Otolaryngology, often known as ear, nose, and throat (ENT) or head and neck surgery, is a surgical specialty in medicine which focuses on the surgical and medical management of conditions of the head and neck [1]. Otolaryngology deals with a broad range of medical conditions, including disease of the upper airway, head and neck, ear, nose, sinuses, and skull base. The scope of otolaryngology spans multiple subspecialties, including pediatric otolaryngology, laryngology, facial and reconstructive plastic surgery, and neuro-otology [2].

The field of otolaryngology has expanded in both scope and practice, especially with the recent pandemic. With hospital utilization limits, otolaryngologists have been forced to triage and prioritize surgical care in efficient manners. The specialty has placed a major emphasis on changing how otolaryngology services are delivered in the operating room and outpatient setting to reduce the burden on the field [3]. These problems collectively served as a catalyst for AI in otolaryngology [3,4]. AI applications may supplement medical professionals in clinical and surgical domains and can provide substantial assistance to doctors when faced with challenging decisions, diagnoses, and treatment protocols [5].

AI aims to create inferences or predictions from new data while considering historical data [6]. The phrase “artificial intelligence” was first used by John McCarthy to refer to “the science and engineering of creating intelligent computer programs” [6]. Since AI is also a collection of numerous technologies meant to automate human intellectual processes, unique AI technologies are developed for certain practices [7].

The landscape of healthcare is changing across many different specialties because of AI's transformational capabilities. In radiology, minor discrepancies in medical images are found, such as early breast cancer in mammograms [8]. Moreover, surgeons can employ AI for accurate preoperative planning, including uses for customized knee replacements specifically in orthopedic surgery [8]. Pathologists can benefit from AI systems that increase the accuracy of tissue sample analysis, particularly in the diagnosis of skin cancers [9]. Moreover, in cardiology, by forecasting the risk of heart illness through patient data analysis, AI can help in early detection of disorders [8]. The combination of AI and genomics also makes it possible to create personalized cancer treatment plans, like those that modify drugs in accordance with a patient's genetic profile [8]. These are only some examples that show how AI is supplementing healthcare and removing certain barriers.

Specifically in otolaryngology, AI seems to be changing how healthcare is provided, both in clinical practice and patient care. AI is expected to have an impact on our specialty, perhaps freeing otolaryngologists from paperwork during office visits, improving skills in the operating room, and changing how we view diseases [10].

The most common applications of AI technologies in otolaryngology are in diagnosing and treating head and neck cancers [5]. AI has been used in conjunction with imaging to distinguish between normal head and neck tissue and thyroid cancer, and it has been found to be 97 % sensitive, 96 % specific, and 96 % accurate [5]. This technology can then allow for more precise resection surgical resection margins [11]. Furthermore, clinicopathological and genetic markers used in this technology can successfully predict the prognosis of established cancer in patients [5]. Another promising application of AI is automatic radiotherapy planning, which can enable the computation of three-dimensional tumor volumes and more accurate dose planning [12].

In otology AI can be used to determine the prognosis of acute sensorineural hearing losses, visualize endolymphatic hydrops, and also detect hearing loss phenotypes [13]. In addition, when AI programs are supplemented in speech analysis and video-stroboscopic images in laryngology, T1a glottic malignancies and other illnesses of the vocal cord can be detected with about 100 % accuracy [14]. While the use of AI in rhinology is limited, current research into paranasal sinuses, sinusitis diagnosis, and sinus diseases is ongoing [15].

Moreover, GPT-4, a natural language model, can transform patient education and research in otolaryngology. GPT 4 can efficiently summarize medical literature and even support professionals in staying up to date with current literature [16]. Although, it is important to make sure GPT-4 in healthcare mandates reliable and current data, considering the privacy and security aspects of patients as well [16]. This review highlights the development of AI, particularly GPT-4, in the field of otolaryngology.

GPT-4 can potentially alter otolaryngology procedures, notably in terms of diagnosis and preoperative planning. GPT-4 can offer doctors with up-to-date, understandable, and pertinent information, enabling educated treatment decisions and enhancing patient care [16]. This ability to provide insights improves patient outcomes while simultaneously lowering the likelihood of medical mistakes. GPT-4 can also enable medical personnel to make quick and sure judgments by providing real-time updates on treatment options and best practices [16].

