Extravalidation and reproducibility results of a commercial deep learning-based automatic detection algorithm for pulmonary nodules on chest radiographs at tertiary hospital.

INTRODUCTION: To extra validate and evaluate the reproducibility of a commercial deep convolutional neural network (DCNN) algorithm for pulmonary nodules on chest radiographs (CRs) and to compare its performance with radiologists. METHODS: This retrospective study enrolled 434 CRs (normal to abnormal ratio, 246:188) from 378 patients that visited a tertiary hospital. DCNN performance was compared with two radiology residents and two thoracic radiologists. Abnormality assessment (using the area under the receiver operating characteristics (AUROC)) and nodule detection (using jackknife alternative free-response ROC (JAFROC)) were compared among three groups (DCNN only, radiologist without DCNN and radiologist with DCNN). A subset of 56 paired cases, having two CRs taken within a 7-day period, were assessed for intraobserver reproducibility using the intraclass correlation coefficient. Independent characteristics of pulmonary nodules detected by DCNN were assessed by multiple logistic regression analysis. RESULTS: The AUROC for abnormality detection for the three groups were 0.87, 0.93 and 0.96, respectively (P < 0.05), whereas the JAFROC analysis of nodule detection was 0.926, 0.929 and 0.964. Reproducibility for the three groups was 0.80, 0.67 and 0.80, which shows an increase in radiologists using DCNN (P < 0.05). Nodules detected by DCNN were more solid, round-shaped and well marginated, not masked and laterally located (P < 0.05). CONCLUSIONS: Extra validation results of DCNN showed high ROC results and there was a significant improvement in the performance when radiologists used DCNN. Reproducibility by DCNN alone showed good agreement, and there was an improvement from moderate to good agreement for radiologists using DCNN.

Acute vertigo and sensorineural hearing loss from infarction of the vestibulocochlear nerve: A case report.

RATIONALE: Acute unilateral audiovestibulopathy is a common neurotological syndrome. Differential diagnoses of acute unilateral audiovestibulopathy include viral infection, vascular insults, and tumors. Regarding vascular causes, ischemic stroke in the anterior inferior cerebellar artery (AICA) territory is known to be the leading cause of acute audiovestibular loss. Previous reports of AICA infarction with audiovestibulopathy failed to demonstrate magnetic resonance imaging (MRI)-positive vestibulocochlear infarctions. Only 1 report demonstrated acute infarction involving the vestibulocochlear nerve on diffusion weighted imaging (DWI)-MRI. PATIENT CONCERNS: A 67 year old man complained of sudden left hearing loss and vertigo. The patient showed left horizontal gaze-evoked nystagmus (GEN) and the head impulse test (HIT) was positive on the left side. Videonystagmography revealed spontaneous rebound nystagmus toward the right side; head-shaking nystagmus toward the right side. The patient presented with left caloric paresis (20.1%). Pure tone audiometry (PTA) revealed severe sensorineural hearing loss on the left side. DIAGNOSIS: MRI of temporal bone showed multifocal acute infarctions in the left inferior cerebellum. Moreover, images revealed tiny infarctions along the left vestibulocochlear nerve and the cochlea, implying acute vestibulocochlear nerve and labyrinthine infarction. There was no evidence of steno-occlusion of major cerebral vessels on MR angiography. INTERVENTIONS: Immediate stroke management was done. OUTCOMES: Neurological symptoms gradually improved after 3 to 5 days. LESSONS: We present a case illustrating a rare but significant finding of vestibulocochlear nerve infarction revealed by DWI-MRI. Prompt imaging protocol enabled the detection of significant findings in this patient with acute unilateral audiovestibulopathy. Clinicians should be aware of the vestibulocochlear nerve and labyrinth on MRI in patients with cerebellar stroke.

Evaluation of neuraminidase antigen based competitive enzyme-linked immunosorbent assay in chickens vaccinated with avian influenza inactivated vaccine.

A competitive enzyme-linked immunosorbent assay (c-ELISA) was developed as a serologic diagnostic tool to detect antibodies against NA subtype 3 of avian influenza virus (AIV). The NA antigen used in this c-ELISA was obtained by pronase treatment of allantoic fluid of specific-pathogen-free (SPF) eggs infected with AIV. The NA specific monoclonal antibodies were produced from purified NA. The N3 c-ELISA was carried out on serum samples collected from both SPF chickens and commercial layers to confirm whether the N3 c-ELISA was capable of detecting specific N3 antibodies. The positive cutoff percentage inhibition value was 6.13%. The sensitivity and specificity of the N3 c-ELISA were 83.7% and 95.6%, respectively, which indicated that N3 c-ELISA can detect the antibodies from SPF chickens or commercial chickens vaccinated with H9N3 subtype of AIV.