor and the agents, built for tool use and function calling. ENTAgent is built on the LangGraph platform. Download figure Open in new tab Figure 1: ENTAgents flow diagram from question input to answer generation. 2.2 RAG To perform RAG, a knowledge base is first built with a variety of sources, including clinical guidelines, textbooks, and published articles, being chunked with the size of 1024 and chunk overlap of 200, including the whole knowledge base as a part of query extends the latency enormously and reaches the context window of Llama model easily. We used Voyage AI’s voyage-large-2-instruct to embed the chunked documents as vectors that represent their semantic meanings, and they were stored in a vector store initiated by the Facebook AI Similarity Search (FAISS) library. Having received the search queries, the embedding model also converts them into embedding vectors. For each of the queries, six chunks from the knowledge base are selected with the highest similarity between their embedding vectors and search queries’ embedding vectors. 2.3 Task Planning: Reflexion We adopted the Reflexion framework proposed by the reinforcement learning proposed by Shinn et al [ 25 ] as the reinforcement learning of our ENTAgents. By leveraging Reflexion, AgenticENT can reflect on their answer and provide verbal feedback, which is stored in the episodic memory and added to the message passed to the next episode. The critiques of what is missing and what is superfluous in the answer were used in the self-reflection for short questions and essay questions, while critiques of inaccuracy and unnecessary were used for multiple choice questions. The Revisor reduces the redundant information in the last answer based on the critique of unnecessary and adds additional information provided by RAG or agents or changes the chosen answer according to the critique of missing and inaccuracy, respectively. 2.4 Multi-Agent System We have three agents that use different tools to search for information from three tools: Google Search, ArXiv, and Pubmed. ArXiv tool searches for relevant articles from its database, which contains preprint and postprint scholarly articles, while the Pubmed tool finds published biomedical literature from the Pubmed database. Google Search tool uses the search engine to find the most relevant information from the website to the query. When the Supervisor decides the best agent to act on next, the search query is passed to that agent. The agent then executes the search query with its tool and concludes what it has found and how it can address the search query in a summary. Finally, the summary is sent to the revisor for revision again. 3. Results The effectiveness and the capability of ENTAgents were evaluated using various types of questions, which are short answer questions, essay questions, and multiple-choice questions. 3.1 Short Answer Questions 19 short answer questions were used to assess the response of ENTAgents in otolaryngology, which includes 11 sample questions from the Royal College of Physicians and Surgeons of Canada (RCPSC). The accuracy of ENTAgents was evaluated with these 11 questions. ENTAgents based on Llama 3.1 scored 30.5 out of 37.5, a correct percentage of 81.3%, which is significantly higher than Llama 3.1 without Agentic, which is 70% ( Figure 2 ). We can see the correct percentage of ENTAgents based on Llama 3.1 is a little higher than the paid model ChatGPT 4.0, which is 79.3%, but there are significant differences between ChatGPT 4.0 (ChatENT based on ChatGPT 4.0) and ENTAgents. The responses from ChatENT and ENTAgents were compared, as shown in Table 1 . As ChatENT is not an open-source model, the responses from ChatENT were extracted from the supplementary information of the original paper. The average length of ENTAgents’s responses was approximately 258.3 and that of ChatENT’s responses was 92.5. The average latency of ENTAgents was 118.7 seconds. From Table 1 , it can be observed that ENTAgents provided a more thorough answer than ChatENT did. The ChatENT failed to identify the number of sub-questions in this question, and it was only able to answer the second part of the question, while our ENTAgents provided complete answers for two parts of the question. For the first part of the question, both correct answers and explanations were provided by ENTAgents, making the answer understandable to the user. Both ENTAgents and ChatENT were able to generate three techniques required from the question and their functions were also provided. However, ENTAgents generated additional information related to the question and its answer. For instance, ENTAgents gave an example of intraoperative imaging and its functions during surgery so that the third technique could be explained clearly. Extra information related to preventing complications and the advantages of osteoplastic flap procedures was also provided. Download figure Open in new tab Figure 2: Accuracy rate of various LLM models for the short questions from the Royal College of Physicians and Surgeons of Canada (RCPSC). View this table: View inline View popup Table 1: Comparison of the responses from ChatENT and ENTAgents to a short answer question. The italics paragraphs are the relevant additional information generated by the ENTAgents besides answering the question. 3.2 Essay Questions The ability to understand and generate subject matter content was also examined with 24 essay questions. The result shows that ENTAgents generated a significant number of relevant words compared to other LLM models ( Figure 3 ). The ENTAgents generated answers with an average of 425.08 words, and the average latency ENTAgents required was 118.67 seconds. An average of 2.67 agents was invoked for ENTAgents to handle an essay question. Download figure Open in new tab Figure 3: Length of responses from different LLMs to the essay questions. The workflow of the ENTAgents is demonstrated in Table 2 . In the initial response, ENTAgents reflected on its answer and stated that the information on aetiopathogenesis was insufficient and that the answer should be more concise and remove repetitive language. The search query was also generated based on the critiques of what is missing. After relevant information was found with the search query, the Revisor edited the previous answer based on the reflection critiques with the aid of the information. The Revisor reviewed its answer again and provided new critiques if it decided the answer required improvements. The supervisor received all the fields from the Revisor and decided that the Google agent should be used. Having revised the answer