Moreover, this technology has gained popularity for potential diagnosis and assisting users and physicians in self-triaging through human-like interactions. In low-income nations with a shortage of specialized care, GPT-4 may boost clinicians' confidence in diagnosis and hasten the start of appropriate treatment. It may also alert clinicians who are overlooking critical diagnoses and provide suggestions [17,18]. In a study at Queen Mary Hospital, GPT-4 showed promise in boosting diagnostic precision for elderly patients with postponed diagnoses [17]. GPT-4 also offered the most likely diagnosis or indicated differential diagnoses after reviewing the medical histories of six patients aged 65 or older who had a delay in a final diagnosis of more than a month, with an accuracy rate of 83.3% [17]. The therapeutic utility of GPT-4 in assisting doctors in difficult diagnostic instances is highlighted by this study, potentially improving patient outcomes, especially in environments with limited access to specialized treatment.

Regarding diagnosis, while not a substitute for radiologists, GPT-4 assists in the initial evaluation of radiological images by highlighting significant structures and anomalies and providing concise descriptions [18]. Radiomic AI programs, enhanced by GPT-4, can enhance image interpretation and pathology detection, promoting efficient and accurate diagnostics [19,20]. Notably, GPT-4's influence in otolaryngology can speed up the interpretation process of medical imaging diagnostics [15,17]. This can improve workflow effectiveness, decrease patient wait times, and eventually contribute to improved patient outcomes [15,17].

The incorporation of GPT-4 into preoperative planning and diagnostics has the potential to change otolaryngology. GPT-4 can improve surgical methods by providing accurate details about the surgical case, potential hazards, and supplemental approaches [21]. This data-driven decision-making in otolaryngology can improve patient-specific care, reduce complications, and create options for better surgical outcomes and patient satisfaction. Additionally on this topic, GPT-4 can be set up to create 3D models from patient-specific information, such as CT scans or MRI pictures [22]. This facilitates the conversion of data into forms suitable for 3D printing or virtual reality simulations [22], and offers surgeons virtual practice and a more realistic preoperative planning [22].

GPT-4 could offer real-time surgical navigation inputs during surgeries, such as highlighting anatomical structures and surgical steps to assist surgeons in making more precise choices [23]. Through intraoperative support, which can be accomplished by superimposing 3D models onto the patient's anatomy, GPT-4 can offer invaluable support in this goal [23]. This addition helps with locating and avoiding crucial structures during surgeries [24]. Additionally, this support promotes operation alignment with expectations, which is especially important for complex otolaryngology surgical procedures.

Take the difficult scenario of endoscopic sinus surgery, a routine otolaryngology operation. While performing this surgery, the surgeon must avoid delicate structures like the optic nerve and carotid artery by navigating the sinus passageways [24]. The surgeon may precisely determine the safest routes and potential pitfalls while using GPT-4's real-time guidance while viewing augmented 3D models of the patient's sinuses [24]. The addition of GPT-4 to this field has the potential to significantly improve surgical performance and improve patient outcomes through novel technology.

GPT-4 also offers critical insights regarding the postoperative recovery phase. Its ability to provide patients with useful information on what to expect after surgery, including pain levels, recovery times, and probable problems, proves to be essential [25]. Beyond providing information, GPT-4 makes a substantial contribution to the postoperative care planning process by aiding in activities like medication management, directing rehabilitation procedures, and providing discharge instructions [26].

In a study by Ayoub et al., GPT-4, was assessed as a medical source for producing postoperative instructions after common otolaryngology procedures [27]. The study evaluated instructions from GPT-4, Google Search, and an institution while focusing on pediatric otolaryngologic procedures like tympanostomy tube placement, tonsillectomy and adenoidectomy, inferior turbinate reduction, tympanoplasty, cochlear implant, neck mass resection, and microdirect laryngoscopy [27]. The greatest scores for procedure-specific items came from GPT-4 (97 %) and institutional instructions (97 %) whereas the lowest came from Google Search (72 %) [27]. GPT displayed effectiveness in improving patient awareness and creating postoperative instructions, especially for groups with poor levels of health or educational literacy [27]. Moreover GPT-4 can provide personalized rehabilitation strategies and early detection of challenges and complications for patients. GPT-4 can have the potential to help optimize the course of recovery and improve the quality of life for patients in the field of otolaryngology [28].

The ability of GPT-4 to process natural language is important for enhancing patient communication. Patients frequently experience confusion and anxiety because of their inability to comprehend complicated medical terminology and available treatments [29]. GPT-4 could streamline interactions for more meaningful exchanges and redefine patient communication. GPT-4 can help patients better comprehend their diseases, symptoms, diagnoses, and therapies by simplifying difficult medical information [18]. This approach enables patients to actively interact with healthcare professionals, which can lead to improved medical conditions and a general enhancement in the quality of care provided to patients [16]. For example, in gastroenterology, GPT-4's function in patient education results in higher treatment adherence [30]. By automating common chores, it frees up clinicians to focus on difficult areas of patient care [30].

