Application of convolutional neural networks for distal radio-ulnar fracture detection on plain radiographs in the emergency room

Objective@#Recent studies have suggested that deep-learning models can satisfactorily assist in fracture diagnosis. We aimed to evaluate the performance of two of such models in wrist fracture detection. @*Methods@#We collected image data of patients who visited with wrist trauma at the emergency department. A dataset extracted from January 2018 to May 2020 was split into training (90%) and test (10%) datasets, and two types of convolutional neural networks (i.e., DenseNet-161 and ResNet-152) were trained to detect wrist fractures. Gradient-weighted class activation mapping was used to highlight the regions of radiograph scans that contributed to the decision of the model. Performance of the convolutional neural network models was evaluated using the area under the receiver operating characteristic curve. @*Results@#For model training, we used 4,551 radiographs from 798 patients and 4,443 radiographs from 1,481 patients with and without fractures, respectively. The remaining 10% (300 radiographs from 100 patients with fractures and 690 radiographs from 230 patients without fractures) was used as a test dataset. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of DenseNet-161 and ResNet-152 in the test dataset were 90.3%, 90.3%, 80.3%, 95.6%, and 90.3% and 88.6%, 88.4%, 76.9%, 94.7%, and 88.5%, respectively. The area under the receiver operating characteristic curves of DenseNet-161 and ResNet-152 for wrist fracture detection were 0.962 and 0.947, respectively. @*Conclusion@#We demonstrated that DenseNet-161 and ResNet-152 models could help detect wrist fractures in the emergency room with satisfactory performance.

Application of convolutional neural networks for distal radio-ulnar fracture detection on plain radiographs in the emergency room

Objective@#Recent studies have suggested that deep-learning models can satisfactorily assist in fracture diagnosis. We aimed to evaluate the performance of two of such models in wrist fracture detection. @*Methods@#We collected image data of patients who visited with wrist trauma at the emergency department. A dataset extracted from January 2018 to May 2020 was split into training (90%) and test (10%) datasets, and two types of convolutional neural networks (i.e., DenseNet-161 and ResNet-152) were trained to detect wrist fractures. Gradient-weighted class activation mapping was used to highlight the regions of radiograph scans that contributed to the decision of the model. Performance of the convolutional neural network models was evaluated using the area under the receiver operating characteristic curve. @*Results@#For model training, we used 4,551 radiographs from 798 patients and 4,443 radiographs from 1,481 patients with and without fractures, respectively. The remaining 10% (300 radiographs from 100 patients with fractures and 690 radiographs from 230 patients without fractures) was used as a test dataset. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of DenseNet-161 and ResNet-152 in the test dataset were 90.3%, 90.3%, 80.3%, 95.6%, and 90.3% and 88.6%, 88.4%, 76.9%, 94.7%, and 88.5%, respectively. The area under the receiver operating characteristic curves of DenseNet-161 and ResNet-152 for wrist fracture detection were 0.962 and 0.947, respectively. @*Conclusion@#We demonstrated that DenseNet-161 and ResNet-152 models could help detect wrist fractures in the emergency room with satisfactory performance.

Validation and modification of HEART score components for patients with chest pain in the emergency department

Objective@#This study aimed to clarify the relative prognostic value of each History, Electrocardiography, Age, Risk Factors, and Troponin (HEART) score component for major adverse cardiac events (MACE) within 3 months and validate the modified HEART (mHEART) score. @*Methods@#This study evaluated the HEART score components for patients with chest symptoms visiting the emergency department from November 19, 2018 to November 19, 2019. All components were evaluated using logistic regression analysis and the scores for HEART, mHEART, and Thrombolysis in Myocardial Infarction (TIMI) were determined using the receiver operating characteristics curve. @*Results@#The patients were divided into a derivation (809 patients) and a validation group (298 patients). In multivariate analysis, age did not show statistical significance in the detection of MACE within 3 months and the mHEART score was calculated after omitting the age component. The areas under the receiver operating characteristics curves for HEART, mHEART and TIMI scores in the prediction of MACE within 3 months were 0.88, 0.91, and 0.83, respectively, in the derivation group; and 0.88, 0.91, and 0.81, respectively, in the validation group. When the cutoff value for each scoring system was determined for the maintenance of a negative predictive value for a MACE rate >99%, the mHEART score showed the highest sensitivity, specificity, positive predictive value, and negative predictive value (97.4%, 54.2%, 23.7%, and 99.3%, respectively). @*Conclusion@#Our study showed that the mHEART score better detects short-term MACE in high-risk patients and ensures the safe disposition of low-risk patients than the HEART and TIMI scores.

