Use of Automated Machine Learning for Classifying Hemoperitoneum on Ultrasonographic Images of Morrison's Pouch: A Multicenter Retrospective Study.

This study evaluated automated machine learning (AutoML) in classifying the presence or absence of hemoperitoneum in ultrasonography (USG) images of Morrison's pouch. In this multicenter, retrospective study, 864 trauma patients from trauma and emergency medical centers in South Korea were included. In all, 2200 USG images (1100 hemoperitoneum and 1100 normal) were collected. Of these, 1800 images were used for training and 200 were used for the internal validation of AutoML. External validation was performed using 100 hemoperitoneum images and 100 normal images collected separately from a trauma center that were not included in the training and internal validation sets. Google's open-source AutoML was used to train the algorithm in classifying hemoperitoneum in USG images, followed by internal and external validation. In the internal validation, the sensitivity, specificity, and area under the receiver operating characteristic (AUROC) curve were 95%, 99%, and 0.97, respectively. In the external validation, the sensitivity, specificity, and AUROC were 94%, 99%, and 0.97, respectively. The performances of AutoML in the internal and external validation were not statistically different (p = 0.78). A publicly available, general-purpose AutoML can accurately classify the presence or absence of hemoperitoneum in USG images of the Morrison's pouch of real-world trauma patients.

Clinical usefulness of deep learning-based automated segmentation in intracranial hemorrhage.

BACKGROUND: Doctors with various specializations and experience order brain computed tomography (CT) to rule out intracranial hemorrhage (ICH). Advanced artificial intelligence (AI) can discriminate subtypes of ICH with high accuracy. OBJECTIVE: The purpose of this study was to investigate the clinical usefulness of AI in ICH detection for doctors across a variety of specialties and backgrounds. METHODS: A total of 5702 patients' brain CTs were used to develop a cascaded deep-learning-based automated segmentation algorithm (CDLA). A total of 38 doctors were recruited for testing and categorized into nine groups. Diagnostic time and accuracy were evaluated for doctors with and without assistance from the CDLA. RESULTS: The CDLA in the validation set for differential diagnoses among a negative finding and five subtypes of ICH revealed an AUC of 0.966 (95% CI, 0.955-0.977). Specific doctor groups, such as interns, internal medicine, pediatrics, and emergency junior residents, showed significant improvement with assistance from the CDLA (p= 0.029). However, the CDLA did not show a reduction in the mean diagnostic time. CONCLUSIONS: Even though the CDLA may not reduce diagnostic time for ICH detection, unlike our expectation, it can play a role in improving diagnostic accuracy in specific doctor groups.

External validation of multimodal termination of resuscitation rules for out-of-hospital cardiac arrest patients in the COVID-19 era.

BACKGROUND: Futile resuscitation for out-of-hospital cardiac arrest (OHCA) patients in the coronavirus disease (COVID)-19 era can lead to risk of disease transmission and unnecessary transport. Various existing basic or advanced life support (BLS or ALS, respectively) rules for the termination of resuscitation (TOR) have been derived and validated in North America and Asian countries. This study aimed to evaluate the external validation of these rules in predicting the survival outcomes of OHCA patients in the COVID-19 era. METHODS: This was a multicenter observational study using the WinCOVID-19 Daegu registry data collected during February 18-March 31, 2020. The subjects were patients who showed cardiac arrest of presumed cardiac etiology. The outcomes of each rule were compared to the actual patient survival outcomes. The sensitivity, specificity, false positive value (FPV), and positive predictive value (PPV) of each TOR rule were evaluated. RESULTS: In total, 170 of the 184 OHCA patients were eligible and evaluated. TOR was recommended for 122 patients based on the international basic life support termination of resuscitation (BLS-TOR) rule, which showed 85% specificity, 74% sensitivity, 0.8% FPV, and 99% PPV for predicting unfavorable survival outcomes. When the traditional BLS-TOR rules and KoCARC TOR rule II were applied to our registry, one patient met the TOR criteria but survived at hospital discharge. With regard to the FPV (upper limit of 95% confidence interval < 5%), specificity (100%), and PPV (> 99%) criteria, only the KoCARC TOR rule I, which included a combination of three factors including not being witnessed by emergency medical technicians, presenting with an asystole at the scene, and not experiencing prehospital shock delivery or return of spontaneous circulation, was found to be superior to all other TOR rules. CONCLUSION: Among the previous nine BLS and ALS TOR rules, KoCARC TOR rule I was most suitable for predicting poor survival outcomes and showed improved diagnostic performance. Further research on variations in resources and treatment protocols among facilities, regions, and cultures will be useful in determining the feasibility of TOR rules for COVID-19 patients worldwide.

