AbstractThe rapid development of large language models such as Generative artificial intelligence (AI) and Chat generative pre-trained transformer (ChatGPT) presents new possibilities and innovative approaches across health care. This trend is no exception in the field of audiology, and the possibility of using artificial intelligence in various domains-such as education, clinical decision-making assistance, patient counseling and audiologic rehabilitation program-is being actively explored. For the effective introduction of Generative AI, prompt engineering is a key factor beyond simple technology implementation. In addition, ethical considerations based on the use of AI must be integrated in fields dealing with patients' personal information, such as audiology. Institutional regulation and expert-led ethical guidelines should be prepared in advance to prevent potential risks. AI should be understood not as a replacement for human experts, but as a companion that complements and enhances clinical judgment and educational activities. Collaboration with generative AI is no longer a choice, but a necessity-and how wisely we integrate and design it will be a pivotal factor in shaping the future of audiology.
INTRODUCTION한국뿐만 아니라 전 세계는 저출산과 의학 기술의 눈부신 발전이라는 두 가지 요인으로 인해 급격하게 고령화 사회로 가고 있다. 이와 같은 인구 구조의 변화는 노화성 질환의 증가로 이어지고 있으며 그중에서도 노화성 난청(age-related hearing loss, presbycusis)은 만성 질환 중 가장 보편적인 문제 중 하나이다. 노인의 청력 저하는 단순한 의사소통의 어려움을 넘어 인지 기능 저하, 사회적 고립, 우울증 등의 다양한 문제를 야기할 수 있어 조기 난청의 진단과 적절한 청력의 재활이 필수적이다.
고령화 사회에서의 노화성 난청 문제가 사회적 문제로 확산될 수 있음을 일찍이 인지한 국가들은 여러 가지 정책을 만들어 지원을 하고 있다. 우리나라의 경우에도 2015년부터 보청기에 대한 보조금의 지원을 최대 131만 원으로 상향 조정하여 지원하고 있다(Ministry of Health and Welfare, 2020). 영국은 National Health Service를 통해 청각 관련 서비스를 제공하고 있으며 2015년에는 action plan on hearing loss를 발표하여 청각 손실에 대한 국가적 대응 전략을 수립하였다 (NHS England, 2015). 또한 Royal National Institute for Deaf People은 청각장애인을 위한 다양한 재정 지원과 보조 기기 제공 프로그램을 운영하고 있다(RNID, 2025). 미국의 경우, Over-the-Counter Hearing Aid Act of 2017을 통해 보청기의 접근성을 높였으며(Warren, 2017), National Institue of Heatlh 산하 National Institute on Deafness and Other Communication Disorders는 노화성 난청에 대한 연구를 지원하고 인공지능을 활용한 청각재활 기술 개발에도 투자하고 있다(National Institute on Deafness and Other Communication Disorders, 2023).
이러한 배경 속에서 청각학은 단순한 청력 측정과 보청기 적합에 국한되지 않고 노인의 전반적인 삶의 질을 향상시키는 중요한 보건의료 분야 중 한 분야로 주목받고 있다. Freeman et al. (2025)은 "청각학의 미래는 생존을 넘어 번영으로 나아가야 한다"고 강조하며 인구 변화와 기술 발전이 맞물리는 이 시점이 청각학 분야에 새로운 전환점이 될 수 있다고 하였다. 특히 디지털 청각기기, 원격 상담 및 적합, 그리고 임상 데이터 기반 의사결정 보조 시스템 등 기술의 도입은 청능사(audiologist)의 역할을 확장시키고 있으며 이에 따라 청각 서비스의 접근성과 효율성 또한 빠르게 변화하고 있다.
이러한 기술적 전환의 중심에는 artificial intelligence (AI), 생성형 AI와 같은 대형 언어 모델(large language model, LLM) 기반 도구들의 활용 가능성이 주목받고 있다. Chat generative pre-trained transformer (ChatGPT)와 같은 챗봇 AI는 LLM을 통해 딥러닝을 하며 방대한 데이터를 활용하여 요약하고 생성하며 예측하는 능력을 갖고 있다. LLM 모델의 경우 인터넷상의 다양한 출처를 통해 수집한 인공 신경망(artificial neural networks)을 훈련하여 만들어지며 LLM은 학습한 데이터에서 패턴과 상관관계를 파악하여 일관성 있고 논리적인 텍스트를 생성할 수 있게 된다. 이러한 언어 이해 및 생성 능력은 여러 응용 분야에서 큰 이점을 제공하고 있다. 이미 다양한 보건의료 분야에서 ChatGPT와 같은 생성형 AI는 상담, 교육, 문서화, 자기주도 학습 등 다양한 영역에서 보완적 도구로 사용 가능성이 제기되고 있다(Patil et al., 2024).
