Artificial intelligence (AI) and data science are being rapidly integrated into various aspects of healthcare, including diagnostic support, treatment optimization, drug discovery, and remote patient monitoring. Recent advancements in deep learning, coupled with the accumulation of largescale medical data, have enabled highly accurate disease detection from retinal fundus images, endoscopy, and CT/MRI scans. Natural language processing applied to electronic health records (EHRs) is also achieving practical success in predicting prognosis and evaluating patient risk. Furthermore, emerging techniques such as reinforcement learning and causal inference are advancing the optimization of individualized treatment strategies, while generative models and federated learning are drawing attention for their potential in privacy-preserving AI development. In the field of laryngology and tracheoesophageal medicine, AI is being applied to the analysis of swallowing and cough sounds, intraoperative image recognition, and voice-based diagnostics. These innovations are expected to support diverse applications, including home care and preoperative planning. This article provides a comprehensive overview of the current applications of AI and data science in medicine, highlights pioneering use cases in the tracheoesophageal domain, and discusses key challenges and future directions for realizing AI-integrated, patient-centered clinical practice.

Laryngectomy still plays an important role in the treatment of head and neck cancer. However, laryngectomized patients experience several difficulties in their lives because of the loss of their original voice. To overcome this problem, we have started the “Save the Voice” project, in which the original voices of patients who planned to undergo laryngectomy were recorded and converted using a voice conversion system. A multicenter prospective observational study has been performed in Japan, and over 40 patients' original voices have been recorded. Also, we have developed recording and converting applications for laryngectomized patients to convert the electrolarynx voice to the original voice. Preoperative recording of the patient's original voice is essential to regenerate the voice after laryngectomy. Collaboration between clinicians and researchers will enable swift implementation of the technology in real-world settings.

Flexible endoscopic evaluation of swallowing (FEES) can be easily performed in the outpatient setting, in the home, and in elderly care facilities. It is widely recognized as a common instrumental evaluation that plays an important role in diagnosing swallowing impairment. However, expertise and experience are required for appropriate evaluation of the swallowing function in order to assess its dynamics during momentary swallowing movements. Thus, assessments can vary among evaluators depending on said expertise and experience. Therefore, the evaluation and interpretation of FEES findings may differ from evaluator to evaluator, even if the findings come from the same subject. It has also been reported that experienced evaluators safely provided texture-modified diets closer to normal diets for patients with swallowing impairment, compared to inexperienced evaluators. The shortage of experts in the field of swallowing in an aging society has become a social problem; however, becoming an expert is not easy and requires much time and effort. As such, in order to make effective use of limited medical resources, it is necessary to provide support for inexperienced evaluators when they are assessing FEES findings. Recently, the usefulness of artificial intelligence (AI)-based imaging, endoscopy, and pathology evaluation assistance systems has been reported. We have developed an AI-assisted computer-aided diagnosis (AI-assisted CAD) system for the evaluation of FEES in collaboration with researchers specializing in AI. The present review discusses the usefulness of AI-assisted CAD for FEES and its future perspectives.

In the field of imaging diagnosis, AI-assisted diagnostic support systems have been introduced to enhance diagnostic accuracy and improve workflow efficiency. This paper presents the utilization of AI in chest imaging at our institution. For chest radiographs, we have implemented ClearRead XR and CTR-AID, while for CT imaging, we use ClearRead CT-VS and SYNAPSE SAI viewer. These software solutions contribute to lesion detection and diagnostic efficiency. Although AI usage may introduce biases, understanding the characteristics of each software and utilizing them appropriately enables AI to function as a valuable diagnostic support tool.

Endoscopic laryngo-pharyngeal surgery (ELPS) is a treatment for superficial laryngo-pharyngeal cancer which is minimally invasive and excellent for preservation of swallowing or speech function. To maximize its benefits, however, it is necessary to achieve both complete and optimal tumor resection without excessive dissection. This paper outlines the development of a system using an AI image processing model to assist in determining the appropriate resection range during ELPS. As a pilot study, a holdout validation test (172 for training, 44 for validation) was conducted using 216 NBI images of superficial laryngo-pharyngeal cancer treated in our department. An upper gastrointestinal endoscopy specialist checked the images and annotated the lesion areas by surrounding them with free-form curves that were judged to be cancerous based on NBI findings. We used DeepLab v3+, a semantic segmentation model, to predict the lesion area, and performed transfer learning using our data on a published pre-trained model using Pascal VOC 2014. The average performance indices of the inference results verified pixel by pixel for 44 verification images were intersection over union (IoU): 0.596, sensitivity: 0.734, and specificity: 0.947. In Japan, two AI surgical support systems for endoscopic surgeries were approved for manufacture and marketing by the Ministry of Health, Labour and Welfare in 2024 as a “surgical image recognition support program.” In the field of otorhinolaryngology, it is necessary to aim for clinical applications of AI surgical support systems that are truly beneficial to society, with a clear strategy for the future of healthcare systems.

In recent years, the application of artificial intelligence (AI) has advanced rapidly, driven by the evolution of deep learning algorithms, the development of high-performance GPUs, and the growing availability of large-scale datasets. In the medical field, there is increasing interest —both in Japan and worldwide— in applying AI technologies, particularly in the field of endoscopy, to aid in the early detection of gastrointestinal cancers and in distinguishing benign from malignant lesions. Clinical studies have suggested that AI-based diagnostic support may contribute to addressing clinical challenges, such as standardizing diagnostic accuracy and reducing the rate of missed lesions. However, careful validation remains essential before AI technologies can be widely implemented in clinical practice. In this article, we provide an overview of the current status and future prospects of AI development for assisting in the detection of pharyngeal and esophageal cancers using endoscopy. In addition, we present our ongoing research efforts and share the results we have obtained so far in developing AI-assisted diagnostic systems for the pharyngeal and esophageal regions.