with the information provided by the agent, the Supervisor saw the answer being the best and finished the execution. View this table: View inline View popup Table 2: Process of ENTAgents handling a question. The bold texts are the revised content in the latest answer compared to the previous answer. An example of the answers to an essay question from Llama 3.1 and ENTAgents was compared in Table 3 . The ENTAgents provided a more comprehensive answer to the question than Llama 3.1. In the pathogenesis part, the answers of ENTAgents included detailed and specific pathological processes and consequences of cholesteatoma and covered the impact on surrounding structures. In addressing the etiology of cholesteatoma, Llama 3.1 provided only a list of factors associated with the condition, whereas ENTAgents offered a detailed explanation. ENTAgents not only detailed how these factors contribute to cholesteatoma development but also discussed the different types of cholesteatoma and their specific causes. Additionally, while Llama 3.1 produced a simple list of clinical features, ENTAgents elaborated on the underlying causes of these features. It then presented a thorough response, addressing complications seen in both typical and advanced cases. In terms of cholesteatoma management, ENTAgents described not only the primary approaches and surgical techniques based on the size and location of the cholesteatoma but also mentioned the use of imaging tools to aid in management. Overall, ENTAgents provided a more comprehensive and structured response that covered all aspects of the query with detailed explanations. View this table: View inline View popup Table 3: Comparison of the responses from ENTAgents and Llama 3.1 to an essay question. 3.3 Multiple Choice Question (MCQ) Finally, we challenged our ENTAgents with 992 MCQs from BoardVitals, which are US board review practice questions that follow the exam content outline for US otolaryngology qualifying examinations. The MCQ question banks consist of 9 sections, each with three levels of difficulty, which are hard, moderate, and easy, and questions from subspecialties including Allergy, Endocrine, Head and Neck, Laryngology, Otology, Pediatrics, Pharmacology, Plastic and Reconstructive and Sleep. The distributions of questions in different subspecialties in the question banks are presented in Table 4 . View this table: View inline View popup Download powerpoint Table 4: Distributions of the MCQs in 9 subspecialties of otolaryngology. The performance of ENTAgents in answering questions from all specialties is summarized in Figure 4 . The results were sorted by the score of ENTAgents from lowest to highest. ENTAgents obtained scores above 70% in all specialties, with the score in Plastic and Reconstructive being the lowest, 68.4%. ENTAgents’s highest score was in Allergy, which is 91.4%. The overall performance of ENTAgents was also calculated and compared to the average correct percentage of humans, obtained from the publicly available data from BoardVitals. ENTAgents had an overall performance of 75.5, 4.5% better than humans, scoring 71%. Download figure Open in new tab Figure 4: The scores of ENTAgent for the otolaryngology MCQ. Table 5 shows an example of the response of ENTAgents to an MCQ. Without RAG and the agents, Llama 3.1 initially chose option B as its answer. Then, the answer was considered as not thorough by ENTAgents. With the retrieval from the knowledge base, it was discovered that the actual likelihood should be 45%. The revisor agent immediately changed its choice to C, which was the correct answer. ArXiv agent was invoked, and it summarized the scholar articles it searched for about ultrasound of the nodule to its response. View this table: View inline View popup Table 5: Example of the self-correction of ENTAgents to an MCQ. 4. Discussion These results show that ENTAgents has the capability to provide comprehensive answers with the aid of reflection critiques. The reflection critiques enable ENTAgents to comment on its answers and suggest what can be improved. In answering short questions and essay questions, ENTAgents can provide the relevant additional information only and remove the abundant texts, making the answer more informative but also concise. ENTAgents provides more informative responses than other AI models, demonstrating enhanced accuracy. For the MCQ question, ENTAgents demonstrates the ability to change its option to the correct one, the Revisor examines these messages from RAG and agents to adjust its answer if needed. Utilizing agents and RAG can supplement the AI framework with up-to-date, highly specific, and detailed information, improving the accuracy and completeness of its responses. Since the RAG database contains knowledge of otolaryngology, ENTAgents is therefore able to draw from verified information and provides accurate answers for this field. The multi-agent system is critical for complex queries, which requires a diverse range of answers and the latest information. The agents can search websites and the literature database to find up-to-date developments related to the question. This approach can effectively address essay questions, allowing inquirers to gain insights into recent advancements and research developments in otolaryngology rather than relying solely on static, pre-existing information from RAG. Thus, our framework offers three distinct advantages over the other LLM agent applications. The first one is extensive search capabilities. For example, ENTAgents draws information from both the RAG knowledge base and external sources through the use of agents. Therefore, it does not require the continuous update of the pertinent sources from the field. This enhances the accuracy, completeness and timeliness of its responses, offering ENTAgents the ability to address a wide range of questions, from fundamental concepts to the latest advancements. The second is self-correction. This means that the reflection framework allows ENTAgents to evaluate their responses according to linguistic feedback. As the use of agents is iterative, the framework ensures that agents are leveraged only when flaws or inaccuracies are detected in the response. Furthermore, it removes superfluous information, ensuring that the final answer is both comprehensive and free of redundancies. Finally, ENTAgents is ease of use. Since the RAG knowledge base is built by external sources and ENTAgents’s ability to self-correct derived from its own reflective critiques, generated through prompting alone, fine-tuning is not required within our framework. The reflection critiques and