A deep learning-based technique for the diagnosis of epidural spinal cord compression on thoracolumbar CT

PURPOSE

To develop a deep learning (DL) model for epidural spinal cord compression (ESCC) on CT, which will aid earlier ESCC diagnosis for less experienced clinicians.

METHODS

We retrospectively collected CT and MRI data from adult patients with suspected ESCC at a tertiary referral institute from 2007 till 2020. A total of 183 patients were used for training/validation of the DL model. A separate test set of 40 patients was used for DL model evaluation and comprised 60 staging CT and matched MRI scans performed with an interval of up to 2 months. DL model performance was compared to eight readers: one musculoskeletal radiologist, two body radiologists, one spine surgeon, and four trainee spine surgeons. Diagnostic performance was evaluated using inter-rater agreement, sensitivity, specificity and AUC.

RESULTS

Overall, 3115 axial CT slices were assessed. The DL model showed high kappa of 0.872 for normal, low and high-grade ESCC (trichotomous), which was superior compared to a body radiologist (R4, κ = 0.667) and all four trainee spine surgeons (κ range = 0.625-0.838)(all p < 0.001). In addition, for dichotomous normal versus any grade of ESCC detection, the DL model showed high kappa (κ = 0.879), sensitivity (91.82), specificity (92.01) and AUC (0.919), with the latter AUC superior to all readers (AUC range = 0.732-0.859, all p < 0.001).

CONCLUSION

A deep learning model for the objective assessment of ESCC on CT had comparable or superior performance to radiologists and spine surgeons. Earlier diagnosis of ESCC on CT could reduce treatment delays, which are associated with poor outcomes, increased costs, and reduced survival.

Electric Scooter Battery Detonation: A Case Series And Review Of Literature

Since 2016 there has been a 20-fold increase in known burns injury from personal mobility device (PMD) related fires. The root cause is the failure of high-density lithium ion (Li-ion) battery packs powering the PMDs. This failure process, known as thermal runaway, is well documented in applied science journals. Importantly, the liberation of hydrogen fluoride from failing Li-ion batteries may contribute to unrecognized chemical burns. A clinical gap in knowledge exists in the understanding of the explosive nature of Li-ion batteries. We reviewed the electrochemical pathophysiology of a failing Li-ion cell as it impacts clinical management of burn injuries. This retrospective study was carried out in two major institutions in Singapore. All admitted PMD-related burns and follow up appointments were captured and reviewed from 2016 - 2020. Thirty patients were admitted to tertiary hospitals, 43% of patients were in the pediatric population and 57% were adult patients, aged from 0.3 to 77 years. TBSA of burns ranged from 0 to 80% with a mean 14.5%. 73% of cases presented with inhalation injury, 8 of whom did not suffer any cutaneous burns. 50% of patients sustained both cutaneous and inhalation burn injuries. 27% of patients sustained major burns of >20% TBSA, with 2 in the pediatric group. Mortali ty rate was 10% from PMD-related fires. This cause of burn injury has proven to be fa tal. Prevention of PMD-related fires by ensuring proper battery utilization, adherence to PMD sanctions for battery standards and public education is vital to reducing the morbidity and mortality of this unique type of thermal injury.

Quantitative investigation on the critical thickness of the dielectric shell for metallic nanoparticles determined by the plasmon decay length

Inert dielectric shells coating the surface of metallic nanoparticles (NPs) are important for enhancing the NPs' stability, biocompatibility, and realizing targeting detection, but they impair NPs' sensing ability due to the electric fields damping. The dielectric shell not only determines the distance of the analyte from the NP surface, but also affects the field decay. From a practical point of view, it is extremely important to investigate the critical thickness of the shell, beyond which the NPs are no longer able to effectively detect the analytes. The plasmon decay length of the shell-coated NPs determines the critical thickness of the coating layer. Extracting from the exponential fitting results, we quantitatively demonstrate that the critical thickness of the shell exhibits a linear dependence on the NP volume and the dielectric constants of the shell and the surrounding medium, but only with a small variation influenced by the NP shape where the dipole resonance is dominated. We show the critical thickness increases with enlarging the NP sizes, or increasing the dielectric constant differences between the shell and surrounding medium. The findings are essential for applications of shell-coated NPs in plasmonic sensing.