Journal of the Chinese Medical Association : JCMA
-
Recently, the rapid advancement in generative artificial intelligence (AI) technology, such as ChatGPT-4, has sparked discussions, particularly in image recognition. Accurate results are critical for hematological diagnosis, particularly for blood morphology identification. Despite advanced hematology analyzers, reliance on professional hematopathologists for manual identification remains in cases of abnormal or rare conditions, a process prone to human subjectivity and potential errors. Consequently, this study aimed to investigate the potential of ChatGPT-4 to assist with blood morphology identification. ⋯ This study shows that although generative AI shows the potential for blood type identification, it has not yet reached the point where it can replace the professional judgment of medical staff. The results showed that ChatGPT-4 is excellent for identifying red blood cell morphology, particularly inclusion bodies. It can be used as an auxiliary tool for clinical diagnosis. However, the overall recognition accuracy must be further improved. Our study produced innovative results in this field, establishing a foundation for future studies and highlighting the potential of generative AI in aiding blood morphology recognition. Future research should focus on enhancing the effectiveness of AI to improve overall standards of medical care.
-
Patients who survive an acute myocardial infarction (AMI) have a higher risk of having a major cardiovascular event (MACE). Cardiopulmonary exercise testing (CPET) could develop prognostic stratification and prescribing exercise prescription. Patients after AMI frequently terminate CPET early with submaximal testing results. We aimed to look at the characteristics of patients' predischarge CPET variables after AMI intervention and find potential CPET variables with prognostic value. ⋯ It is critical to identify a submaximal predictor during CPET for patients who survive AMI. Our findings suggested that OUES could be a significant prognostic indicator in patients after first AMI in both the short and long term.
-
Intensive care unit (ICU) mortality prediction helps to guide therapeutic decision making for critically ill patients. Several scoring systems based on statistical techniques have been developed for this purpose. In this study, we developed a machine-learning model to predict patient mortality in the very early stage of ICU admission. ⋯ The XGBoost model most accurately predicted ICU mortality and was superior to traditional scoring systems. Our results highlight the utility of machine learning for ICU mortality prediction in the Asian population.
-
Duodenal adenocarcinoma is rare and its prognostic factors remain controversial. In our study, the role of cell-free deoxyribonucleic acid (cfDNA) as prognostic factor in duodenal adenocarcinoma was evaluated. ⋯ cfDNA analysis is simple and noninvasive. High cfDNA level is a strong independent prognostic factor for decreased overall survival and it should be integrated into clinical care.