search queries are provided along with the answers so that users can understand clearly the generation of the response of ENTAgents, enhancing its explainability. There are many opportunities for ENTAgents to be utilized across different fields or situations; adjustments to the knowledge base and the reflection prompts only allow ENTAgents to provide its response according to other criteria as needed. In addition, the model we used, Llama 3.1, is an open source LLM model, which offers full access and capability to develop advanced solutions, unlike other proprietary models, and hence we can modify ENTAgents easily for future applications. Our work illustrates the potential applications of ENTAgents, including medical education, patient education, and clinical support. First, ENTAgents can be leveraged to provide medical education to students effectively. With the ability to retrieve solid knowledge from guidelines, books or academic papers in the knowledge base in a second, the learning process can be streamlined, and students can absorb concentrated information efficiently. ENTAgents can also be used in revision for medical students, given its ability to respond to various types of questions comprehensively. Healthcare professionals can also benefit from learning about the advancements in their field. The latest literature and academic symposiums can be found by the agents, and they will be summarized to the users. Professionals can then learn about cutting-edge technology and recent developments in the field. On top of that, the public can be educated by ENTAgents. The flexibility of ENTAgents to change its response makes it easy to explain difficult medical terms or knowledge to the layman, and it can be developed into the mobile application. The patients can thus easily access ENTAgents to have a better understanding of precautions and treatments of the diseases. Lastly, the accessibility of ENTAgents to various sources of scholarly articles provides insights to clinicians to handle rare cases and assist them in decision-making. Since the explainable concerns can be dampened with the reflection critiques provided, physicians can understand the workflow of ENTAgents and be assured to take the recommendations of ENTAgents. Although our framework excels in generating comprehensive responses to users, the latency of ENTAgents needs to be improved. ENTAgents consumes a lot of time searching the information and revising the answers by the agents. Thus, it will not be a user-friendly application to wait such a long time for its response, further improvement in its speed is paramount. Moreover, the responses of ENTAgents deviated from the prescribed format several times. While the response structure is organized into distinct sections such as answer, reflective critique, and search query, the agent occasionally produces reflective content as an answer, resulting in wrong outputs and altering the workflow. Prompting techniques should be refined to ensure consistent adherence to the designated response format. Lastly, ENTAgents focuses on integrating verbal reinforcement learning within a multi-agent system to deliver comprehensive responses. Therefore, it lacks multimodal capabilities to store visual or audio information within its knowledge base and respond to queries beyond text-based inputs. Despite these limitations, the ENTAgents presents an advanced framework capable of generating accurate and thorough responses, demonstrating significant potential for diverse applications. Our future research will first emphasize enhancing the multimodal capabilities of ENTAgents to enable the interpretation of visual data, particularly medical images sourced from both the RAG and the agents, and provide visual explanations to the users. As a result, ENTAgents will be better equipped to tackle complex medical cases across different modalities, offering clearer and more comprehensible responses for users. In addition, we aim to scale our knowledge database to encompass diverse entities, including diseases and conditions, imaging data, anatomical features, and more. This data will be structured well and interlinked to organize a healthcare knowledge graph for ENTAgents. With this approach, ENTAgents will deliver higher quality and more relevant answers effectively. Conflicts of interest There are no conflicts to declare. Data Availability All data produced in the present study are available upon reasonable request to the authors Supplementary Material 1: Used documents for RAG knowledge base View this table: View inline View popup Table 1: Short questions and answers from ENTAgent. View this table: View inline View popup Table 2: Essay questions and answers from ENTAgent. Acknowledgements We gratefully acknowledge the funding provided by the Research Grant Council of Hong Kong, General Research Fund (Ref No. 14211223). References 1. ↵ Gebreab SA , Salah K , Jayaraman R , et al. LLM-based framework for Administrative Task Automation in Healthcare . 2024 12th International Symposium on Digital Forensics and Security (ISDFS) . 2023 Apr 29 ; 1 – 7 . DOI: 10.1109/isdfs60797.2024.10527275 OpenUrl CrossRef 2. ↵ Goel A , Gueta A , Omry Gilon et al. LLMs Accelerate Annotation for Medical Information Extraction . arXiv (Cornell University) . 2023 Dec 4 ; DOI: 10.48550/arXiv.2312.02296 OpenUrl CrossRef 3. ↵ Panagoulias , Dimitrios P , Virvou M , Tsihrintzis GA . Evaluating LLM -- Generated Multimodal Diagnosis from Medical Images and Symptom Analysis . arXiv (Cornell University) . 2024 Jan 28 ; DOI: 10.48550/arXiv.2402.01730 OpenUrl CrossRef 4. ↵ Wang S , Zhao Z , Ouyang X , Wang Q , Shen D. ChatCAD: Interactive Computer-Aided Diagnosis on Medical Image using Large Language Models . 2023 Feb 14 ; DOI: 10.48550/arXiv.2302.07257 OpenUrl CrossRef 5. ↵ Kim JK , Chua M , Rickard M , Lorenzo A . Chatgpt and Large Language Model (LLM) Chatbots: The current state of acceptability and a proposal for guidelines on utilization in Academic Medicine . Journal of Pediatric Urology . 2023 Sept 26 ; 19 ( 5 ): 598 – 604 . DOI: 10.1016/j.jpurol.2023.05.018 OpenUrl CrossRef PubMed 6. ↵ Cascella M , Montomoli J , Bellini V , Bignami E . Evaluating the feasibility of CHATGPT in Healthcare: An analysis of multiple clinical and research scenarios . Journal of Medical Systems . 2023 Mar 4 ; 47 ( 1 ). DOI: 10.1007/s10916-023-01925-4 OpenUrl CrossRef PubMed 7. ↵ Thirunavukarasu AJ , Ting DS , Elangovan K , Gutierrez L , Tan TF , Ting DS . Large language models in medicine . Nature Medicine . 2023 Jul 17 ; 29 ( 8 ): 1930 – 40 . DOI: 10.1038/s41591-023-02448-8 OpenUrl CrossRef PubMed 8. ↵ Hallucination Leaderboard . 