본 단보에서는 변화하는 환경과 그 속에서 부각되고 있는 AI 기술의 도입 가능성을 검토하고 생성형 AI의 구조와 프롬프트 엔지니어링(prompt engineering)의 개념과 우리가 앞으로 나아가야 할 연구와 임상에서의 활용에 대하여 제안하고자 한다.
청각학 임상과 교육에서의 생성형 인공지능(ChatGPT) 활용 가능성임상에서의 판단 결정의 지원청각학의 분야에서도 이러한 인공지능을 청각재활에 적용하기 위한 여러 연구가 시도되었다. 청각학적 지식을 검증하기 위하여 타이완 청각학 자격시험을 ChatGPT 4 (Open AI, L.L.C., San Francisco, CA, USA)에게 제시하여 얼마나 논리적으로 응답할 수 있는가를 평가하였다. 타이완 청능사 자격 시험은 총 299문제로 구성되며 60% 이상의 정답률을 보일 시 합격이다. 본 연구에서 ChatGPT는 75%의 정답률을 보였으며 오답의 원인을 분석했을 때 논리적 판단의 오류가 아닌 잘못된 정보 혹은 상충된 정보로 인한 것으로 나타났다. 본 연구 결과는 정보 제공 도구로서 ChatGPT의 활용 가능성을 보여주지만 개인의 병력, 동반 질환, 문화적 배경 등을 고려한 개인 맞춤형 판단은 불가능하다는 한계를 지닌다. 또한 데이터 편향, 개인정보 보호, 법적 책임 부재, 오진 가능성 등의 측면에서 실제 임상 적용에는 전문가의 보완적 판단이 반드시 수반되어야 함을 제안한다(Wang et al., 2024b).
자가 청력 모니터링 도구와의 통합청력 자가 측정은 디지털 헬스케어 환경에서 점차 중요해지고 있으며 ChatGPT를 해당 도구에 통합하려는 시도가 이루어지고 있다. Pastucha et al.(2024)는 청력 모니터링 앱과 ChatGPT를 결합하여 개인이 전문가에게 상담을 받아야 하는지에 대하여 ChatGPT가 판단할 수 있는가를 평가하였다. 연구자는 12개월간 1,000개의 청력도가 포함된 난청인 시뮬레이션 데이터를 사용하였으며 ChatGPT가 데이터를 분석하고 전문가에게 상담할 시기를 결정하게 하였다. 이 결과를 5명의 전문가의 판정과 비교해 본 결과 99% 일치하는 결과를 보였다. 본 연구에서는 인공지능이 난청의 조기 선별 및 전문가 상담을 유도하여 적절한 청각 재활을 이끌 수 있다는 장점을 보여주었지만 연구에 사용한 자료가 실제의 청력 데이터가 아닌 난청을 시뮬레이션이라는 한계점이 있다. 또한 청각재활에서 인공지능은 재활의 결정의 어떠한 책임을 질 수 없으며 개개인의 배경 정보 등을 모두 고려하지 못한다는 점에 있어서 전문가의 판단은 대체할 수 없으며 보조적 수단으로 사용해야 한다고 제안하였다.
관련 정보 제공 및 정서적 지원개인과의 의사소통 능력 역시 생성형 AI의 주요 강점 중 하나이다. ChatGPT는 단순 정보 제공을 넘어 정서적 지지 및 건강 문해력 향상 측면에서 주목받고 있다. Jedrzejczak et al.(2024)은 이명(tinnitus) 관련 정보를 ChatGPT에 입력하고 신뢰성 있는 답변을 제공하는지에 대하여 전문가들이 판단하는 연구를 실시하였다. 연구자들은 이명에 대한 질문 중 총 10개의 질문을 선정하였으며 질문의 답을 6명의 청각 전문가들이 5점 척도로 평가하였다. 이때 모든 질문의 답에서 4점 이상의 점수를 획득하여 ChatGPT 답변이 신뢰성이 있음을 증명하였다. 또한 ChatGPT의 답변은 비전문가가 이해할 수 있도록 친근감 있게 쉬운 설명의 답변으로 나타났다. 그러나 일부의 답변은 모호한 표현이나 혼동을 줄 수 있는 답변이 포함되어 있어 전문가 상담에 보조적인 수단만으로 활용을 제안하였다.
