Search

In recent years, there have been major advances in deep learning algorithms for image recognition in traumatic brain injury (TBI). Interest in this area has increased due to the potential for greater objectivity, reduced interpretation times and, ultimately, higher accuracy. Triage algorithms that can re-order radiological reading queues have been developed, using classification to prioritize exams with suspected critical findings. Localization models move a step further to capture more granular information such as the location and, in some cases, size and subtype, of intracranial hematomas that could aid in neurosurgical management decisions. In addition to the potential to improve the clinical management of TBI patients, the use of algorithms for the interpretation of medical images may play a transformative role in enabling the integration of medical images into precision medicine. Acute TBI is one practical example that can illustrate the application of deep learning to medical imaging. This review provides an overview of computational approaches that have been proposed for the detection and characterization of acute TBI imaging abnormalities, including intracranial hemorrhage, skull fractures, intracranial mass effect, and stroke. [ABSTRACT FROM AUTHOR]

Objectives: Subglottic stenosis (SGS) is one of the most common airway disorders in pediatric patients. Currently, treatment decisions rely primarily on the Cotton-Myer scale, which classifies SGS severity based on percentage reduction in airspace cross-sectional area (CSA). However, the precise relationship between upper airway resistance and subglottic CSA is unknown. We hypothesize that airway resistance can be described by the Bernoulli Obstruction Theory, which predicts that airway resistance is inversely proportional to airspace CSA ( R∝A-1) in cases of severe constriction.Methods: Computed tomography (CT) scans of six healthy subjects and five SGS patients were used to create three-dimensional models of the respiratory tract from nostrils to carina. Cylindrical segments of varying lengths and varying diameters were digitally inserted in the subglottis of the healthy subjects to create simulated SGS models. Computational fluid dynamics simulations were run, and airway resistance was computed in the simulated SGS models and actual SGS models.Results: Constriction diameter had a greater impact in airway resistance than constriction length. In agreement with the Bernoulli Obstruction Theory, airway resistance in the simulated SGS models was well represented by the power law R=aAb, where a is a constant and the exponent b ranged from -0.85 to -1.07. The percentage reduction in airflow (QOBSTRUCTIONQHEALTHY) at a constant pressure drop was found to be directly proportional to the percentage reduction in CSA (AOBSTRUCTIONAHEALTHY) in the limit of severe constrictions, namely QOBSTRUCTIONQHEALTHY=kAOBSTRUCTIONAHEALTHY, where k=2.25 ± 0.15. Airway resistances in the simulated SGS models were similar to resistances in models based on CT scans of actual SGS patients, suggesting that our simulated SGS models were representative of airway resistance in actual SGS patients.Conclusion: Our computer simulations suggest that the degree of airflow limitation in SGS patients may be estimated based on anatomic measurements alone. Future studies are recommended to test these predictions in larger cohorts.Level Of Evidence: 4. Laryngoscope, 128:1551-1557, 2018. [ABSTRACT FROM AUTHOR]