2024 Nov 6 . Available from: https://github.com/vectara/hallucination-leaderboard 9. ↵ Li J , Yuan Y , Zhang Z . Enhancing LLM Factual Accuracy with RAG to Counter Hallucinations: A Case Study on Domain-Specific Queries in Private Knowledge-Bases . arXiv.org . 2024 . DOI: 10.48550/arXiv.2403.10446 OpenUrl CrossRef 10. Shi L , Kazda M , Sears B , Shropshire N , Puri R. Ask-EDA: A Design Assistant Empowered by LLM, Hybrid RAG and Abbreviation De-hallucination . arXiv (Cornell University) . 2024 Jun 3 . DOI: 10.48550/arXiv.2406.06575 OpenUrl CrossRef 11. Niu C , Wu Y , Zhu J , et al. Ragtruth: A hallucination corpus for developing trustworthy retrieval-augmented language models . Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) . 2024 Aug ; 10862 – 78 . DOI: 10.18653/v1/2024.acl-long.585 OpenUrl CrossRef 12. Alkhalaf M , Yu P , Yin M , Deng C . Applying generative AI with retrieval augmented generation to summarize and extract key clinical information from electronic health records . Journal of Biomedical Informatics . 2024 Aug ; 156 : 104662 . DOI: 10.1016/j.jbi.2024.104662 OpenUrl CrossRef 13. Miao J , Thongprayoon C , Suppadungsuk S , Garcia Valencia OA , Cheungpasitporn W . Integrating retrieval-augmented generation with large language models in nephrology: Advancing Practical Applications . Medicina . 2024 Mar 8 ; 60 ( 3 ): 445 . DOI: 10.3390/medicina60030445 OpenUrl CrossRef 14. ↵ Long C , Subburam D , Lowe K et al. Chatent: Augmented large language model for expert knowledge retrieval in otolaryngology - head and Neck Surgery . American Academy of Otolaryngology—Head and Neck Surgery . 2023 Jun 19 ; DOI: 10.2139/ssrn.4571725 OpenUrl CrossRef 15. ↵ Unlu O , Shin J , Mailly CJ , Oates MF , Tucci MR , Varugheese M , et al. Retrieval-augmented generation–enabled GPT-4 for clinical trial screening . NEJM AI . 2024 Jun 27 ; 1 ( 7 ). DOI: 10.1056/aioa2400181 OpenUrl CrossRef 16. ↵ Kaur D , Uslu S , Durresi M , Durresi A . LLM-based agents utilized in a trustworthy artificial conscience model for controlling AI in medical applications . Lecture Notes on Data Engineering and Communications Technologies . 2024 Apr 10 ; 198 – 209 . DOI: 10.1007/978-3-031-57870-0_18 OpenUrl CrossRef 17. ↵ Mahyar Abbasian , Azimi I , Rahmani AM , Jain R. Conversational Health Agents: A Personalized LLM-Powered Agent Framework . arXiv (Cornell University) . 2023 Oct 3 . DOI: 10.48550/arXiv.2310.02374 OpenUrl CrossRef 18. Pandey HG , Amod A , Kumar S . Advancing Healthcare Automation: Multi-agent system for medical necessity justification . Proceedings of the 23rd Workshop on Biomedical Natural Language Processing . 2024 Aug 16 ; 39 – 49 . DOI: 10.18653/v1/2024.bionlp-1.4 OpenUrl CrossRef 19. Tang X , Zou A , Zhang Z , et al. MedAgents: Large language models as collaborators for Zero-Shot Medical Reasoning . Findings of the Association for Computational Linguistics ACL 2024 . 2024 Jun 4 ; 599 – 621 . DOI: 10.18653/v1/2024.findings-acl.33 OpenUrl CrossRef 20. ↵ Schmidgall S , Ziaei R , Harris C , Reis E , Jopling J , Moor M . AgentClinic: a multimodal agent benchmark to evaluate AI in simulated clinical environments . arXiv (Cornell University) . 2024 May 13 ; DOI: 10.48550/arXiv.2405.07960 OpenUrl CrossRef 21. ↵ Alghamdi HM , Mostafa A . Towards reliable healthcare LLM agents: A case study for pilgrims during Hajj . Information . 2024 Jun 26 ; 15 ( 7 ): 371 . DOI: 10.3390/info15070371 OpenUrl CrossRef 22. Borkowski A , Alon Ben-Ari . Muli-Agent AI Systems in Healthcare: Technical and Clinical Analysis . 2024 Oct 2 . DOI: 10.20944/preprints202410.0182.v1 OpenUrl CrossRef 23. ↵ Li J , Wang S , Zhang M , et al. Agent Hospital: A Simulacrum of Hospital with Evolvable Medical Agents . arXiv.org . 2024 May 5 . DOI: 10.48550/arXiv.2405.02957 OpenUrl CrossRef 24. ↵ Hiesinger W , Zakka C , Chaurasia A , et al. Almanac: Retrieval-augmented language models for Clinical Medicine . NEJM AI . 2024 Jan 25 ; 1 ( 2 ). DOI: 10.1056/AIoa2300068 OpenUrl CrossRef 25. ↵ Shinn N , Labash B , Gopinath A . Reflexion: Language Agents with Verbal Reinforcement Learning . arXiv (Cornell University) . 2023 Mar 20 ; 1 ( 1 ). DOI: 10.48550/arXiv.2303.11366 OpenUrl CrossRef 26. Rosenfeld RM , Schwartz SR , Cannon CR , et al. Clinical practice guideline: acute otitis externa executive summary . Otolaryngol Head Neck Surg . 2014 ; 150 ( 2 ): 161 – 168 . DOI: 10.1177/0194599813517659 OpenUrl CrossRef PubMed Web of Science 27. Rosenfeld RM , Schwartz SR , Cannon CR , et al. Clinical practice guideline: acute otitis externa [published correction appears in Otolaryngol Head Neck Surg. 2014 Mar; 150(3):504] . Otolaryngoloy Head Neck Surg . 2014 ; 150 ( 1 Suppl ): S1 - S24 . DOI: 10.1177/0194599813517083 OpenUrl CrossRef PubMed Web of Science 28. Almeida JR de , Guyatt GH , Sud S , et al. Management of Bell palsy: clinical practice guideline . CMAJ . 2014 ; 186 ( 12 ): 917 – 922 . DOI: 10.1503/cmaj.131801 OpenUrl FREE Full Text 29. Bhattacharyya N , Baugh RF , Orvidas L , et al. Clinical practice guideline: benign paroxysmal positional vertigo . Otolaryngol Head Neck Surg . 2008 ; 139 ( 5 Suppl 4 ): S47 - S81 . DOI: 10.1016/j.otohns.2008.08.022 OpenUrl CrossRef PubMed Web of Science 30. Califano L , Salafia F , Mazzone S , Melillo MG , Califano M . Anterior canal BPPV and apogeotropic posterior canal BPPV: two rare forms of vertical canalolithiasis . Acta Otorhinolaryngol Ital . 2014 ; 34 ( 3 ): 189 – 197 . OpenUrl PubMed 31. Roland PS , Smith TL , Schwartz SR , et al. Clinical practice guideline: cerumen impaction . Otolaryngol Head Neck Surg . 2008 ; 139 ( 3 Suppl 2 ): S1 - S21 . DOI: 10.1016/j.otohns.2008.06.026 OpenUrl CrossRef PubMed Web of Science 32. Stachler RJ , Chandrasekhar SS , Archer SM , et al. Clinical practice guideline: sudden hearing loss . Otolaryngology–Head and Neck Surgery . 2012 ; 146 ( 3_suppl ): S1 - S35 . DOI: 10.1177/0194599812436449 OpenUrl CrossRef PubMed Web of Science 33. Basura GJ , Adams ME , Monfared A , et al. Clinical practice guideline: Ménière’s Disease executive summary . Otolaryngology–Head and Neck Surgery . 2020 ; 162 ( 4 ): 415 – 434 . DOI: 10.1177/0194599820909439 OpenUrl CrossRef PubMed 34. Bhattacharyya N , Gubbels SP , Schwartz SR , et al. Clinical practice guideline: benign paroxysmal positional vertigo (Update) . Otolaryngol Head Neck Surg . 2017 ; 156 ( 3_suppl ): S1 - S47 . DOI: 10.1177/0194599816689667 OpenUrl CrossRef 35. Chandrasekhar SS , Tsai Do BS , Schwartz SR , et al. Clinical practice guideline: sudden hearing loss (update) executive summary . Otolaryngol Head Neck Surg . 