청각학 교육에서의 활용ChatGPT는 청각학 교육 분야에서의 활용 가능성이 제기되었다. Ratuszniak et al.(2025)은 아동 청각학에 대한 질문을 ChatGPT에게 질문하고 청각학 전공 학생과 전문가에게 ChatGPT의 답변을 정확성, 완전성, 관련성 그리고 언어 표현력의 영역에서 평가하도록 하였다. 학생들의 경우 ChatGPT의 응답에 높은 점수를 부여하였지만 전문가의 경우 엄격한 기준으로 평가하여 상대적으로 학생들보다 낮은 점수로 평가한 것으로 나타났다. 또한 영어로 한 질문의 답이 폴란드어로 답변한 경우보다 문장의 완성도 측면에서 높게 나타나 아직 ChatGPT의 비영어권의 답변이 완전하지 않다고 보고하였다. 이 결과는 현재는 ChatGPT는 학습보조 수단으로서의 활용은 가능하나 아직 완벽하게 전문가를 대신할 정도의 수준은 도달하지 못했음을 의미한다.
청각훈련 도구로의 활용ChatGPT는 자연어를 이해하고 생성할 수 있는 능력이 있어 사용자 수준에 맞는 맞춤형 문장을 제공하거나 청취훈련에 사용할 수 있는 환경음, 문장, 이야기 등을 생성하여 청능훈련의 도구로의 활용이 가능할 수 있다. 최근 AI 기반 훈련 도구의 개발이 활발하게 이루어지고 있다. 예를 들어 Starkey Hearing Techologies에서 개발한 HearCoach의 경우 AI 기반 청각 환경을 분석할 수 있으며 다양한 언어 번역이 가능하여 사용자가 실시간 의사소통이 가능하다(Starkey Hearing Technologies, 2025). Amptify는 디지털 청각재활 플랫폼으로 사용자의 청각 손실 정도, 인지 능력, 학습 속도를 분석하여 개인 맞춤형 청능훈련 프로그램을 제공하며 AI를 활용하여 사용자의 진행 상황을 실시간 모니터링이 가능하다. 또한 AI를 활용하여 사용자 간의 상호작용을 촉진할 수 있는 기능을 가지고 있다(Amptify Inc., 2025).
이처럼 AI 기반 훈련 도구의 기술은 청능훈련을 개인 맞춤형이 가능하며 접근성의 향상, 훈련 지속성의 가능성을 제시한다. AI 기반 청능훈련은 기존의 청능훈련 도구를 보완하거나 대체할 수 있을 것이다.
앞으로의 과제: 프롬프트 엔지니어링(prompt engineering)의 기술적 과제와 청각학 적용 전략지금까지 많은 생성형 AI의 도입이 청각학에 본격적으로 활용되지 않았지만 교육, 임상, 재활 전반에 생성형 AI의 도입 가능성은 충분하다. 본격적인 활용에 앞서 우리는 프롬프트 엔지니어링에 관한 연구가 필요하다. 프롬프트 엔지니어링은 상용자가 LLM에게 정확하고 일관된 답변을 받기 위해서 질문을 만드는 과정을 의미한다. 이 질문은 단순하게 질문이 아니며 모델의 역할(role prompting), 문맥 정보, 출력 형식 지시, 목표 대상 등을 명확하게 표현하면서 AI에게 인간의 수준의 답변을 얻기 위한 기술적인 전략이다.
