Objectives: De-identifying protected health information (PHI) in medical documents is important, and a prerequisite to deidentification is the identification of PHI entity names in clinical documents. This study aimed to compare the performance of three pre-training models that have recently attracted significant attention and to determine which model is more suitable for PHI recognition. Methods: We compared the PHI recognition performance of deep learning models using the i2b2 2014 dataset. We used the three pre-training models—namely, bidirectional encoder representations from transformers (BERT), robustly optimized BERT pre-training approach (RoBERTa), and XLNet (model built based on Transformer-XL)—to detect PHI. After the dataset was tokenized, it was processed using an inside-outside-beginning tagging scheme and WordPiecetokenized to place it into these models. Further, the PHI recognition performance was investigated using BERT, RoBERTa, and XLNet. Results: Comparing the PHI recognition performance of the three models, it was confirmed that XLNet had a superior F1-score of 96.29%. In addition, when checking PHI entity performance evaluation, RoBERTa and XLNet showed a 30% improvement in performance compared to BERT. Conclusions Among the pre-training models used in this study, XLNet exhibited superior performance because word embedding was well constructed using the two-stream self-attention method. In addition, compared to BERT, RoBERTa and XLNet showed superior performance, indicating that they were more effective in grasping the context.

Targeted alpha therapy (TAT) is an active area of drug development as a highly specific and highly potent therapeutic modality that can be applied to many types of late-stage cancers. In order to properly evaluate its safety and efficacy, understanding biokinetics of alpha-emitting radiopharmaceuticals is essential. Quantitative imaging of alpha-emitting radiopharmaceuticals is often possible via imaging of gammas and positrons produced during complex decay chains of these radionuclides. Analysis of the complex decay chains for alpha-emitting radionuclides (Tb-149, At-211, Bi-212 (decayed from Pb-212), Bi-213, Ra-223, Ac- 225, and Th-227) with relevance to imageable signals is attempted in this mini-review article. Gamma camera imaging, single-photon emission computed tomography, positron emission tomography, bremsstrahlung radiation imaging, Cerenkov luminescence imaging, and Compton cameras are briefly discussed as modalities for imaging alpha-emitting radiopharmaceuticals.
Electrons ; Luminescence ; Positron-Emission Tomography ; Radioisotopes ; Radionuclide Imaging ; Radiopharmaceuticals ; Tomography, Emission-Computed ; Tomography, Emission-Computed, Single-Photon

Targeted alpha therapy (TAT) is an active area of drug development as a highly specific and highly potent therapeutic modality that can be applied to many types of late-stage cancers. In order to properly evaluate its safety and efficacy, understanding biokinetics of alpha-emitting radiopharmaceuticals is essential. Quantitative imaging of alpha-emitting radiopharmaceuticals is often possible via imaging of gammas and positrons produced during complex decay chains of these radionuclides. Analysis of the complex decay chains for alpha-emitting radionuclides (Tb-149, At-211, Bi-212 (decayed from Pb-212), Bi-213, Ra-223, Ac- 225, and Th-227) with relevance to imageable signals is attempted in this mini-review article. Gamma camera imaging, single-photon emission computed tomography, positron emission tomography, bremsstrahlung radiation imaging, Cerenkov luminescence imaging, and Compton cameras are briefly discussed as modalities for imaging alpha-emitting radiopharmaceuticals.

Purpose: This study evaluated aggressive hyperbaric oxygen therapy (HBOT) by understanding various exposure routes of acute carbon monoxide (CO) poisoning, the risk factors causing acute cardiovascular, and neurological toxicity caused by poisoning. Methods: A retrospective study was conducted based on the medical records of 417 acute CO poisoning patients who visited the emergency care unit from March 2017 to August 2019. The exposure routes, HBOT performance, age, sex, medical history (hypertension, diabetes mellitus, ischemic heart disease, heart failure), intentionality, loss of consciousness (LOC), intake with alcohol or sedatives, and initial test results (carboxyhemoglobin (COHb), troponin- I, electrocardiography, echocardiography, brain MRI) were examined. Comparative analysis of the clinical information was conducted between the groups that showed acute cardiovascular toxicity and neurological toxicity, and groups that did not. Results: Among 417 patients diagnosed with acute CO poisoning, 201 cases (48.2%) were intentional, and charcoal briquette was the most common route (169 patients (40.5%)). Two hundred sixteen cases (51.8%) were accidental, and fire was the most common route (135 patients (32.4%)). The exposure route was more diverse with accidental poisoning. Three hundred ninety-nine patients were studied for acute cardiovascular toxicity, and 62 patients (15.5%) were confirmed to be positive. The result was statistically significant in intentionality, LOC, combined sedatives, initial COHb, HTN, and IHD. One hundred two patients were studied for acute neurological toxicity, which was observed in 26 patients (25.5%). The result was statistically significant in age and LOC. Conclusion Active HBOT should be performed to minimize damage to the major organs by identifying the various exposure routes of CO poisoning, risk factors for acute cardiovascular toxicity (intentionality, LOC, combined sedatives, initial COHb, HTN, IHD), and the risk factors for acute neurological toxicity (age, LOC).

Objective: Emergency physicians experience difficulty in determining the disposition of patients with elevated troponin I levels using emergency room tests. In this study, we aimed to investigate factors that could discriminate between the occurrence of type 1 myocardial infarction (T1MI) and type 2 myocardial infarction (T2MI) in patients with elevated troponin I levels. Methods: Patients admitted to the emergency department between January 1, 2017 and June 30, 2017 with elevated troponin I levels who underwent subsequent cardiac biomarker testing were included. Samples for baseline blood tests, such as cardiac biomarker levels, were collected within approximately 10 minutes of admission. Electrocardiogram, transthoracic echocardiography, and percutaneous coronary intervention results were retrospectively examined via patient report and chart reviews. Results: During the study period, 169 of 234 (72%) patients were diagnosed with T2MI and 65 (28%) were diagnosed with T1MI. Among various factors, typical chest pain (odds ratio [OR], 4.40; 95% confidence interval [CI], 1.46 to 13.24; P=0.008), high troponin I levels (OR, 1.50; 95% CI, 1.19 to 1.90; P<0.001), high cholesterol (OR, 1.01; 95% CI, 1.00 to 1.02; P=0.008), and low D-dimer levels (OR, 0.87; 95% CI, 0.77 to 0.98; P=0.027) were significantly associated with T1MI incidence. Conclusion Our findings in this study indicate that typical chest pain, high levels of troponin I and cholesterol, and low levels of D-dimer were associated with the diagnosis of T1MI. Further studies are suggested to determine the cut-off values for accurate diagnosis of T1MI in the ED.