2019 ; 161 ( 2 ): 195 – 210 . DOI: 10.1177/0194599819859883 OpenUrl CrossRef PubMed 36. Tunkel DE , Bauer CA , Sun GH , et al. Clinical practice guideline: tinnitus executive summary . Otolaryngol Head Neck Surg . 2014 ; 151 ( 4 ): 533 – 541 . DOI: 10.1177/0194599814547475 OpenUrl CrossRef PubMed 37. Lempert T , Olesen J , Furman J , et al. Vestibular migraine: diagnostic criteria . J Vestib Res . 2012 ; 22 ( 4 ): 167 – 172 . DOI: 10.3233/VES-2012-0453 OpenUrl Abstract / FREE Full Text 38. Laursen BB , Danstrup CS , Hoffmann S , Nørskov-Lauritsen N , Christensen ALB , Ovesen T . The effect of pneumococcal conjugate vaccines on incidence and microbiology associated with complicated acute otitis media . International Journal of Pediatric Otorhinolaryngology . 2017 ; 101 : 249 – 253 . DOI: 10.1016/j.ijporl.2017.07.002 OpenUrl CrossRef PubMed 39. Nieminen TA , Kivekäs I , Artama M , Nohynek H , Kujansivu J , Hovi P . Sudden hearing loss following vaccination against COVID-19 . JAMA Otolaryngology–Head & Neck Surgery . 2023 ; 149 ( 2 ): 133 – 140 . DOI: 10.1001/jamaoto.2022.4154 OpenUrl CrossRef 40. Yanir Y , Doweck I , Shibli R , Najjar-Debbiny R , Saliba W . Association between the BNT162b2 messenger RNA COVID-19 vaccine and the risk of sudden sensorineural hearing loss . JAMA Otolaryngology–Head & Neck Surgery . 2022 ; 148 ( 4 ): 299 – 306 . DOI: 10.1001/jamaoto.2021.4278 OpenUrl CrossRef 41. Coelho DH , Lalwani AK . Medical management of Ménière’s Disease . The Laryngoscope . 2008 ; 118 ( 6 ): 1099 – 1108 . DOI: 10.1097/MLG.0b013e31816927f0 OpenUrl CrossRef PubMed 42. Hadjipanayis CG , Carlson ML , Link MJ , et al. Congress of Neurological Surgeons systematic review and evidence-based guidelines on surgical resection for the treatment of patients with vestibular schwannomas . Neurosurgery . 2018 ; 82 ( 2 ): E40 . DOI: 10.1093/neuros/nyx512 OpenUrl CrossRef PubMed 43. Tunkel DE , Bauer CA , Sun GH , et al. Clinical practice guideline: tinnitus executive summary . Otolaryngology–Head and Neck Surgery . 2014 ; 151 ( 4 ): 533 – 541 . DOI: 10.1177/0194599814547475 OpenUrl CrossRef PubMed 44. Schwartz SR , Magit AE , Rosenfeld RM , et al. Clinical practice guideline (update): earwax (cerumen impaction) executive summary . Otolaryngology–Head and Neck Surgery . 2017 ; 156 ( 1 ): 14 – 29 . DOI: 10.1177/0194599816678832 OpenUrl CrossRef PubMed 45. Tucci DL , McCoul ED , Rosenfeld RM , et al. Clinical consensus statement: balloon dilation of the eustachian tube . Otolaryngology–Head and Neck Surgery . 2019 ; 161 ( 1 ): 6 – 17 . DOI: 10.1177/0194599819848423 OpenUrl CrossRef PubMed 46. Haugen BR , Alexander EK , Bible KC , et al. 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer . Thyroid . 2016 ; 26 ( 1 ): 1 - 133 . DOI: 10.1089/thy.2015.0020 OpenUrl CrossRef PubMed 47. Wells SA Jr , Asa SL , Dralle H , et al. Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma . Thyroid . 2015 ; 25 ( 6 ): 567 – 610 . DOI: 10.1089/thy.2014.0335 OpenUrl CrossRef PubMed 48. American Thyroid Association Taskforce On Radioiodine Safety , Sisson JC , Freitas J , et al. Radiation safety in the treatment of patients with thyroid diseases by radioiodine 131I: practice recommendations of the American Thyroid Association [published correction appears in Thyroid. 2011 Jun;21(6):689] . Thyroid . 2011 ; 21 ( 4 ): 335 - 346 . DOI: 10.1089/thy.2010.0403 OpenUrl CrossRef PubMed Web of Science 49. American Thyroid Association (ATA) Guidelines Taskforce on Thyroid Nodules and Differentiated Thyroid Cancer , Cooper DS , Doherty GM , et al. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer [published correction appears in Thyroid. 2010 Aug;20(8):942. Hauger, Bryan R [corrected to Haugen, Bryan R]] [published correction appears in Thyroid. 2010 Jun;20(6):674-5] . Thyroid . 2009 ; 19 ( 11 ): 1167 - 1214 . DOI: 10.1089/thy.2009.0110 OpenUrl CrossRef PubMed Web of Science 50. The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer , Haugen BR , Alexander EK , Bible KC , et al. 2015 American Thyroid Association management guidelines for adult patients with thyroid nodules and differentiated thyroid cancer . Thyroid . 2016 ; 26 ( 1 ): 1 - 133 . DOI: 10.1089/thy.2015.0020 OpenUrl CrossRef PubMed 51. Wells SA Jr , Asa SL , Dralle H , et al. Revised American Thyroid Association guidelines for the management of medullary thyroid carcinoma . Thyroid . 2015 ; 25 ( 6 ): 567 – 610 . DOI: 10.1089/thy.2014.0335 OpenUrl CrossRef PubMed 52. Bible KC , Kebebew E , Brierley J , et al. 2021 American Thyroid Association guidelines for management of patients with anaplastic thyroid cancer: American Thyroid Association Anaplastic Thyroid Cancer Guidelines Task Force . Thyroid . 2021 ; 31 ( 3 ): 337 - 386 . DOI: 10.1089/thy.2020.0944 OpenUrl CrossRef PubMed 53. van Dijk SPJ , Coerts HI , Gunput STG , et al. Assessment of radiofrequency ablation for papillary microcarcinoma of the thyroid: A systematic review and meta-analysis . JAMA Otolaryngology–Head & Neck Surgery . 2022 ; 148 ( 4 ): 317 – 325 . DOI: 10.1001/jamaoto.2021.4381 OpenUrl CrossRef 54. Alexander EK , Pearce EN , Brent GA , et al. 2017 Guidelines of the American Thyroid Association for the diagnosis and management of thyroid disease during pregnancy and the postpartum . Thyroid . 2017 ; 27 ( 3 ): 315 - 389 . DOI: 10.1089/thy.2016.0457 OpenUrl CrossRef PubMed 55. Chandrasekhar SS , Randolph GW , Seidman MD , et al. Clinical practice guideline: improving voice outcomes after thyroid surgery . Otolaryngol Head Neck Surg . 2013 ; 148 ( 6 Suppl ): S1 - S37 . DOI: 10.1177/0194599813487301 OpenUrl CrossRef PubMed 56. Tunkel DE , Anne S , Payne SC , et al. Clinical practice guideline: nosebleed (epistaxis) executive summary . Otolaryngology–Head and Neck Surgery . 2020 ; 162 ( 1 ): 8 – 25 . DOI: 10.1177/0194599819889955 OpenUrl CrossRef PubMed 57. Zidar N , Gale N . Update from the 5th Edition of the World Health Organization classification of head and neck tumors: hypopharynx, Larynx, trachea and parapharyngeal space . Head and Neck Pathol . 2022 ; 16 ( 1 ): 31 – 39 . DOI: 10.1007/s12105-021-01405-6 OpenUrl CrossRef 58. Cibas ES , Ali SZ. The 2017 Bethesda system for reporting thyroid cytopathology . Thyroid . 2017 ; 27 ( 11 ): 1341 - 1346 . DOI: 10.1089/thy.2017.0500 OpenUrl CrossRef PubMed 59. Zhang SL , Ng HW . Primary thrombolysis for free flap surgery in head and neck reconstruction: a case report and review . Arch Plast Surg . 2021 ; 48 ( 5 ): 511 – 517 . DOI: 10.5999/aps.2021.00171 OpenUrl CrossRef PubMed 60. Seethala RR , Baloch ZW , Barletta JA , et al. Noninvasive follicular thyroid neoplasm with papillary-like nuclear features: a review for pathologists . Mod Pathol . 