GPT-4 also enables patients to engage in virtual consultations thanks to its smooth integration into telemedicine platforms [26]. For example, a patient recovering from a recent tonsillectomy can speak with GPT-4 via a telemedicine app to get suggestions on how to handle postoperative pain, ask about dietary restrictions, and get assurance about the recovery process. This type of on-demand assistance is especially helpful for those who live in rural places with little access to specialist medical care or for people who have trouble traveling because of physical limitations. The availability of GPT-4 for advice and communication acts as a link between patients and medical professionals, promoting thorough treatment even in difficult situations [31]. Lastly, the potential of GPT-4 to translate between languages would be a development in global healthcare systems [30]. Although this ability requires more improvement before deployment, it will eventually speed up and increase access to such services in multicultural and multilingual contexts for both patients and healthcare professionals [30]. The availability of GPT-4 for advice and communication acts as a link between patients and medical professionals, promoting thorough treatment even in difficult situations [31].

Otolaryngologists may quickly obtain current information from medical databases, research articles, and clinical guidelines using GPT-4. With this technology, colleagues who reside abroad can participate in virtual case discussions, present patient cases, share clinical outcomes, and ask for ideas and comments. Otolaryngologists can use GPT-4 to promote the sharing of best practices and therapeutic modalities, ensuring that evidence-based recommendations are followed throughout clinical practice. GPT-4 is incredibly useful for remote consultations since it enables collaboration with experts to give healthcare practitioners in underserved areas professional advice. Additionally, GPT-4 aids otolaryngologists with conference and presentation preparation by producing slide content, distilling study results, and providing communication advice [32]. GPT-4 supports the growth of professional networks by connecting otolaryngologists with peers, mentors, and subject-matter specialists, creating opportunities for cooperation and knowledge sharing.

GPT-4 can also act as a virtual research assistant in the world of research, providing important assistance throughout the process. Artificial intelligence models help researchers develop and implement research plans, carry out investigations, and analyze data. The involvement of GPT-4 in data analysis and interpretation can speed up the discovery of important insights from outcomes, increasing efficiency and efficacy [33]. This simplifies the study procedure and frees up researcher's time to focus on new ideas and innovations. The ability of GPT-4 to search and summarize the literature is a novel use in otolaryngology. Due to its natural language processing, it is possible to conduct searches of the medical literature and to condense pertinent results into formats that are simple to understand [33]. Access to the most recent discoveries and research benefits otolaryngologists and encourages proactive participation with innovations. This procedure also improves the effectiveness of healthcare delivery, which benefits patient outcomes [16].

AI integration into otolaryngology education can alter how students interact with and learn the complex intricacies of this specialty. GPT-4 and its virtual teaching features stand out as revolutionary tools that may alter how education in otolaryngology is delivered through providing students with tailored, in-depth information as well as interactive simulations that will make studying more engaging. GPT-4 has the potential to improve student engagement and increase learning in otolaryngology, although additional research is needed to confirm this [34].

GPT-4 in otolaryngology education includes providing medical students, aspiring otolaryngologists, and other healthcare professionals with up-to-date information on developments in this area. Additionally, GPT-4's ability to evaluate clinical competencies [35] emphasizes its crucial significance in enhancing otolaryngology education. Students in otolaryngology can use GPT-4 to delve into the specifics of the discipline and promote learning for better understanding. GPT-4 can help medical students and otolaryngology residents by providing them with specialized questions, quick answers, and access to specialized instructional materials. The model has the potential to improve long-term retention of the specialized subject matter in addition to making comprehension easier [36,37].

Otolaryngology could benefit greatly from the capabilities of GPT-4, especially when it comes to customizing treatment approaches for each patient. Consider a patient who has a history of chronic sinusitis; in this case, GPT-4 can be crucial in creating a highly individualized diagnostic and treatment plan. It considers important elements such the patient's age, medical background, and current sinus health. Since each patient's specific circumstances are considered, this individualized approach ensures that treatment recommendations precise and successful. For instance, GPT-4 can produce patient-friendly summaries of medical facts, enabling people to make better healthcare decisions [16,38].

The effects of GPT-4 on otolaryngology also include administrative task simplification, which boosts overall practice effectiveness. GPT-4 can effectively lighten the administrative load on medical staff by scheduling patients, delivering reminders on time, and coordinating calendars. Due to easier and more responsive healthcare exchanges, this resource allocation not only optimizes patient care but also increases patient satisfaction. To further enhance patients' overall experiences, GPT-4 can assist patients with administrative tasks like insurance questions or appointment rescheduling.

