Automatic Classification of the Korean Triage Acuity Scale in Simulated Emergency Rooms Using Speech Recognition and Natural Language Processing: a Proof of Concept Study

Background@#Rapid triage reduces the patients' stay time at an emergency department (ED). The Korean Triage Acuity Scale (KTAS) is mandatorily applied at EDs in South Korea.For rapid triage, we studied machine learning-based triage systems composed of a speech recognition model and natural language processing-based classification. @*Methods@#We simulated 762 triage cases that consisted of 18 classes with six types of the main symptom (chest pain, dyspnea, fever, stroke, abdominal pain, and headache) and three levels of KTAS. In addition, we recorded conversations between emergency patients and clinicians during the simulation. We used speech recognition models to transcribe the conversation. Bidirectional Encoder Representation from Transformers (BERT), support vector machine (SVM), random forest (RF), and k-nearest neighbors (KNN) were used for KTAS and symptom classification. Additionally, we evaluated the Shapley Additive exPlanations (SHAP) values of features to interpret the classifiers. @*Results@#The character error rate of the speech recognition model was reduced to 25.21% through transfer learning. With auto-transcribed scripts, support vector machine (area under the receiver operating characteristic curve [AUROC], 0.86; 95% confidence interval [CI], 0.81–0.9), KNN (AUROC, 0.89; 95% CI, 0.85–0.93), RF (AUROC, 0.86; 95% CI, 0.82–0.9) and BERT (AUROC, 0.82; 95% CI, 0.75–0.87) achieved excellent classification performance.Based on SHAP, we found “stress”, “pain score point”, “fever”, “breath”, “head” and “chest” were the important vocabularies for determining KTAS and symptoms. @*Conclusion@#We demonstrated the potential of an automatic KTAS classification system using speech recognition models, machine learning and BERT-based classifiers.

Automatic Classification of the Korean Triage Acuity Scale in Simulated Emergency Rooms Using Speech Recognition and Natural Language Processing: a Proof of Concept Study

Background@#Rapid triage reduces the patients' stay time at an emergency department (ED). The Korean Triage Acuity Scale (KTAS) is mandatorily applied at EDs in South Korea.For rapid triage, we studied machine learning-based triage systems composed of a speech recognition model and natural language processing-based classification. @*Methods@#We simulated 762 triage cases that consisted of 18 classes with six types of the main symptom (chest pain, dyspnea, fever, stroke, abdominal pain, and headache) and three levels of KTAS. In addition, we recorded conversations between emergency patients and clinicians during the simulation. We used speech recognition models to transcribe the conversation. Bidirectional Encoder Representation from Transformers (BERT), support vector machine (SVM), random forest (RF), and k-nearest neighbors (KNN) were used for KTAS and symptom classification. Additionally, we evaluated the Shapley Additive exPlanations (SHAP) values of features to interpret the classifiers. @*Results@#The character error rate of the speech recognition model was reduced to 25.21% through transfer learning. With auto-transcribed scripts, support vector machine (area under the receiver operating characteristic curve [AUROC], 0.86; 95% confidence interval [CI], 0.81–0.9), KNN (AUROC, 0.89; 95% CI, 0.85–0.93), RF (AUROC, 0.86; 95% CI, 0.82–0.9) and BERT (AUROC, 0.82; 95% CI, 0.75–0.87) achieved excellent classification performance.Based on SHAP, we found “stress”, “pain score point”, “fever”, “breath”, “head” and “chest” were the important vocabularies for determining KTAS and symptoms. @*Conclusion@#We demonstrated the potential of an automatic KTAS classification system using speech recognition models, machine learning and BERT-based classifiers.

Trend of Prevalence and Antifungal Drug Resistance of Candida Species Isolated from Candidemia Patients at a Tertiary Care Hospital During Recent Two Decades / 대한임상미생물학회지

BACKGROUND: Candidemia has increased with an increasing number of people in the high risk group and so has become more important. This study was conducted to investigate the isolation rate of Candida species from candidemia patients and the change in rate of antifungal resistance. METHODS: At a single tertiary care hospital, 1,120 blood cultures positive for Candida species from 1997 to 2016 were investigated according to date of culture, gender, age, and hospital department. RESULTS: During the investigation period, the number of candidemia patients increased from 14 in 1997 to 84 in 2016. The most common organism identified during the two decades was Candida albicans (40.8%), followed by Candida parapsilosis (24.1%), Candida tropicalis (13.2%), and Candida glabrata (12.8%). C. glabrata was relatively common in females (45.5%) compared to males. The age group 40-89 years was more frequently infected than other age groups, and the most frequent isolates according to age group were C. albicans in neonate (66.7%), C. parapsilosis in 1-9-year-olds (41.7%), and C. glabrata in those aged ≥60 years (range; 13.3%–20.0%). According to the visited departments, C. albicans, C. glabrata, and Candida haemulonii were more common in medical departments, while C. parapsilosis was more common in surgical departments. In the antifungal susceptibility test, a rising trend of azole resistance among C. albicans and C. glabrata was observed in recent years. CONCLUSION: In this study, it was confirmed that the isolation rate of Candida species in blood is different by age, gender, and hospital department, and the distribution of isolated Candida species changed over time. The resistance patterns of antifungal agents are also changing, and continuous monitoring and proper selection of antifungal agents are necessary.

TiO2 nanotube stimulate chondrogenic differentiation of limb mesenchymal cells by modulating focal activity

Vertically aligned, laterally spaced nanoscale titanium nanotubes were grown on a titanium surface by anodization, and the growth of chondroprogenitors on the resulting surfaces was investigated. Surfaces bearing nanotubes of 70 to 100 nm in diameter were found to trigger the morphological transition to a cortical actin pattern and rounded cell shape (both indicative of chondrocytic differentiation), as well as the up-regulation of type II collagen and integrin beta4 protein expression through the down-regulation of Erk activity. Inhibition of Erk signaling reduced stress fiber formation and induced the transition to the cortical actin pattern in cells cultured on 30-nm-diameter nanotubes, which maintained their fibroblastoid morphologies in the absence of Erk inhibition. Collectively, these results indicate that a titanium-based nanotube surface can support chondrocytic functions among chondroprogenitors, and may therefore be useful for future cartilaginous applications.

Curcumin inhibits cellular condensation and alters microfilament organization during chondrogenic differentiation of limb bud mesenchymal cells

Curcumin is a well known natural polyphenol product isolated from the rhizome of the plant Curcuma longa, anti-inflammatory agent for arthritis by inhibiting synthesis of inflammatory prostaglandins. However, the mechanisms by which curcumin regulates the functions of chondroprogenitor, such as proliferation, precartilage condensation, cytoskeletal organization or overall chondrogenic behavior, are largely unknown. In the present report, we investigated the effects and signaling mechanism of curcumin on the regulation of chondrogenesis. Treating chick limb bud mesenchymal cells with curcumin suppressed chondrogenesis by stimulating apoptotic cell death. It also inhibited reorganization of the actin cytoskeleton into a cortical pattern concomitant with rounding of chondrogenic competent cells and down-regulation of integrin beta1 and focal adhesion kinase (FAK) phosphorylation. Curcumin suppressed the phosphorylation of Akt leading to Akt inactivation. Activation of Akt by introducing a myristoylated, constitutively active form of Akt reversed the inhibitory actions of curcumin during chondrogenesis. In summary, for the first time, we describe biological properties of curcumin during chondrogenic differentiation of chick limb bud mesenchymal cells. Curcumin suppressed chondrogenesis by stimulating apoptotic cell death and down-regulating integrin-mediated reorganization of actin cytoskeleton via modulation of Akt signaling.