White et al.(2023)에 따르면 ChatGPT는 구조적으로 설계된 프롬프트를 사용할 때 구체적이고 정확한 답변이 추출 가능하며 Kung et al.(2023)의 질의에 역할을 명시할 경우(예: you are a certified audiologist), 일반 질의보다 정보의 깊이, 표현의 일관성을 높이는 것으로 보고하였다. 이러한 프롬프트 엔지니어링에 관한 연구 결과는 정보의 정확한 표현과 신뢰도가 매우 중요한 청각학을 비롯한 보건의료 분야에 생성형 AI를 본격적으로 도입하기 위해서는 프롬프트에 대한 구조적 설계 연구가 핵심 변수임을 보여준다(White et al., 2023). 이러한 구조화된 프롬프트에 관한 연구는 일관성 있는 결과를 얻어 연구, 임상 교육 등에서 반복해서 사용 가능하며 AI로부터 질문보다는 더 정밀하고 유용한 응답을 얻을 수 있게 성능을 향상시킬 수 있다. 또한 반복 업무 자동화 기능을 통해 시간을 절약할 수 있고 의료, 교육 등에서 특화된 응답을 유도하여 전문성을 유지할 수 있도록 할 수 있다.
다음은 OpenAI(2025)의 ChatGPT를 활용하여 청각학 관련 프롬프트 엔지니어링 생성한 예시이다.
예시 1) 청력 검사 해석 요청 및 상담 유도: You are a licensed audiologist. A 72-year-old patient has a hearing threshold of 55 dB HL at 1 kHz and 70 dB HL at 4 kHz in both ears. Based on this audiogram, summarize the degree of hearing loss and recommend next steps.
예시 2) 보청기 착용 적합성 판단: You are an audiologist. A patient has moderate sensorineural hearing loss and complains about difficulty hearing in noisy environments. Would you recommend hearing aids, and if so, which features should be prioritized?
예시 3) 이명에 대한 설명: You are an audiologist. Please explain what tinnitus is to a patient in plain language, including common causes and how it can be managed at home.
예시 4) 난청 조기 발견 교육: You are a hearing health educator. Create a short explanation to help older adults understand the early signs of hearing loss and why early testing is important.
예시 5) 노화성 난청 환자에게 보청기 적응 교육: You are an audiologist. Provide a step-by-step counseling script for helping an elderly patient adapt to new hearing aids, focusing on comfort and realistic expectations.
Kuka(2023)와 Ouyang et al.(2022)이 제안한 구조화된 프롬프트의 주요 요소이다. 1) You are a... → 전문가 역할 부여(Kuka, 2023). 2) Explain to a patient/parent → 대상자 설정(Kuka, 2023). 3) Using simple language → 언어 수준 조정(Kuka 2023). 4) Recommend next steps → 의사결정 유도(Ouyang et al., 2022).
다음은 위의 연구에서 제시한 사항을 청각학에 적용하여 프롬프트를 구조화해서 질문하였을 때와 그렇지 않았을 때의 답변의 차이를 보여준 예시이다(OpenAI, 2025).
1) 프롬프트 A (단순한 질문).질문: What is tinnitus?
답: Tinnitus is the perception of noise or ringing in the ears. It is a common problem that affects about 15 to 20 percent of people.
이러한 단순한 질문은 백과사전식의 단순한 답변만을 생성한다.
2) 프롬프트 B (역할+대상+어조 포함).질문: You are an audiologist (전문가의 역할 부여). Explain tinnitus to a parent whose child has just been diagnosed, using simple and reassuring language (어조 포함).
답: Hello! I understand that your child has recently been diagnosed with tinnitus, and that can sound a little scary at first - but let me reassure you. Tinnitus means your child is hearing a sound, like ringing, buzzing, or humming, even when there’s nothing around making that noise. It’s not a disease, and in most cases, it’s not dangerous.
Children may describe it in different ways, and sometimes they don’t mention it until it bothers them. The good news is, tinnitus in children often gets better on its own, especially when the cause - like an ear infection or hearing issue - is treated.
What’s most important is to help your child feel calm and supported. We can work together to monitor how they’re feeling, check their hearing, and explore ways to make the sound less noticeable if needed. You’re not alone, and we’ll guide you every step of the way.
프롬프트에 역할, 대사, 어조를 포함시켰을 때 좀 더 이해하기 쉽고 감정적으로 안정을 유도하며 대상 맞춤으로 답변이 생성됨을 알 수 있다.
3) 프롬프트 C (임상 판단 유도)질문: You are a licensed audiologist (전문가 역할 부여). A 50-year-old patient presents with unilateral tinnitus lasting over 3 months (대상자 설정). No hearing loss is reported. What is your clinical interpretation, and should this patient be referred for further evaluation (의사결정 유도)?