Obstruction of the Hepatic Venous Flow Caused by Intravenous Leiomyomatosis.

Budd-Chiari syndrome (BCS) is a rare intrahepatic vascular disease that is characterized by a hepatic venous outflow obstruction. Intravenous leiomyomatosis (ILs) is a rare complication of a myoma. Here, we report a case of BCS that was caused by intracaval ILs. A woman presented to the emergency department (ED) with abdominal distension that had gradually progressed over a period of 3 years. Bedside ultrasonography and contrast-enhanced computed tomography (CECT) showed a large ascites and pelvic mass. The mass continued to the inferior vena cava and the right atrium. The intracaval mass was obstructing the left and middle hepatic veins. We established a tentative diagnosis of BCS caused by intracaval ILs and attempted surgical resection. Complete resection of the intracaval mass failed because of adhesion; however, she was discharged from the hospital without any postoperative complications. After 3 months, a pelvic ultrasonography showed a recurrence of a 4 × 3 cm pelvic mass. The mass size increased to 6 cm after 30 months. ILs can cause secondary BCS and can lead to life-threatening conditions. Owing to its extreme rarity, early detection in the ED is challenging. Bedside ultrasonography and CECT can enable the early recognition of BCS by ILs.

Deep-learning algorithms for the interpretation of chest radiographs to aid in the triage of COVID-19 patients: A multicenter retrospective study.

The recent medical applications of deep-learning (DL) algorithms have demonstrated their clinical efficacy in improving speed and accuracy of image interpretation. If the DL algorithm achieves a performance equivalent to that achieved by physicians in chest radiography (CR) diagnoses with Coronavirus disease 2019 (COVID-19) pneumonia, the automatic interpretation of the CR with DL algorithms can significantly reduce the burden on clinicians and radiologists in sudden surges of suspected COVID-19 patients. The aim of this study was to evaluate the efficacy of the DL algorithm for detecting COVID-19 pneumonia on CR compared with formal radiology reports. This is a retrospective study of adult patients that were diagnosed as positive COVID-19 cases based on the reverse transcription polymerase chain reaction among all the patients who were admitted to five emergency departments and one community treatment center in Korea from February 18, 2020 to May 1, 2020. The CR images were evaluated with a publicly available DL algorithm. For reference, CR images without chest computed tomography (CT) scans classified as positive for COVID-19 pneumonia were used given that the radiologist identified ground-glass opacity, consolidation, or other infiltration in retrospectively reviewed CR images. Patients with evidence of pneumonia on chest CT scans were also classified as COVID-19 pneumonia positive outcomes. The overall sensitivity and specificity of the DL algorithm for detecting COVID-19 pneumonia on CR were 95.6%, and 88.7%, respectively. The area under the curve value of the DL algorithm for the detection of COVID-19 with pneumonia was 0.921. The DL algorithm demonstrated a satisfactory diagnostic performance comparable with that of formal radiology reports in the CR-based diagnosis of pneumonia in COVID-19 patients. The DL algorithm may offer fast and reliable examinations that can facilitate patient screening and isolation decisions, which can reduce the medical staff workload during COVID-19 pandemic situations.

Transtracheal ultrasonographic confirmation of endotracheal intubation using I-gel and an endotracheal tube introducer.