2018 ; 31 ( 1 ): 39 – 55 . DOI: 10.1038/modpathol.2017.130 OpenUrl CrossRef PubMed 61. Robbins KT , Clayman G , Levine PA , et al. Neck dissection classification update: revisions proposed by the American Head and Neck Society and the American Academy of Otolaryngology-Head and Neck Surgery . Arch Otolaryngol Head Neck Surg . 2002 ; 128 ( 7 ): 751 – 758 . DOI: 10.1001/archotol.128.7.751 OpenUrl CrossRef PubMed Web of Science 62. Maghami E , Ismaila N , Alvarez A , et al. Diagnosis and management of squamous cell carcinoma of unknown primary in the head and neck: ASCO Guideline . JCO . 2020 ; 38 ( 22 ): 2570 – 2596 . DOI: 10.1200/JCO.20.00275 OpenUrl CrossRef PubMed 63. Robbins KT , Shaha AR , Medina JE , et al. Consensus statement on the classification and terminology of neck dissection . Archives of Otolaryngology–Head & Neck Surgery . 2008 ; 134 ( 5 ): 536 – 538 . DOI: 10.1001/archotol.134.5.536 OpenUrl CrossRef PubMed 64. Dellon ES , Gonsalves N , Hirano I , et al. ACG clinical guideline: Evidenced based approach to the diagnosis and management of esophageal eosinophilia and eosinophilic esophagitis (EoE) . Am J Gastroenterol . 2013 ; 108 ( 5 ): 679 – 693 . DOI: 10.1038/ajg.2013.71 OpenUrl CrossRef PubMed 65. Stachler RJ , Francis DO , Schwartz SR , et al. Clinical practice guideline: hoarseness (dysphonia) (update) executive summary . Otolaryngol Head Neck Surg . 2018 ; 158 ( 3 ): 409 – 426 . DOI: 10.1177/0194599817751031 OpenUrl CrossRef PubMed 66. Chandrasekhar SS , Randolph GW , Seidman MD , et al. Clinical practice guideline: improving voice outcomes after thyroid surgery . Otolaryngol Head Neck Surg . 2013 ; 148 ( 6 Suppl ): S1 - S37 . DOI: 10.1177/0194599813487301 OpenUrl CrossRef PubMed 67. O’Hara J , Stocken DD , Watson GC , et al. Use of proton pump inhibitors to treat persistent throat symptoms: multicentre, double blind, randomised, placebo controlled trial . BMJ . 2021 ; 372 : m4903 . DOI: 10.1136/bmj.m4903 OpenUrl Abstract / FREE Full Text 68. Cergan R , Dumitru M , Vrinceanu D , Neagos A , Jeican II , Ciuluvica RC . Ultrasonography of the larynx: novel use during the SARS-CoV-2 pandemic (review) . Exp Ther Med . 2021 ; 21 ( 3 ): 273 . DOI: 10.3892/etm.2021.9704 OpenUrl CrossRef PubMed 69. Khan NC , Vukkadala N , Saxena A , Damrose EJ , Nekhendzy V , Sung CK . Safety and utility of transnasal humidified rapid-insufflation ventilatory exchange (THRIVE) for laser laryngeal surgery . Otolaryngol--head neck surg. Published online April 6 , 2023 : ohn.324 . DOI: 10.1002/ohn.324 OpenUrl CrossRef 70. Biller HF , Barnhill Jr . . FR , Ogura JH , Perez CA . Hemilaryngectomy following radiation failure for carcinoma of the vocal cords . The Laryngoscope . 1970 ; 80 ( 2 ): 249 – 253 . DOI: 10.1288/00005537-197002000-00009 OpenUrl CrossRef PubMed Web of Science 71. Hu A , Morrison M , Honey CR . Hemi-laryngopharyngeal spasm (HeLPS): defining a new clinical entity . Ann Otol Rhinol Laryngol . 2020 ; 129 ( 9 ): 849 – 855 . DOI: 10.1177/0003489420916207 OpenUrl CrossRef PubMed 72. Kuhn MA , Gillespie MB , Ishman SL , et al. Expert consensus statement: management of dysphagia in head and neck cancer patients . Otolaryngol--head neck surg . 2023 ; 168 ( 4 ): 571 – 592 . DOI: 10.1002/ohn.302 OpenUrl CrossRef PubMed 73. Lund VJ . European Position Paper on the Anatomical Terminology of the Internal Nose and Paranasal Sinuses . 2014 . 74. Gupta S , Lee JJ , Perrin A , et al. Efficacy and safety of saline nasal irrigation plus theophylline for treatment of COVID-19-related olfactory dysfunction: The SCENT2 Phase 2 Randomized Clinical Trial . JAMA Otolaryngol Head Neck Surg . 2022 ; 148 ( 9 ): 830 – 837 . DOI: 10.1001/jamaoto.2022.1573 OpenUrl CrossRef PubMed 75. Bichakjian CK , Olencki T , Aasi SZ , et al. Basal cell skin cancer, version 1.2016, NCCN clinical practice guidelines in oncology . Journal of the National Comprehensive Cancer Network . 2016 May ; 14 ( 5 ): 574 – 97 . DOI: 10.6004/jnccn.2016.0065 OpenUrl Abstract / FREE Full Text 76. Jozwik M , Bednarczuk K , Osierda Z . Dermatofibrosarcoma protuberans: An updated review of the literature . Cancers . 2024 Sept 11 ; 16 ( 18 ): 3124 . DOI: 10.3390/cancers16183124 OpenUrl CrossRef PubMed 77. Pfister DG , Spencer S , Brizel DM , et al. Head and neck cancers, version 1.2015 . Journal of the National Comprehensive Cancer Network . 2015 Jul ; 13 ( 7 ): 847 – 56 . DOI: 10.6004/jnccn.2015.0102 OpenUrl Abstract / FREE Full Text 78. Coit DG , Thompson JA , Algazi A , et al. Melanoma, version 2.2016, NCCN clinical practice guidelines in oncology . Journal of the National Comprehensive Cancer Network . 2016 Apr ; 14 ( 4 ): 450 – 73 . DOI: 10.6004/jnccn.2016.0051 OpenUrl Abstract / FREE Full Text 79. Zaggana E , Konstantinou MP , Krasagakis GH , et al. Merkel cell carcinoma—update on diagnosis, management and future perspectives . Cancers . 2022 Dec 23 ; 15 ( 1 ): 103 . DOI: 10.3390/cancers15010103 OpenUrl CrossRef PubMed 80. Schmults CD , Blitzblau R , Aasi SZ , Alam M , Andersen JS , Baumann BC , et al. NCCN Guidelines® insights: Squamous cell skin cancer, version 1.2022 . Journal of the National Comprehensive Cancer Network . 2021 Dec ; 19 ( 12 ): 1382 – 94 . DOI: 10.6004/jnccn.2021.0059 OpenUrl CrossRef PubMed 81. Haddad RI , Lydiatt WM , Ball DW , Busaidy NL , Byrd D , Callender G , et al. Anaplastic thyroid carcinoma, version 2.2015 . Journal of the National Comprehensive Cancer Network . 2015 Sept ; 13 ( 9 ): 1140 – 50 . DOI: 10.6004/jnccn.2015.0139 OpenUrl Abstract / FREE Full Text 82. Wald ER , Applegate KE , Bordley C , et al. Clinical practice guideline for the diagnosis and management of acute bacterial sinusitis in children aged 1 to 18 years . Pediatrics . 2013 ; 132 ( 1 ): e262 – e280 . DOI: 10.1542/peds.2013-1071 OpenUrl CrossRef PubMed Web of Science 83. Lieberthal AS , Carroll AE , Chonmaitree T , et al. The diagnosis and management of acute otitis media . Pediatrics . 2013 ; 131 ( 3 ): e964 – e999 . DOI: 10.1542/peds.2012-3488 OpenUrl CrossRef PubMed Web of Science 84. Le Saux N , Robinson JL ; Canadian Paediatric Society, Infectious Diseases and Immunization Committee. Management of acute otitis media in children six months of age and older . Paediatr Child Health . 2016 ; 21 ( 1 ): 39 – 50 . DOI: 10.1093/pch/21.1.39 OpenUrl CrossRef PubMed 85. American Academy of Family Physicians, American Academy of Otolaryngology-Head and Neck Surgery, American Academy of Pediatrics Subcommittee on Otitis Media With Effusion . Otitis media with effusion . Pediatrics . 2004 ; 113 ( 5 ): 1412 - 1429 . DOI: 10.1542/peds.113.5.1412 OpenUrl CrossRef PubMed Web of Science 86. Rosenfeld RM , Shin JJ , Schwartz SR , et al. Clinical practice guideline: otitis media with effusion executive summary (update) . Otolaryngology–Head and Neck Surgery . 