Perioperative Outcomes of Different Surgical Methods Among Bladder Cancer Patients Undergoing Radical Cystectomy With Neobladder Urinary Diversion

Purpose@#To compare perioperative outcomes according to surgical methods among bladder cancer patients who underwent radical cystectomy (RC) with neobladder urinary diversion. @*Materials and Methods@#Between June 2007 and January 2020, 89 bladder cancer patients who received RC with neobladder urinary diversion were enrolled in this study. Patients were stratified into surgical methods – (1) open RC with neobladder (ONB) reconstruction, (2) robotassisted RC (RARC) with extracorporeal neobladder (ECNB) reconstruction, and (3) RARC with intracorporeal neobladder (ICNB) reconstruction. Perioperative outcomes were compared among the 3 groups, with major complications defined according to Clavien-Dindo grades III–V within 90 days. Logistic regression analysis was performed to identify significant factors for postoperative complications. @*Results@#Of 89 patients, 28 (31%) had ONB, 31 (35%) had ECNB, and 30 (34%) had ICNB. The median operative time was 471 minutes, and the ICNB group (424.5 minutes) was significantly less than ONB (444.5 minutes) and ECNB groups (542.9 minutes) (p=0.001). Transfusion rate was also significantly less in the ICNB group (13%) (p=0.001). Complications were recorded in 67 patients (75%) and major complications in 22 of all patients (25%). The major complication rate was significantly less in ICNB (13.4%) than in ONB (25%) and ECNB (35%) (p=0.003). Multivariate analysis showed surgical methods (ICNB) (odds ratio [OR], 0.709; p=0.003) and age (OR, 1.150; p=0.001) were significant factors related to occurrence of major postoperative complications. @*Conclusions@#RARC with ICNB reduces postoperative complications compared to ONB and ECNB.

Long-Term Outcomes of Acute Ischemic Stroke in Patients Aged 80 Years and Older

PURPOSE: Short life expectancy influences decision-making when treating very old patients with acute ischemic stroke (AIS). We investigated mortality and survival duration in very old AIS patients (> or =80 years) who received hospital care. PATIENTS AND METHODS: Mortality data were obtained from medical records, structured telephone inquiries, death certificates from the Korean National Statistical Office, and social security data 5+/-1.9 years after stroke onset. Age, gender, vascular risk factors, and functional outcomes from modified Rankin scales (MRS) at discharge were analyzed as predictors of mortality. RESULTS: Among 134 patients, 92 (68.7%) died. On Kaplan- Meier analysis, duration of survival of patients aged 80-84 years was longer than those aged 85-89 or 90-94 (24+/-6.4, 8+/-7.3, 7+/-2.0 months, respectively, p=0.002). Duration of survival of patients discharged in a state of MRS 0-1 was longer than the remaining groups at 47+/-4.8 months (p<0.001). In Cox proportional hazard analysis, age and MRS at discharge were independent predictors of mortality. CONCLUSION: Long-term outcomes of very old patients with AIS are not uniformly grave, therefore predictors of mortality and estimated duration of survival should be considered during decision- making for treatment.

Analysis of 10,811 Cases with Acute Ischemic Stroke from Korean Stroke Registry: Hospital-Based Multicenter Prospective Registration Study

BACKGROUND: Although several hospital-based stroke studies were published, there has not been any reliable data representing the clinical characteristics of stroke in Korea. We analyzed the clinical characteristics of patients with ischemic stroke registered in the Korean Stroke Registry (KSR), which is the largest prospective hospital-based nation-wide stroke registry in Korea. METHODS: The KSR provided standardized protocols for collecting data, which includes the data of demographics, subtypes of stroke, risk factors, and neurological outcome at discharge. The brain imaging studies, including CT or MRI, were performed in all cases. RESULTS: KSR registered 10,811 patients of acute ischemic stroke between Nov. 2002 and Jun. 2004. The large-artery atherosclerosis was the most common subtype (37.3%), followed by small vessel occlusion (30.8%). Hypertension (65.4%) was the most common risk factor, followed by smoking (34.5%) and diabetes (28.3%). Although most of the hypertensive and diabetic patients had been diagnosed before the stroke, less than 45.4% and 32.5% of them were under regular control. The steno-occlusive lesion of extracranial carotid artery was only 29.3% and the ratio of intra- to extracranial artery disease was more than 2 in KSR. Only 20.5% of patients were admitted within 3 hours after stroke onset and 2.1% were treated with intravenous thrombolysis. In-hospital case-fatality was 5.2%, which is relatively comparable to those of previous studies. CONCLUSIONS: The KSR provided informative data in understanding the clinical characteristics of ischemic stroke in Korea. Further analysis of KSR will facilitate clinical trials and development of guidelines for the management of stroke patients.