The adoption of GPT-4 AI systems could require collection of large amounts of personal data, including conversation logs, browsing histories, and location data. This functionality raises privacy issues due to the possibility of data misuse or unauthorized sharing with third parties. As a result, there is a need for increased caution in the acquisition and use of data by GPT-4 AI systems. In order to address these privacy concerns, calls for openness, user-defined restrictions, and adherence to data protection laws need to be accounted for [39].

GPT-4 may be used in the field of otolaryngology to get access to and control over private patient data. As a result, it becomes important to guarantee the security and confidentiality of this information in compliance with patient privacy rules and regulations. It is crucial to define the duties and responsibilities because GPT-4 functions as an AI model rather than a trained healthcare professional. This will allow for patients to get reliable and effective care with the privacy that is required. Patients have a right to know if GPT-4 is used in their treatment and to understand the extent of that use and the data it provides. Patients must get thorough education and give their informed consent before GPT-4 is included into their care, underlining the value of ethical practice and patient autonomy.

The principle of respect for autonomy intersects with fundamental ethical considerations in medical care, such as informed consent and patient confidentiality [40]. Consideration must be given to how to maintain patient confidentiality and privacy. Consent mechanisms are required for the integration of medical history in AI processes [41]. As a result, the conversation should move away from passive opt-in/opt-out terms and toward a method that allows for explicit and informed consent.

In addition, it is important to consider how to protect the autonomy of medical professionals in addition to patient concerns. Clinicians frequently make evidence-based decisions influenced by their familiarity with or deviation from new recommendations, personal experiences, and existing practices [42]. If clinicians decide to incorporate GPT-4 recommendations, the sheer exposure to the proposals could potentially steer them toward previously unresearched results or therapies [43]. This demonstrates the complex interaction between GPT-4 insights and the protection of the autonomy and judgment of medical personnel.

Moreover, legal concerns can arise if GPT-4 were used in the clinic and were to give incorrect medical advice in the field of otolaryngology [44]. Clinicians can still be held liable since AI does not have the same legal standing as a clinician. It becomes difficult to determine who is legally responsible when a patient is harmed, raising concerns about who should be held accountable—the patient, the otolaryngology clinic, or GPT-4. To define and assign responsibility for AI use, legal frameworks that are specific to otolaryngology are required. Although OpenAI has taken steps to address issues, users are currently responsible for errors. To protect practitioners and patients in the field of otolaryngology, regulations must be clear and spell out the responsibilities of developers, otolaryngologists, and others involved.

There are also significant ethical questions raised, particularly in the context of humanistic treatment [44]. Otolaryngologists use AI to improve the productivity and information accessibility of their practices. It's important to be aware that AI's expanding influence could unintentionally undermine the humanistic elements of medical care. Fundamental to otolaryngology, humanistic ethics stress the value of compassionate, patient-centered care, and honesty dealings with patients.

Healthcare practitioners can use AI in otolaryngology to supplement their care, but must defend the humanistic values that are fundamental to the profession. While GPT-4 and other AI tools can deliver information quickly and accurately, they are unable to offer patients the kind of emotional support and empathy that is essential to the doctor-patient interaction in otolaryngology. Therefore, it's important to make sure GPT-4 enhances rather than replaces this vital connection. To uphold trust and integrity as well as the humanistic principles that characterize otolaryngology treatment, adherence to evidence-based procedures and clinical recommendations become crucial. Otolaryngologists must consciously address these issues, using AI as a supplementary tool to improve their knowledge and communication while respecting the principles of humanism. By doing this, they can make sure that GPT-4 can enhance patient care while maintaining the fundamental values of compassion and empathy.

An in-depth assessment of GPT-4's function in the field of otolaryngology demands a thorough examination of both its reliability and efficiency [44]. Particularly in a medical profession where accuracy can mean the difference between life and death, validity is very important. To reduce the possibility of inaccurate information and its potential effects on patient health, it is crucial to make sure that GPT-4's responses follow accepted clinical principles and evidence-based procedures. However, due to the amount of data that GPT-4 creates, validating its output is a difficult task that consumes a lot of resources.

In addition, GPT-4 has limits when it comes current research, such as current clinical trials, research findings, or publications, which can make it less dependable for remaining up to date with the most recent data [45]. Given that its expertise is based on a fixed dataset and it might not have access to recently published or real-time data, it is less ideal for giving the most up-to-date information in domains that are continuously evolving [45]. This restriction can be especially challenging in professions like otolaryngology, where recommendations, therapies, and surgical techniques are frequently updated and improved. To give healthcare professionals and patients the required legal and ethical information, it is essential to establish thorough ethical guidelines. These rules will make it easier to use GPT-4 s potential while protecting people's safety.