답: Based on the information provided, a 50-year-old patient experiencing unilateral tinnitus (tinnitus in one ear only) lasting for more than 3 months - and without any reported hearing loss - warrants further clinical evaluation.
While tinnitus can be benign, persistent unilateral tinnitus raises the possibility of underlying retrocochlear pathology, such as a vestibular schwannoma (acoustic neuroma), even in the absence of measurable hearing loss. It is important to rule out any serious causes.
Therefore, I would recommend referring this patient to an otolaryngologist (ear, nose and throat specialist) for further assessment. A comprehensive audiologic evaluation should also be performed if it hasn’t been done already, possibly followed by an magnetic resonance imaging to evaluate the internal auditory canal. Early detection is key for appropriate management.
전문적 임상 판단 포함 및 의학적 근거 제시가 된 답변을 받을 수 있음을 알 수 있다.
4) 프롬프트 D (구조화된 프롬프트의 요소 모두 포함)질문: You are an experienced audiologist (전문가 역할 부여). Please explain to the parent of a 4-year-old child who has just been diagnosed with moderate bilateral sensorineural hearing loss (대상자 설정). Use simple language (언어 수준 조정) and recommend the next steps they should take for their child’s hearing and communication development (의사결정 유도).
답변: Hello, I'm glad you came in today. I understand it can be overwhelming to hear that your child has hearing loss, but you're not alone - and there are many ways we can help.
Your child has moderate hearing loss in both ears, which means they may miss out on some everyday sounds, especially soft speech. This can affect how they learn to speak and interact with others.
Here’s what I recommend. 1) Hearing aids: these small devices can make sounds clearer and louder for your child. We can fit and adjust them to meet your child’s needs. 2) Speech and language therapy: a speech therapist can help your child develop strong communication skills. 3) Regular follow-ups: we’ll schedule visits to check your child’s hearing and make sure everything is working well. 4) Early intervention programs: these are special services for young children with hearing loss that support their development at home and school.
The earlier we begin, the better the outcomes. You're taking an important step just by being here - and we’ll guide you through the rest.
좀 더 쉬운 용어를 사용하여 구체적으로 답변이 생성되었음을 알 수 있다.
최근 보건의료 분야에서는 프롬프트 엔지니어링의 기법을 체계적으로 검토하고 의료 AI 시스템의 정확성과 일관성을 높이기 위한 연구가 활발히 이루어지고 있다(Zaghir et al., 2024). 이러한 연구들은 치료의 지침을 일관성을 향상시키고(Wang et al., 2024a), 임상 정보의 추출 기능을 향상 등에 이용 가능하다(Sivarajkumar et al., 2024). 이러한 연구들은 프롬프트 엔지니어링이 단순한 응답 생성 이상의 의미를 가지며 의료 분야에서 정확성, 공감성, 임상적 활용 가능성을 높이는 핵심 전략임을 시사한다. 다만 모든 적용에는 전문가의 검증과 윤리적 고려가 병행되어야 한다.
현재 대부분의 LLM은 영어로 학습을 하였기 때문에 한국어 프롬프트에 대해서는 표현 뉘앙스, 문법 구조, 의미 해석 등에 있어 한계를 가지고 있다. 또한 한국어로 질문하였을 때 일관적인 답변을 생성해 내는 데에도 문제가 있는 것으로 나타났다(Park et al., 2023). 한국어뿐만 아니라 비영어권에서의 ChatGPT의 답변의 경우 데이터 품질에도 차이가 있을 수 있다(Lai et al., 2023; Zhang et al, 2023). 청각학과 같은 보건의료 분야에서는 개인에게 맞춤형 데이터를 이용해야 하기 때문에 언어의 정확성이 더욱 요구된다. 이러한 연구를 근간으로 하여 다양한 한국어 청각학 시나리오를 바탕으로 한 표준화된 한국어 프롬프트 템플릿을 개발하고 AI의 반응을 정량적으로 평가하는 검증모델에 대한 연구가 필요할 것이다.
