Intracerebral hemorrhage is a severe problem where more than one-third of patients die within a month. In diagnosing intracranial hemorrhage, neuroimaging examinations are essential. As a result, the interpretation of neuroimaging becomes a crucial process in medical procedures. However, human-based image interpretation has inherent limitations, as it can only handle a restricted range of tasks. To address this, a study on medical image captioning has been conducted, but it primarily focused on single medical images. However, actual medical images often consist of continuous sequences, such as CT scans, making it challenging to directly apply existing studies. Therefore, this paper proposes a CT image captioning model that utilizes a 3D-CNN model and distilGPT-2. In this study, four combinations of 3D-CNN models and language models were compared and analyzed for their performance. Additionally, the impact of applying penalties to the loss function and adjusting penalty values during the training process was examined. The proposed CT image captioning model demonstrated a maximum BLEU score of 0.35 on the in-house dataset, and it was observed that the text generated by the model became more similar to human interpretations in medical image reports with the application of loss function penalties.

Intracerebral hemorrhage (ICH) is a severe cerebrovascular disorder that poses a life-threatening risk, necessitating swift diagnosis and treatment. While CT scans are the most effective diagnostic tool for detecting cerebral hemorrhage, their interpretation typically requires the expertise of skilled professionals. However, in regions with a shortage of such experts or situations with time constraints, delays in diagnosis may occur. In this paper, we propose a method that combines a pre-trained CNN classifier and GPT-2 to generate text for sequentially acquired ICH CT images. Initially, CNN undergoes fine-tuning by learning the presence of ICH in publicly available single CT images, and subsequently, it extracts feature vectors (i.e., matrix) from 3D ICH CT images. These vectors are input along with text into GPT-2, which is trained to generate text for consecutive CT images. In experiments, we evaluated the performance of four models to determine the most suitable image captioning model: (1) In the N-gram-based method, ReseNet50V2 and DenseNet121 showed relatively high scores. (2) In the embedding-based method, DenseNet121 exhibited the best performance. (3) Overall, the models showed good performance in BERT score. Our proposed method presents an automatic and valuable approach for analyzing 3D ICH CT images, contributing to the efficiency of ICH diagnosis and treatment.

Purpose: This study develops a chatbot for school violence prevention (C-SVP) among elementary school students.Methods: Among the analysis, design, development, implementation, and evaluation (ADDIE) models, ADD phases were applied to develop a C-SVP. Students’ learning needs were identified by constructing content with a design that attracted their attention. Subsequently, a formative evaluation was conducted on the developed C-SVP to test its applicability by ten elementary school students targeting the 5th and 6th grades.Results: The chatbot was designed using KakaoTalk and named “School Guardian Angel.” The formative evaluation revealed that the developed C-SVP was easily accessible and useful for elementary school students.Conclusion: The developed C-SVP is expected to be effective in preventing violence among elementary school students. However, further research involving children of various age groups is required.

Objectives: Although language sample analysis is essential for language evaluation, it is not widely conducted due to time constraints and difficulties in analysis. This study was conducted to evaluate the Korean-Automatic Language System (K-ALAS), which was developed to support language sample analysis.Methods: 50 language samples of young children aged 2 to 6 were analyzed by K-ALAS and a language analysis expert. The percentage agreement of the analysis results between K-ALAS and the expert on Eojeol, words, and morpheme, which are the basic units of language measurement, were calculated, and t-test and Pearson correlation analysis were conducted on the main language measures (MLUe, MLUw, MLUm, NTJ, NTE, NDJ, NDE, NTW, NDW, TTR) measured by K-ALAS and the expert.Results: The percentage agreement of the eogeol, word, and morpheme analysis between K-ALAS and the expert, was found to be very high at 97.87-99.88%, and the t-test results showed that there was no significant difference between all language measures measured by K-ALAS and the expert, and the result of correlation analysis was also very high with r = .972-1.000.Conclusion: This study confirmed that K-ALAS analyze language samples and measures the main language measures very accurately. We discussed the implications of clinical use of K-ALAS in relation to language diagnosis and intervention.