The opacity of deep learning makes its application challenging in the medical field. Therefore, there is a need to enable explainable artificial intelligence (XAI) in the medical field to ensure that models and their results can be explained in a manner that humans can understand. This study uses a high-accuracy computer vision algorithm model to transfer learning to medical text tasks and uses the explanatory visualization method known as gradient-weighted class activation mapping (Grad-CAM) to generate heat maps to ensure that the basis for decision-making can be provided intuitively or via the model. The system comprises four modules: pre-processing, word embedding, classifier, and visualization. We used Word2Vec and BERT to compare word embeddings and use ResNet and 1Dimension convolutional neural networks (CNN) to compare classifiers. Finally, the Bi-LSTM was used to perform text classification for direct comparison. With 25 epochs, the model that used pre-trained ResNet on the formalized text presented the best performance (recall of 90.9%, precision of 91.1%, and an F1 score of 90.2% weighted). This study uses ResNet to process medical texts through Grad-CAM-based explainable artificial intelligence and obtains a high-accuracy classification effect; at the same time, through Grad-CAM visualization, it intuitively shows the words to which the model pays attention when making predictions.

Background: Apps for real-time continuous glucose monitoring (CGM) on smartphones and other devices linked to CGM systems have recently been developed, and such CGM apps are also coming into use in Japan. In comparison with conventional retrospective CGM, the use of CGM apps improves patients’ own blood glucose control, which is expected to help slow the progression of type 2 diabetes mellitus (DM) and prevent complications, but the effect of their introduction on medical costs remains unknown. Objective: Our objective in this study was to perform an economic appraisal of CGM apps from the viewpoint of assessing public medical costs associated with type 2 DM, using the probability of developing type 2 DM–associated complications, and data on medical costs and utility value to carry out a medical cost simulation using a Markov model in order to ascertain the cost-effectiveness of the apps. Methods: We developed a Markov model with the transition states of insulin therapy, nephrosis, dialysis, and cardiovascular disease, all of which have a major effect on medical costs, to identify changes in medical costs and utility values resulting from the introduction of a CGM app and calculated the incremental cost-effectiveness ratio (ICER). Results: The ICER for CGM app use was US $33,039/quality-adjusted life year (QALY). Conclusions: Sensitivity analyses showed that, with the exception of conditions where the transition probability of insulin therapy, utility value, or increased medical costs increases, the ICER for the introduction of CGM apps was below the threshold of US $43,478/QALY used by the Central Social Insurance Medical Council. Our results provide basic data on the cost-effectiveness of introducing CGM apps, which are currently starting to come into use.

Background: Named entity recognition (NER) plays an important role in extracting the features of descriptions such as the name and location of a disease for mining free-text radiology reports. However, the performance of existing NER tools is limited because the number of entities that can be extracted depends on the dictionary lookup. In particular, the recognition of compound terms is very complicated because of the variety of patterns. Objective: The aim of this study is to develop and evaluate an NER tool concerned with compound terms using RadLex for mining free-text radiology reports. Methods: We leveraged the clinical Text Analysis and Knowledge Extraction System (cTAKES) to develop customized pipelines using both RadLex and SentiWordNet (a general purpose dictionary). We manually annotated 400 radiology reports for compound terms in noun phrases and used them as the gold standard for performance evaluation (precision, recall, and F-measure). In addition, we created a compound terms–enhanced dictionary (CtED) by analyzing false negatives and false positives and applied it to another 100 radiology reports for validation. We also evaluated the stem terms of compound terms by defining two measures: occurrence ratio (OR) and matching ratio (MR). Results: The F-measure of cTAKES+RadLex+general purpose dictionary was 30.9% (precision 73.3% and recall 19.6%) and that of the combined CtED was 63.1% (precision 82.8% and recall 51%). The OR indicated that the stem terms of effusion, node, tube, and disease were used frequently, but it still lacks capturing compound terms. The MR showed that 71.85% (9411/13,098) of the stem terms matched with that of the ontologies, and RadLex improved approximately 22% of the MR from the cTAKES default dictionary. The OR and MR revealed that the characteristics of stem terms would have the potential to help generate synonymous phrases using the ontologies. Conclusions: We developed a RadLex-based customized pipeline for parsing radiology reports and demonstrated that CtED and stem term analysis has the potential to improve dictionary-based NER performance with regard to expanding vocabularies.

Background: The COVID-19 pandemic prompted global behavioral restrictions, impacting public mental health. Sentiment analysis, a tool for assessing individual and public emotions from text data, gained importance amid the pandemic. This study focuses on Japan’s early public health interventions during COVID-19, utilizing sentiment analysis in infodemiology to gauge public sentiment on social media regarding these interventions. Objective: This study aims to investigate shifts in public emotions and sentiments before and after the first state of emergency was declared in Japan. By analyzing both user-generated tweets and retweets, we aim to discern patterns in emotional responses during this critical period. Methods: We conducted a day-by-day analysis of Twitter (now known as X) data using 4,894,009 tweets containing the keywords “corona,” “COVID-19,” and “new pneumonia” from March 23 to April 21, 2020, approximately 2 weeks before and after the first declaration of a state of emergency in Japan. We also processed tweet data into vectors for each word, employing the Fuzzy-C-Means (FCM) method, a type of cluster analysis, for the words in the sentiment dictionary. We set up 7 sentiment clusters (negative: anger, sadness, surprise, disgust; neutral: anxiety; positive: trust and joy) and conducted sentiment analysis of the tweet groups and retweet groups. Results: The analysis revealed a mix of positive and negative sentiments, with “joy” significantly increasing in the retweet group after the state of emergency declaration. Negative emotions, such as “worry” and “disgust,” were prevalent in both tweet and retweet groups. Furthermore, the retweet group had a tendency to share more negative content compared to the tweet group. Conclusions: This study conducted sentiment analysis of Japanese tweets and retweets to explore public sentiments during the early stages of COVID-19 in Japan, spanning 2 weeks before and after the first state of emergency declaration. The analysis revealed a mix of positive (joy) and negative (anxiety, disgust) emotions. Notably, joy increased in the retweet group after the emergency declaration, but this group also tended to share more negative content than the tweet group. This study suggests that the state of emergency heightened positive sentiments due to expectations for infection prevention measures, yet negative information also gained traction. The findings propose the potential for further exploration through network analysis.

[This corrects the article DOI: 10.2196/37881.].

Aiming at the problems of poor retrieval intelligence and low recall rate of irrigation fertilization strategy in apple orchard, an intelligent retrieval algorithm of apple orchard irrigation fertilization strategy is proposed to realize the intelligent retrieval of irrigation fertilization strategy. Firstly, the retrieval of the user's irrigation fertilization strategy is divided into title retrieval and comprehensive retrieval, and the comprehensive retrieval is a compound search with the name of the heirloom, the upload time range and the evaluation value. Secondly, the semantic similarity retrieval algorithm based on ontology is proposed for the title retrieval of apple orchard irrigation fertilization strategy, and the domain ontology of apple orchard is constructed. Lastly, the semantic similarity between the retrieval statement and the title is calculated by using the concept similarity algorithm based on ontology and the sentence similarity algorithm, and the title of irrigation fertilization strategy is sorted according to the similarity degree. This paper compares the algorithm with the traditional keyword retrieval, and the experimental results show that the retrieval algorithm proposed in this paper has higher recall rate of irrigation fertilization strategy than the traditional keyword retrieval recall rate, and can realize the intelligent retrieval of the irrigation fertilization strategy of apple orchard.