DISCUSSIONS본 단보에서는 청각학 분야에서 생성형 인공지능인 ChatGPT와 같은 LLM이 갖는 활용 가능성에 대하여 알아보았다. 선행 연구를 통해서 청능사의 교육, 상담 및 자가 모니터링, 청능훈련 도구로서 ChatGPT가 활용 가능함을 확인하였다. 이러한 활용이 적극적으로 수행될 수 있기 위해서는 프롬프트 엔지니어링의 설계가 필요하다. 또한 현재 영어로 학습한 ChatGPT에게 한국어로 된 체계적인 프롬프트의 설계는 우리가 LLM의 활용에 앞서 넘어야 할 산이 될 수도 있다. 그러나 앞으로 이러한 인공지능은 우리 보건의료 분야에서 꼭 활용될 수단인 만큼 우리 청각학 분야에서 빠르게 연구가 진행되어야 할 것이다. 또한 이와 함께 본 단보에서는 다루지 않았지만 사생활 침해 등 윤리적 문제에 대한 명확한 가이드라인과 규제가 마련되어야 한다. ChatGPT 등을 보건의료 분야에 활용할 시 오진 시 법적 책임 소재의 불명확함, 알고리즘 편향 문제, 비밀 보장 미흡 등의 다양한 윤리 사생활 문제가 나타날 수 있다(Haltaufderheide & Ranisch, 2024; Wang et al., 2023; Wei et al., 2024). 이때 AI의 역할이 전문가의 교육 활동과 임상적 판단을 대체하는 것이 아닌 보완하는 방향으로 나아가야 한다.
청각학의 실무 현장 및 원격 재활 영역에서의 생성형 AI의 사용은 선택이 아니며 필수이다. 앞으로 생성형 AI와의 상생에 있어 사회적 신뢰를 동시에 추구하는 전략이 필요하며 책임 있는 생태계 구축을 하는 것이 중요하겠다. 우리는 이제 인공지능과 현명하게 함께 살아가는 방법을 고민해야 할 때이며 이는 향후 청각학의 미래, 더 나아가 보건의료의 미래를 결정짓는 중요한 계기가 될 것이다.
NotesREFERENCESAmptify Inc. (2025, Mar 1). Amptify Digital Hearing Health Program. Amptify Inc.: Retrieved from https://amptify.com/.
Freeman, B. A., Danesh, A. A., Jones, K., & Danesh, F. (2025). The future of audiology surviving and thriving. Audiology Today, 37(2), 29.
Haltaufderheide, J. & Ranisch, R. (2024). The ethics of ChatGPT in medicine and healthcare: A systematic review on large language models (LLMs). Nature Partner Journals Digital Medicine, 7(1), 183.
Jedrzejczak, W. W., Skarzynski, P. H., Raj-Koziak, D., Sanfins, M. D., Hatzopoulos, S., & Kochanek, K. (2024). ChatGPT for tinnitus information and support: Response accuracy and retest after three and six months. Brain Sciences, 14(5), 465.
Kuka, V. (2023, Mar 1). Understanding Prompt Structure: Key Parts of A Prompt. Learn Prompting. Retrieved from https://learnprompting.org/docs/basics/prompt_structure.
Kung, T. H., Cheatham, M., Medenilla, A., Sillos, C., De Leon, L., Elepaño, C., et al. (2023). Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLoS Digital Health, 2(2), e0000198.
Lai, V. D., Ngo, Nf. T., Veyseh, A. P. B., Man, H., Dernoncourt, F., Bui, T., et al. (2023). ChatGPT Beyond English: Towards A Comprehensive Evaluation of Large Language Models on Multilingual Learning. Findings of the Association for Computational Linguistics. Retrieved from https://aclanthology.org/2023.findings-emnlp.878/.
Ministry of Health and Welfare. (2020, Jun 30). Enhancing Hearing Aid Reimbursement System for the Hearing-Impaired. Ministry of Health and Welfare. Retrieved from https://www.mohw.go.kr/react/al/sal0301vw.jsp?PAR_MENU_ID=04&MENU_ID=0403&CONT_SEQ=354146.
National Institute on Deafness and Other Communication Disorders. (2023, Mar 1). Age-Related Hearing Loss. National Institutes of Health. Retrieved from https://www.nidcd.nih.gov/health/age-related-hearing-loss.
NHS England. (2015, Mar 1). Action Plan on Hearing Loss. National Health Service. Retrieved from https://www.england.nhs.uk/wpcontent/uploads/2015/03/act-plan-hearing-loss-upd.pdf.