BACKGROUND: This study aimed to investigate the efficacy of transtracheal ultrasonography in confirming the placement of an endotracheal tube introducer during endotracheal intubation using the I-gel supraglottic airway as a guide. METHODS: In this prospective study, endotracheal intubation using an endotracheal tube introducer through the I-gel was performed in patients with return of spontaneous circulation after I-gel insertion for out-of-hospital cardiac arrest. The introducer placement was assessed by the occurrence of hyperechoic artifacts within the trachea or esophagus in transtracheal ultrasonography. Results of ultrasonography were confirmed by waveform capnography in the case of tracheal artifacts and direct visualization by laryngoscopy in the case of esophageal artifacts. RESULTS: One hundred and six patients were enrolled in this study. In transtracheal ultrasonography, artifacts of introducer in the trachea and esophagus were observed in 80 (75.5%) and 26 (24.5%) patients, respectively. Transtracheal ultrasonography in identifying the placement of introducers revealed a sensitivity of 100% (95% CI 95.4 to 100), specificity of 96.3% (95% CI 81.0 to 99.9), a PPV of 98.7% (95% CI 92.0 to 99.8), and a NPV of 100% (95% CI 94.8 to 99.9). CONCLUSION: Transtracheal ultrasonography is an accurate method for identifying introducer placement during endotracheal intubation using an endotracheal tube introducer through the I-gel.

CT window trainable neural network for improving intracranial hemorrhage detection by combining multiple settings.

Window settings to rescale and contrast stretch raw data from radiographic images such as Computed Tomography (CT), X-ray and Magnetic Resonance images is a crucial step as data pre-processing to examine abnormalities and diagnose diseases. We propose a distant-supervised method for determining automatically the best window settings by attaching a window estimator module (WEM) to a deep convolutional neural network (DCNN)-based lesion classifier and training them in conjunction. Aside from predicting a flexible window setting for each raw image, we statistically identify the top four window settings by calculating the mean and standard deviations for the entire dataset. Images are scaled on each of the top settings estimated by WEM and following lesion classifiers are subsequently trained. We study the effects of only using the flexible window, the single fixed window as either a known default window used by radiologists or an estimated mean value, and two different approaches to combine results from the top window settings to improve the detection of intracranial hemorrhage (ICH) from brain CT images. Experimental results showed that using the top four window settings identified from the window estimator module and combining the results had the best performance.

Preparedness of personal protective equipment and implementation of new CPR strategies for patients with out-of-hospital cardiac arrest in the COVID-19 era.

BACKGROUND: In February and March 2020, healthcare providers and citizens in Daegu, South Korea, experienced the onslaught of a large-scale community epidemic of COVID-19. This had a profound impact on patients who experienced out-of-hospital cardiac arrest (OHCA). METHODS: We conducted a retrospective observational study of 171 OHCA patients based on the multicenter WinCOVID registry. Demographic and clinical characteristics, overall survival, COVID-19 related data, as well as personal protective equipment (PPE) and resuscitation techniques used during the COVID-19 outbreak were evaluated and compared with outcomes from a 2018 historical cohort (n â€‹= â€‹158). RESULTS: Among the interventions, high-level PPE was introduced and standard cardiopulmonary resuscitation was changed to chest compressions using mechanical devices. All OHCA patients were treated as confirmed or suspicious for COVID-19 regardless of symptoms. Furthermore, complete or partial closures of emergency centers and the number of medical personnel requiring self-isolation decreased in response to the introduction of isolated resuscitation units. However, the adjusted odds ratio and 95% confidence intervals for survival discharge and favorable neurologic outcome were 0.51 (0.25-0.97) and 0.45 (0.21-1.07) compared with those in the 2018 historical cohort. CONCLUSIONS: Responses to the COVID-19 pandemic included changes to current PPE strategies and introduction of isolated resuscitation units; the latter intervention reduced the number of unexpected closures and quarantines of emergency resources early on during the COVID-19 outbreak. Given the possibility of future outbreaks, we need to have revised resuscitation strategies and the capacity to commandeer emergency resources for OHCA patients.

Recurrent hemoptysis in a 26-year-old woman with a ground-glass opacity lesion of the lung.

Hemoptysis is a major reason for emergency department (ED) visits. Catamenial hemoptysis (CH), a rare condition of thoracic endometriosis, can cause recurrent hemoptysis but is difficult to diagnose in the ED due to the scarcity of cases and nonspecific clinical findings. We report a case of a 26-year-old woman who presented to the ED with recurrent hemoptysis since 2 years without a definite cause. Her vital signs and blood test findings were unremarkable. Chest computed tomography (CT) did not show any specific lesions other than a non-specific ground-glass opacity pattern in her right lung. She was on day 4 of her menstrual cycle and her hemoptysis frequently occurred during menstruation. Although there was no histological confirmation, based on her history of hemoptysis during menstruation and no other cause of the hemoptysis, the patient was tentatively diagnosed with CH and was administered gonadotropin-releasing hormone. She had no recurrence of hemoptysis for 3 months. While CH is difficult to diagnose in the ED, the patient's recurrent hemoptysis related to menstruation was a clue to the presence of CH. Therefore, physicians should determine the relationship between hemoptysis and menstruation for women of childbearing age presenting with repeated hemoptysis without a definite cause.

