Mucosal melanoma of the head and neck: a retrospective analysis of 34 cases in Japan.

PURPOSE: Mucosal melanoma of the head and neck (MMHN) is a rare condition. This study aimed to investigate oncological outcomes of surgical intervention in patients with MMHN. MATERIALS AND METHODS: The study included 34 patients with MMHN who underwent surgical resection as initial treatment at 10 institutions in Japan between July 2005 and June 2015. Results: The 5-year overall survival (OS), local control rate (LCR), disease-free survival (DFS), and disease-specific survival (DSS) rates were 48.7%, 53.4%, 32.4%, and 55.1%, respectively. Based on multivariate analysis, no independent prognostic factors for the 5-year OS and DSS were found. Based on univariate analysis, the 5-year LCR was worse in patients with lesions in the nasal cavity and paranasal sinuses than in the oral cavity and pharynx. However, no differences in oncological outcomes were identified in relation to primary sites, and postoperative radiotherapy (PORT) and adjuvant systemic therapy did not contribute to improvements in the 5-year OS. CONCLUSIONS: No independent prognostic factors for the 5-year OS or DSS were identified. Regional or distant recurrences are often identified, regardless of local control with surgical resection. Difficult control of MMHN with conventional therapeutic strategies, such as surgical intervention, PORT, and systemic therapy, has been suggested.

The Effect of Noise on Deep Learning for Classification of Pathological Voice.

OBJECTIVE: This study aimed to evaluate the significance of background noise in machine learning models assessing the GRBAS scale for voice disorders. METHODS: A dataset of 1406 voice samples was collected from retrospective data, and a 5-layer 1D convolutional neural network (CNN) model was constructed using TensorFlow. The dataset was divided into training, validation, and test data. Gaussian noise was added to test samples at various intensities to assess the model's noise resilience. The model's performance was evaluated using accuracy, F1 score, and quadratic weighted Cohen's kappa score. RESULTS: The model's performance on the GRBAS scale generally declined with increasing noise intensities. For the G scale, accuracy dropped from 70.9% (original) to 8.5% (at the highest noise), F1 score from 69.2% to 1.3%, and Cohen's kappa from 0.679 to 0.0. Similar declines were observed for the remaining RBAS components. CONCLUSION: The model's performance was affected by background noise, with substantial decreases in evaluation metrics as noise levels intensified. Future research should explore noise-tolerant techniques, such as data augmentation, to improve the model's noise resilience in real-world settings. LEVEL OF EVIDENCE: This study evaluates a machine learning model using a single dataset without comparative controls. Given its non-comparative design and specific focus, it aligns with Level 4 evidence (Case-series) under the 2011 OCEBM guidelines Laryngoscope, 2024.

Unique environmental Symbiodiniaceae diversity at an isolated island in the northwestern Pacific.

Dinoflagellates in the family Symbiodiniaceae are intensively investigated as algal symbionts of corals and other invertebrates, underpinning coral reef ecosystems as primary producers. Diversity, including regional diversification, of free-living communities is less studied. In this study, an environmental Symbiodiniaceae community at an isolated island, Okinotori Island, Japan, was investigated to determine whether the community is endemic or common with other locations near continents and major ocean currents. Symbiotic algae in common corals at the island were the same type as those of the corals from other Japanese waters. In the environmental samples, genera Symbiodinium (formerly clade A), Cladocopium (clade C), Durusdinium (clade D), and clades F (including Freudenthalidium), G, and I, were identified through analysis of internal transcribed spacer region 2 of nuclear ribosomal RNA gene (ITS2) sequences. Interestingly, some sequences found were genetically different from those of previously reported genera/clades. These unknown sequences were genetically included in the Symbiodiniaceae linage, but they were differentiated from the previously known nine clades. The sequences formed a cluster in the phylogenetic tree based on 28S nrDNA. These sequences were thus considered members of a novel clade in the family (clade J). In total, 120 kinds of ITS2 sequences were produced; while 10 were identical to previously reported sequences, the majority were highly divergent. These genetically unique Symbiodiniaceae types, including novel clade J, may have evolved in isolation and reflect the environmental characteristics of the Okinotori Island.

Real-time analysis application for identifying bursty local areas related to emergency topics.

Since social media started getting more attention from users on the Internet, social media has been one of the most important information source in the world. Especially, with the increasing popularity of social media, data posted on social media sites are rapidly becoming collective intelligence, which is a term used to refer to new media that is displacing traditional media. In this paper, we focus on geotagged tweets on the Twitter site. These geotagged tweets are referred to as georeferenced documents because they include not only a short text message, but also the documents' posting time and location. Many researchers have been tackling the development of new data mining techniques for georeferenced documents to identify and analyze emergency topics, such as natural disasters, weather, diseases, and other incidents. In particular, the utilization of geotagged tweets to identify and analyze natural disasters has received much attention from administrative agencies recently because some case studies have achieved compelling results. In this paper, we propose a novel real-time analysis application for identifying bursty local areas related to emergency topics. The aim of our new application is to provide new platforms that can identify and analyze the localities of emergency topics. The proposed application is composed of three core computational intelligence techniques: the Naive Bayes classifier technique, the spatiotemporal clustering technique, and the burst detection technique. Moreover, we have implemented two types of application interface: a Web application interface and an android application interface. To evaluate the proposed application, we have implemented a real-time weather observation system embedded the proposed application. we used actual crawling geotagged tweets posted on the Twitter site. The weather observation system successfully detected bursty local areas related to observed emergency weather topics.