2016 ; 154 ( 2 ): 201 – 214 . DOI: 10.1177/0194599815624407 OpenUrl CrossRef PubMed 87. Roland PS , Rosenfeld RM , Brooks LJ , et al. Clinical practice guideline: polysomnography for sleep-disordered breathing prior to tonsillectomy in children . Otolaryngol Head Neck Surg . 2011 ; 145 ( 1 Suppl ): S1 - S15 . DOI: 10.1177/0194599811409837 OpenUrl CrossRef PubMed Web of Science 88. Baugh RF , Archer SM , Mitchell RB , et al. Clinical practice guideline: tonsillectomy in children . Otolaryngol Head Neck Surg . 2011 ; 144 ( 1 Suppl ): S1 - S30 . DOI: 10.1177/0194599810389949 OpenUrl CrossRef PubMed Web of Science 89. Mitchell RB , Archer SM , Ishman SL , et al. Clinical practice guideline: tonsillectomy in children (update)-executive summary . Otolaryngol Head Neck Surg . 2019 ; 160 ( 2 ): 187 – 205 . DOI: 10.1177/0194599818807917 OpenUrl CrossRef PubMed 90. Rosenfeld RM , Schwartz SR , Pynnonen MA , et al. Clinical practice guideline: tympanostomy tubes in children--executive summary . Otolaryngol Head Neck Surg . 2013 ; 149 ( 1 ): 8 – 16 . DOI: 10.1177/0194599813490141 OpenUrl CrossRef PubMed Web of Science 91. Rosenfeld RM , Tunkel DE , Schwartz SR , et al. Executive summary of clinical practice guideline on tympanostomy tubes in children (update) . Otolaryngol Head Neck Surg . 2022 ; 166 ( 2 ): 189 – 206 . DOI: 10.1177/01945998211065661 OpenUrl CrossRef PubMed 92. Joint Committee on Infant Hearing . Year 2007 Position Statement: Principles and guidelines for early hearing detection and intervention programs . Pediatrics . 2007 ; 120 ( 4 ): 898 - 921 . DOI: 10.1542/peds.2007-2333 OpenUrl CrossRef PubMed Web of Science 93. van de Berg R , Widdershoven J , Bisdorff A , et al. Vestibular migraine of childhood and recurrent vertigo of childhood: diagnostic criteria consensus document of the Committee for the Classification of Vestibular Disorders of the Bárány Society and the International Headache Society . J Vestib Res . 2021 ; 31 ( 1 ): 1 – 9 . DOI: 10.3233/VES-200003 OpenUrl CrossRef PubMed 94. Francis GL , Waguespack SG , Bauer AJ , et al. Management guidelines for children with thyroid nodules and differentiated thyroid cancer . Thyroid . 2015 ; 25 ( 7 ): 716 – 759 . DOI: 10.1089/thy.2014.0460 OpenUrl CrossRef PubMed 95. Baldassari CM , Lam DJ , Ishman SL , et al. Expert consensus statement: pediatric drug-induced sleep endoscopy . Otolaryngol Head Neck Surg . 2021 ; 165 ( 4 ): 578 – 591 . DOI: 10.1177/0194599820985000 OpenUrl CrossRef PubMed 96. Ishman SL , Maturo S , Schwartz S , et al. Expert consensus statement: management of pediatric persistent obstructive sleep apnea after adenotonsillectomy . Otolaryngol--head neck surg . 2023 ; 168 ( 2 ): 115 – 130 . DOI: 10.1002/ohn.159 OpenUrl CrossRef PubMed 97. Messner AH , Walsh J , Rosenfeld RM , et al. Clinical consensus statement: ankyloglossia in children . Otolaryngol--head neck surg . 2020 ; 162 ( 5 ): 597 – 611 . DOI: 10.1177/0194599820915457 OpenUrl CrossRef PubMed 98. The Joint Committee on Infant Hearing . Year 2019 position statement: principles and guidelines for early hearing detection and intervention programs . JEHDI . 2019 ; 4 ( 2 ): 1 - 44 . OpenUrl 99. Desrosiers M , Evans GA , Keith PK , et al. Canadian clinical practice guidelines for acute and chronic rhinosinusitis . Allergy Asthma Clin Immunol . 2011 ; 7 ( 1 ): 2 . Published 2011 Feb 10. DOI:10.1186/1710-1492-7-2 OpenUrl CrossRef PubMed 100. Rosenfeld RM , Piccirillo JF , Chandrasekhar SS , et al. Clinical practice guideline (update): adult sinusitis executive summary . Otolaryngoloy Head Neck Surg . 2015 ; 152 ( 4 ): 598 – 609 . DOI: 10.1177/0194599815574247 OpenUrl CrossRef PubMed 101. Seidman MD , Gurgel RK , Lin SY , et al. Clinical practice guideline: allergic rhinitis . Otolaryngol Head Neck Surg . 2015 ; 152 ( 1 Suppl ): S1 - S43 . DOI: 10.1177/0194599814561600 OpenUrl CrossRef PubMed 102. Brozek JL , Bousquet J , Baena-Cagnani CE , et al. Allergic rhinitis and its impact on asthma (ARIA) guidelines: 2010 revision . J Allergy Clin Immunol . 2010 ; 126 ( 3 ): 466 – 476 . DOI: 10.1016/j.jaci.2010.06.047 OpenUrl CrossRef PubMed Web of Science 103. Fokkens WJ , Lund VJ , Hopkins C , et al. European Position Paper on rhinosinusitis and nasal polyps 2020 . Rhinology . 2020 ; 58 ( Suppl S29 ): 1 – 464 . Published 2020 Feb 20 . DOI: 10.4193/Rhin20.600 OpenUrl CrossRef PubMed 104. Kaplan A . Canadian guidelines for acute bacterial rhinosinusitis: clinical summary . Can Fam Physician . 2014 ; 60 ( 3 ): 227 – 234 . OpenUrl Abstract / FREE Full Text 105. Wormald PJ , Hoseman W , Callejas C , et al. The international frontal sinus anatomy classification (IFAC) and classification of the extent of endoscopic frontal sinus surgery (EFSS) . Int Forum Allergy Rhinol . 2016 ; 6 ( 7 ): 677 – 696 . DOI: 10.1002/alr.21738 OpenUrl CrossRef PubMed 106. Orlandi RR , Kingdom TT , Smith TL , et al. International consensus statement on allergy and rhinology: rhinosinusitis 2021 [published correction appears in Int Forum Allergy Rhinol. 2022 Jul;12(7):974] . Int Forum Allergy Rhinol . 2021 ; 11 ( 3 ): 213 - 739 . DOI: 10.1002/alr.22741 OpenUrl CrossRef PubMed 107. Butowt R , Bilinska K , Bartheld CS von. Olfactory dysfunction in COVID-19: new insights into the underlying mechanisms . Trends in Neurosciences . 2023 ; 46 ( 1 ): 75 – 90 . DOI: 10.1016/j.tins.2022.11.003 OpenUrl CrossRef 108. Grayson JW , Cavada M , Harvey RJ . Clinically relevant phenotypes in chronic rhinosinusitis . J Otolaryngol Head Neck Surg . 2019 ; 48 ( 1 ): 23 . DOI: 10.1186/s40463-019-0350-y OpenUrl CrossRef PubMed 109. Piccirillo JF , Payne SC , Rosenfeld RM , et al. Clinical consensus statement: balloon dilation of the sinuses . Otolaryngol Head Neck Surg . 2018 ; 158 ( 2 ): 203 – 214 . DOI: 10.1177/0194599817750086 OpenUrl CrossRef PubMed 110. Thamboo A , Kilty S , Witterick I , et al. Canadian Rhinology Working Group consensus statement: biologic therapies for chronic rhinosinusitis . J of Otolaryngol - Head & Neck Surg . 2021 ; 50 ( 1 ): 15 . DOI: 10.1186/s40463-021-00493-2 OpenUrl CrossRef 111. Pinkiewicz M , Dorobisz K , Zatoński T . A comprehensive approach to facial reanimation: a systematic review . J Clin Med . 2022 ; 11 ( 10 ): 2890 . DOI: 10.3390/jcm11102890 OpenUrl CrossRef PubMed 112. Anne S , Mims JW , Tunkel DE , et al. Clinical Practice Guideline: Opioid prescribing for analgesia after common otolaryngology operations-executive summary . Otolaryngol Head Neck Surg . 