Improving Sensitivity on Identification and Delineation of Intracranial Hemorrhage Lesion Using Cascaded Deep Learning Models.

Highly accurate detection of the intracranial hemorrhage without delay is a critical clinical issue for the diagnostic decision and treatment in an emergency room. In the context of a study on diagnostic accuracy, there is a tradeoff between sensitivity and specificity. In order to improve sensitivity while preserving specificity, we propose a cascade deep learning model constructed using two convolutional neural networks (CNNs) and dual fully convolutional networks (FCNs). The cascade CNN model is built for identifying bleeding; hereafter the dual FCN is to detect five different subtypes of intracranial hemorrhage and to delineate their lesions. Using a total of 135,974 CT images including 33,391 images labeled as bleeding, each of CNN/FCN models was trained separately on image data preprocessed by two different settings of window level/width. One is a default window (50/100[level/width]) and the other is a stroke window setting (40/40). By combining them, we obtained a better outcome on both binary classification and segmentation of hemorrhagic lesions compared to a single CNN and FCN model. In determining whether it is bleeding or not, there was around 1% improvement in sensitivity (97.91% [± 0.47]) while retaining specificity (98.76% [± 0.10]). For delineation of bleeding lesions, we obtained overall segmentation performance at 80.19% in precision and 82.15% in recall which is 3.44% improvement compared to using a single FCN model.

Ultrasonographic findings in Fitz-Hugh-Curtis syndrome: a thickened or three-layer hepatic capsule.

Fitz-Hugh-Curtis syndrome (FHCS) is characterized by inflammation of the perihepatic capsules associated with the pelvic inflammatory disease (PID). FHCS is not a serious disease, but if not treated properly, it can result in increased medical costs, prolonged treatment, and dissatisfaction with treatment. However, early recognition of FHCS in the emergency department can be difficult because its symptoms or physical findings may mimic many other diseases. Although contrast-enhanced computed tomography (CECT) is the useful imaging modality for recognition of FHCS, it is available only when a high suspicion is established. We performed point-of-care ultrasonography in an 18-year-old woman who had a sharp right upper quadrant (RUQ) abdominal pain without PID symptoms and found a thickened or three-layer hepatic capsule. These findings coincided with areas showing increased hepatic capsular enhancement in the arterial phase of CECT. These results show that if the thickened or three-layer hepatic capsule without evidence of a common cause of RUQ pain is observed on ultrasonography in women of childbearing age with RUQ abdominal pain, the physician can consider the possibility of FHCS.

Renal assessment using CKD-EPI equation is useful as an early predictor of contrast- induced nephropathy in elderly patients with cancer.

OBJECTIVES: To assess respective roles of serum creatinine (SCr) alone and estimated glomerular filtration rate (eGFR) as an early predictor for contrast-induced nephropathy (CIN) in elderly patients with cancer. MATERIALS AND METHODS: eGFR of 348 patients at 65years or older with malignancy who underwent contrast-enhanced computed tomography (CECT) were calculated. eGFR was calculated based on the following three equations: Chronic Kidney Disease Epidemiology Collaboration equation (CKD-EPI); Modification of Diet in Renal Disease Study (MDRD); Cockcroft-Gault (CG). CIN was subdivided into two groups: CIN25% (SCr increase >25% but ≤0.5mg/dl), and CIN0.5 (SCr increase >0.5mg/dl). The occurrence and clinical outcomes of CIN were determined according to SCr and eGFR. RESULTS: After CECT, CIN occurred in 50 (14.4%) patients, including 33 CIN25% patients and 17 CIN0.5 patients. CIN0.5 was significantly correlated with prolonged hospitalizations and increased in-hospital mortality, but not CIN25%. Despite SCr<1.5mg/dl, preexisting renal insufficiency (RI) was observed in 47 (13.5%) patients based on CKD-EPI equation, 50 (14.4%) patients based on MDRD equation, and 144 (41.4%) patients based on CG formula. In preexisting RI, the prevalence of CIN0.5 had an odds ratio of 15.02 (5.24 to 43.07) based on CKD-EPI equation, 13.73 (4.81 to 39.20) based on MDRD equation, and 5.03 (1.60 to 15.75) based on CG formula. CONCLUSION: In elderly patients with cancer who visit the emergency department, renal assessment before CECT using CKD-EPI equation was superior to SCr alone, MDRD equation, or CG formula in predicting the occurrence of CIN related CECT.