Newly Onset Seizures in the Elderly: A Hospital-based Study

BACKGROUND: Increasing incidences of epilepsy in the elderly are well-known. However, the causes of newly onset seizures in the elderly have rarely been described in Korea. METHODS: We selected 160 cases of individuals who had their first seizure at over the age of 60 (male : female = 1.5 : 1 ; mean age : 69.1 years), who were admitted to Hallym University Hospital from July 1, 1994 to June 31, 1998. We analyzed the etiology, type of seizures, EEG, neuro-imag-ing, morbidity and mortality of the patients. RESULTS: The etiologies of seizures were remote symptomatic in 87 (54.3%), acute symptomatic in 38 (23.8%), progressive encephalopathy in 21 (13.1%), and idiopathic in 14 (8.8%). Status epilepticus occurred in 34 cases, including 8 cases of multifocal myoclonic status after hypoxic brain damage. The most common single cause of seizure was old stroke (35%, infarction in 41 and hemorrhage in 15 cases). Partial seizure was more common in patients with remote symptomatic than with other causes. Newly developed neurological deficits were present in 30 of the 151 who survived, including 15 acute symptomatic, 9 remote symptomatic, and 6 pro-gressive encephalopathy cases. Morbidity and mortality were highest in the acute symptomatic group (P<0.05) and tend to be low in the idiopathic group. CONCLUSIONS: We conclude that newly onset seizures in the elderly requiring hospitalization occur mainly with acute and remote symptomatic neurological insults. Acute symptomatic neurological insults are associated with a significant morbidity and mortality, while the morbidity is low in the absence of any asso-ciated neurological insults.

A Case of Posterior Spinal Artery Infarction after Cervical Trauma

Clinically, the infarction of posterior spinal arteries is rarely recognized due to rich anastomosis. As a result, there have been few clinical reports of posterior spinal artery infarction. A 38-year-old man experienced severe transitory neck and occipital pain after his friend had struck him on the cervical area. A few days later, he developed dysmetria, dysdiadochokinesia, and decreased vibration and position senses on the right side of his body. Routine laboratory find-ings, an echocardiogram, a work-up for connective tissue diseases, and CSF studies were all found to be normal. A MRI showed increased signals in the right posterior and posterolateral part of the lower medulla and some portion of the first cervical cord on T2- and proton-weighted images without significant enhancements. A cerebral angiogram showed a long narrow thread-like segment in the distal portion of the right vertebral artery, which was indicative of a dissection. The right posterior inferior cerebellar artery was not visualized.

Anterior Spinal Artery Syndrome Associated With Cervical Spinal Stenosis

BACKGROUND: The anterior spinal artery infarction constitutes a classical syndrome of vascular myelopathy. The causes of the anterior spinal artery syndrome are various, but most episode probably occur as the result of atherosclerosis or dissection of the aorta and its branches. However, few cases reported developed with spinal structural abnormalities. CASE: A 65-year-old man presented with sudden paraparesis. There was no evidence of hypertension, diabetes and smoking. Motor weakness was more prominent on the left side and progressed. Loss of pain and temperature senses were shown at the level of 71 with preservation of touch, joint perception and vibration senses. The DTR's of legs were depressed and extensor toe signs were presented. A C-spine MRI showed high signal intensity on 72 weighted image and low signal on 71 weighted image(C6-71) with cervical spinal stenosis at the C4-C7 spinal level and mild cervical disc protrusion (C6-C7, C7-T1). After three months later, follow up cervical MRI showed somewhat decreased size of high signal intensity on 72 weighted image and more prominent low signal on 71 image. DISCUSSION: In our case, we could not find any usual cause of anterior spinal artery infarction. However only cervical spinal stenosis associated with mild cervical disc protrusion was present. In stenotic cervical canal, the anterior spinal artery can be more vulnerable to extrinsic compression and the infarction may early develop with insignificant trigger event, such as disc protrusion. We concluded that the ischemic change of anterior two thirds of cervical spinal cord might develop due to the compression of the anterior spinal artery by cervical stenosis and mild cervical intervertebral disc protrusion.