2021 ; 164 ( 4 ): 687 – 703 . DOI: 10.1177/0194599821996303 OpenUrl CrossRef PubMed 113. American Society of Anesthesiologists Task Force on Operating Room Fires , Caplan RA , Barker SJ , et al. Practice advisory for the prevention and management of operating room fires . Anesthesiology . 2008 ; 108 ( 5 ): 786 - 972 . DOI: 10.1097/01.anes.0000299343.87119.a9 OpenUrl CrossRef PubMed Web of Science 114. Wu Q , Wang X , Liang G , et al. Advances in image-based artificial intelligence in otorhinolaryngology–head and neck surgery: a systematic review . Otolaryngol--head neck surg . Published online June 8 , 2023 : ohn.391 . DOI: 10.1002/ohn.391 OpenUrl CrossRef 115. Bur AM , Shew M , New J . Artificial intelligence for the otolaryngologist: a state of the art review . Otolaryngol--head neck surg . 2019 ; 160 ( 4 ): 603 – 611 . DOI: 10.1177/0194599819827507 OpenUrl CrossRef PubMed 116. Gölaç H , Atalık G , Özcebe E , Gündüz B , Karamert R , Kemaloğlu YK . Vocal outcomes after COVID-19 infection: acoustic voice analyses, durational measurements, self-reported findings, and auditory-perceptual evaluations . Eur Arch Otorhinolaryngol . 2022 ; 279 ( 12 ): 5761 – 5769 . DOI: 10.1007/s00405-022-07468-7 OpenUrl CrossRef PubMed 117. Sommer DD , Cote D , McHugh T , et al. Revised recommendations from the CSO-HNS taskforce on performance of tracheotomy during the COVID-19 pandemic – what a difference a year makes . J of Otolaryngol - Head & Neck Surg . 2021 ; 50 ( 1 ): 59 . DOI: 10.1186/s40463-021-00531-z OpenUrl CrossRef 118. Position Statements . CSOHNS . Published June 29 , 2023 . Accessed June 29, 2023 . https://www.entcanada.org/cso/position-statements/ 119. Otolaryngology . Choosing Wisely Canada . Published June 29 , 2023 . Accessed June 29, 2023 . https://choosingwiselycanada.org/recommendation/otolaryngology/ 120. John R. de Almeida , Gordon H. Guyatt , Sachin Sud , et al. Management of Bell palsy: clinical practice guideline . CMAJ . 2014 ; 186 ( 12 ): 917 . DOI: 10.1503/cmaj.131801 OpenUrl FREE Full Text 121. Almeida JR de , Guyatt GH , Sud S , et al. Management of Bell palsy: clinical practice guideline . CMAJ . 2014 ; 186 ( 12 ): 917 – 922 . DOI: 10.1503/cmaj.131801 OpenUrl FREE Full Text 122. Baugh RF , Basura GJ , Ishii LE , et al. Clinical practice guideline: Bell’s Palsy executive summary . Otolaryngol Head Neck Surg . 2013 ; 149 ( 5 ): 656 – 663 . DOI: 10.1177/0194599813506835 OpenUrl CrossRef PubMed 123. Bilezikian JP , Brandi ML , Eastell R , et al. Guidelines for the management of asymptomatic primary hyperparathyroidism: summary statement from the fourth international workshop . J Clin Endocrinol Metab . 2014 ; 99 ( 10 ): 3561 – 3569 . DOI: 10.1210/jc.2014-1413 OpenUrl CrossRef PubMed Web of Science 124. Sauve L , Forrester AM , Top KA . Group A streptococcal pharyngitis: a practical guide to diagnosis and treatment . Paediatrics & Child Health . 2021 ; 26 ( 5 ): 319 – 319 . DOI: 10.1093/pch/pxab025 OpenUrl CrossRef PubMed 125. Fitzgerald N , Hussain S , Memon S , et al. Evaluating clinical and cost impacts of achieving 90% HPV vaccination rate against cervical cancer in Canada using the OncoSim cancer simulation model . JGO . 2018 ; 4 ( Supplement 2 ): 21s – 21s . DOI: 10.1200/jgo.18.27600 OpenUrl CrossRef 126. Khurshid M , Ansari I , Ahmad H , et al. Development of facial palsy following COVID-19 vaccination: A systematic review . Ann Med Surg (Lond) . 2022 ; 82 : 104758 . DOI: 10.1016/j.amsu.2022.104758 OpenUrl CrossRef 127. Irwin RS , French CL , Chang AB , et al. Classification of cough as a symptom in adults and management algorithms . Chest . 2018 ; 153 ( 1 ): 196 – 209 . DOI: 10.1016/j.chest.2017.10.016 OpenUrl CrossRef PubMed 128. Allori AC , Mulliken JB , Meara JG , Shusterman S , Marcus JR . Classification of cleft lip/palate: then and now . The Cleft Palate-Craniofacial Journal . 2017 ; 54 ( 2 ): 175 – 188 . DOI: 10.1597/14-080 OpenUrl CrossRef 129. Kahrilas PJ , Altman KW , Chang AB , et al. Chronic cough due to gastroesophageal reflux in adults: CHEST guideline and expert panel report . Chest . 2016 ; 150 ( 6 ): 1341 – 1360 . DOI: 10.1016/j.chest.2016.08.1458 OpenUrl CrossRef PubMed 130. Classification and approach to patients with functional voice disorders . DOI: 10.1177/000348948209100409 OpenUrl CrossRef PubMed 131. Canadian Headache Society Guideline for Migraine Prophylaxis: Supplement 2 . Canadian Journal of Neurological Sciences . 2012 ; 39 ( S2 ): i - 63 . DOI: 10.1017/S0317167100015109 OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted January 07, 2025. Download PDF Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following ENTAgents: AI Agents for Complex Knowledge Otolaryngology Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share ENTAgents: AI Agents for Complex Knowledge Otolaryngology Tsz Kin Chan , Ngoc-Duy Dinh medRxiv 2025.01.01.25319863; doi: https://doi.org/10.1101/2025.01.01.25319863 Share This Article: Copy Citation Tools ENTAgents: AI Agents for Complex Knowledge Otolaryngology Tsz Kin Chan , Ngoc-Duy Dinh medRxiv 2025.01.01.25319863; doi: https://doi.org/10.1101/2025.01.01.25319863 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Otolaryngology Subject Areas All Articles Addiction Medicine (568) Allergy and Immunology (863) Anesthesia (299) Cardiovascular Medicine (4423) Dentistry and Oral Medicine (443) Dermatology (382) Emergency Medicine (607) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1507) Epidemiology (15219) Forensic Medicine (30) Gastroenterology (1123) Genetic and Genomic Medicine (6587) Geriatric Medicine (667) Health Economics (997) Health Informatics (4524) Health Policy (1368) Health Systems and Quality Improvement (1612) Hematology (540) HIV/AIDS (1264) Infectious Diseases (except HIV/AIDS) (15910) Intensive Care and Critical Care Medicine (1103) Medical Education (623) Medical Ethics (145) Nephrology (667) Neurology (6588) Nursing (345) Nutrition (998) Obstetrics and Gynecology (1143) Occupational and Environmental Health (956) Oncology (3331) Ophthalmology (970) Orthopedics (369) Otolaryngology (420) Pain Medicine (435) Palliative Medicine (129) Pathology (663) Pediatrics (1690) Pharmacology and Therapeutics (691) Primary Care Research (710) Psychiatry and Clinical Psychology (5438) Public and Global Health (9218) Radiology and Imaging (2195) Rehabilitation Medicine and Physical Therapy (1369) Respiratory Medicine (1195) Rheumatology (593) Sexual and Reproductive Health (709) Sports Medicine (529) Surgery (709) Toxicology (99) Transplantation (289) Urology (265) (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9ff80e5a1bbb0db4',t:'MTc3OTQxMzQzMQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();