OpenAI. (2025, Mar 1). ChatGPT (May 6 version). OpenAI. Retrieved from https://chat.openai.com/chat.
Ouyang, L., Wu, J., Jiang, X., Almeida, D., Wainwright, C., Mishkin, P., et al. (2022). Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35, 27730-27744.
Park, D., An, G. T., Kamyod, C., & Kim, C. G. (2023). A study on performance improvement of prompt engineering for generative AI with a large language model. Journal of Web Engineering, 22(8), 1187-1206.
Pastucha, M., Ratuszniak, A., Ganc, M., Piłka, E., Drohobycka, I., Skarżyński, H., et al. (2024, Dec 16). ChatGPT As A Decision-Support Tool for Better Self-Monitoring Of Hearing. medRxiv. Retrieved from ttps://www.medrxiv.org/content/10.1101/2024.12.16.24319073v1.
Patil, R., Heston, T. F., & Bhuse, V. (2024). Prompt engineering in healthcare. Electronics, 13(15), 2961.
Ratuszniak, A., Gos, E., Lorens, A., Skarzynski, P. H., Skarzynski, H., & Jedrzejczak, W. W. (2025). Performance of ChatGPT in pediatric audiology as rated by students and experts. Journal of Clinical Medicine, 14(3), 875.
RNID. (2025, Jan 28). Benefits. Royal National Institute for Deaf People. Retrieved from https://rnid.org.uk/information-and-support/benefits/.
Sivarajkumar, S., Kelley, M., Samolyk-Mazzanti, A., Visweswaran, S., & Wang, Y. (2024). An empirical evaluation of prompting strategies for large language models in zero-shot clinical natural language processing: Algorithm development and validation study. Journal of Medical Internet Research, 12, e49858.
Starkey Hearing Technologies. (2025, Mar 1). Hear Coach 3.3.0. Starkey Hearing Technologies. Retrieved from https://hear-coach-ios.soft112.com/.
Wang, C., Liu, S., Yang, H., Guo, J., Wu, Y., & Liu, J. (2023). Ethical considerations of using ChatGPT in health care. Journal of Medical Internet Research, 25, e48009.
Wang, L., Chen, X., Deng, X., Wen, H., You, M., Liu, W., et al. (2024a). Prompt engineering in consistency and reliability with the evidence‑based guideline for large language models. Nature Partner Journals Digital Medicine, 7, 41.
Wang, S., Mo, C., Chen, Y., Dai, X., Wang, H., & Shen, X. (2024b). Exploring the performance of ChatGPT-4 in the Taiwan audiologist qualification examination: Preliminary observational study highlighting the potential of AI chatbots in hearing care. Journal of Medical Internet Research Medical Education, 10(1), e55595.
Warren, E. (2017, Mar 21). Over-The-Counter Hearing Aids: One-Pager. United States Senate. Retrieved from https://www.warren.senate.gov/files/documents/3_21_17_Hearing_Aids_One_Pager.pdf.
Wei, Q., Yao, Z., Cui, Y., Wei, B., Jin, Z., & Xu, X. (2024). Evaluation of ChatGPT generated medical responses: A systematic review and meta analysis. Journal of Biomedical Informatics, 151, 104620.
White, J., Fu, Q., Hays, S., Sandborn, M., Olea, C., Gilbert, H., et al. (2023, Mar 1). A Prompt Pattern Catalog to Enhance Prompt Engineering with Chatgpt. arXiv Preprint. Retrieved from chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://omekas-test.sba.unipi.it/files/original/9473424cea8d562f876a4bca4bedd9e2336910af.pdf.
Zaghir, J., Naguib, M., Bjelogrlic, M., Névéol, A., Tannier, X., & Lovis, C. (2024). Prompt engineering paradigms for medical applications: Scoping review. Journal of Medical Internet Research, 26, e60501.
Zhang, X., Li, S., Hauer, B., Shi, N., & Kondrak, G. (2023, Mar 1). Don’t Trust ChatGPT When Your Question is Not in English: A Study of Multilingual Abilities And Types of LLMs. arXiv Preprint, Retrieved from https://doi.org/10.48550/arXiv.2305.16339. arxiv.org.
|
|