Procalcitonin determined at emergency department as na early indicator of progression to septic shock in patient with sepsis associated with ureteral calculi.

INTRODUCTION: To investigate the role of initial procalcitonin (PCT) level as an early predictor of septic shock for the patient with sepsis induced by acute pyelonephritis (APN) secondary to ureteral calculi. MATERIALS AND METHODS: The data from 49 consecutive patients who met criteria of sepsis due to APN following ureteral stone were collected and divided into two groups: with (n=15) or without (n=34) septic shock. The clinical variables including PCT level for this outcome were retrospectively compared by univariate analysis, followed by multivariable logistic regression model. RESULTS: All subjects had hydronephrosis, and were hospitalized with the mean of 11.8 days (3-42 days). The mean size of the ureteral stones was 7.5mm (3-30mm), and 57% were located in upper ureter. At univariate analysis, patients with septic shock were significantly older, a higher proportion had hypertension, lower platelet count and serum albumin level, higher CRP and PCT level, and higher positive blood culture rate. Multivariate models indicated that lower platelet count and higher PCT level are independent risk factors (p=0.043 and 0.046, respectively). In ROC curve, the AUC was significantly wider in PCT (0.929), compared with the platelet count (0.822, p=0.004). At the cut-off of 0.52ng/mL, the sensitivity and specificity were 86.7% and 85.3%. CONCLUSION: Our study demonstrated elevated initial PCT levels as an early independente predictor to progress into septic shock in patients with sepsis associated with ureteral calculi.

Procalcitonin determined at emergency department as an early indicator of progression to septic shock in patient with sepsis associated with ureteral calculi

ABSTRACT Introduction: To investigate the role of initial procalcitonin (PCT) level as an early predictor of septic shock for the patient with sepsis induced by acute pyelonephritis (APN) secondary to ureteral calculi. Materials and Methods: The data from 49 consecutive patients who met criteria of sepsis due to APN following ureteral stone were collected and divided into two groups: with (n=15) or without (n=34) septic shock. The clinical variables including PCT level for this outcome were retrospectively compared by univariate analysis, followed by multivariable logistic regression model. Results: All subjects had hydronephrosis, and were hospitalized with the mean of 11.8 days (3–42 days). The mean size of the ureteral stones was 7.5mm (3–30mm), and 57% were located in upper ureter. At univariate analysis, patients with septic shock were significantly older, a higher proportion had hypertension, lower platelet count and serum albumin level, higher CRP and PCT level, and higher positive blood culture rate. Multivariate models indicated that lower platelet count and higher PCT level are independent risk factors (p=0.043 and 0.046, respectively). In ROC curve, the AUC was significantly wider in PCT (0.929), compared with the platelet count (0.822, p=0.004). At the cut-off of 0.52ng/mL, the sensitivity and specificity were 86.7% and 85.3%. Conclusion: Our study demonstrated elevated initial PCT levels as an early independent predictor to progress into septic shock in patients with sepsis associated with ureteral calculi.

High-dose vitamin C management in dapsone-induced methemoglobinemia.

Methylene blue is the first-choice treatment of methemoglobinemia, but it is not readily available in most Korean emergency departments because of an import suspension. An 84-year-old woman with dapsone-induced massive methemoglobinemia visited our emergency department for unclear mentality and cyanosis. Because methylene blue was not available, we intravenously administrated vitamin C (VC) for symptomatic methemoglobinemia, although VC is not a universally accepted treatment. Vitamin C (10 g intravenously) administered 6 hourly successfully treated the dapsone-induced methemoglobinemia and did not adversely affect renal functions. Thus,we recommend that if methylene blue is unavailable, 6 hourly intravenous administrations of 10 g of VC should be considered for dapsone-induced methemoglobinemia.