Computed tomography pseudofractures of the mid face and skull base

Computational Approaches for Acute Traumatic Brain Injury Image Recognition

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.

Imaging of Skull Base Trauma: Fracture Patterns and Soft Tissue Injuries

This article provides an overview of the patterns of skull base trauma and provides a review of the pertinent soft tissue injuries and complications that can ensue. A brief review of skull base anatomy is provided with subsequent focus on the important findings in anterior, central, and posterior skull base trauma.

Bone Age Assessment Empowered with Deep Learning: A Survey, Open Research Challenges and Future Directions

Deep learning is a quite useful and proliferating technique of machine learning. Various applications, such as medical images analysis, medical images processing, text understanding, and speech recognition, have been using deep learning, and it has been providing rather promising results. Both supervised and unsupervised approaches are being used to extract and learn features as well as for the multi-level representation of pattern recognition and classification. Hence, the way of prediction, recognition, and diagnosis in various domains of healthcare including the abdomen, lung cancer, brain tumor, skeletal bone age assessment, and so on, have been transformed and improved significantly by deep learning. By considering a wide range of deep-learning applications, the main aim of this paper is to present a detailed survey on emerging research of deep-learning models for bone age assessment (e.g., segmentation, prediction, and classification). An enormous number of scientific research publications related to bone age assessment using deep learning are explored, studied, and presented in this survey. Furthermore, the emerging trends of this research domain have been analyzed and discussed. Finally, a critical discussion section on the limitations of deep-learning models has been presented. Open research challenges and future directions in this promising area have been included as well.

Validity of computed tomography in diagnosing midfacial fractures

The aim of this study was to evaluate the sensitivity, accuracy, and reliability of two-dimensional computed tomography (2D-CT) scans (axial, coronal, sagittal planes) and three-dimensional computed tomography (3D-CT) reconstructions in diagnosing midfacial fractures in relation to actual fractures identified clinically and during surgery (gold standard). The imaging diagnosis was performed by a radiologist and an oral and maxillofacial surgeon. Sixty-two patients with a total of 429 midfacial fractures were included. Frontal sinus and nose fractures were easily diagnosed. For the three CT planes, there was a statistically significant difference between the CT examination and the gold standard for five to seven of the nine bones evaluated, while for 3D-CT, a difference was observed only for fractures of the orbital floor. The inter-observer agreement between the oral and maxillofacial surgeon and the radiologist was 75.5%. In conclusion, in this study 3D-CT reconstructions showed significantly the best sensitivity, accuracy, and reliability for the diagnosis of midfacial fractures. The sagittal reconstructions were the least diagnostic of the 2D-CT images. For areas where the parameters studied showed less agreement and hence a more difficult diagnosis, we recommend a combination of 3D and 2D-CT images to improve diagnostic accuracy.

A Protocol to Reduce Interobserver Variability in the Computed Tomography Measurement of Orbital Floor Fractures

Orbital fracture detection and size determination from computed tomography (CT) scans affect the decision to operate, the type of surgical implant used, and postoperative outcomes. However, the lack of standardization of radiological signs often leads to the false-positive detection of orbital fractures, while nonstandardized landmarks lead to inaccurate defect measurements. We aim to design a novel protocol for CT measurement of orbital floor fractures and evaluate the interobserver variability on CT scan images. Qualitative aspects of this protocol include identifying direct and indirect signs of orbital fractures on CT scan images. Quantitative aspects of this protocol include measuring the surface area of pure orbital floor fractures using computer software. In this study, 15 independent observers without clinical experience in orbital fracture detection and measurement measured the orbital floor fractures of three randomly selected patients following the protocol. The time required for each measurement was recorded. The intraclass correlation coefficient of the surface area measurements is 0.999 (0.997-1.000) with p-value < 0.001. This suggests that any observer measuring the surface area will obtain a similar estimation of the fractured surface area. The maximum error limit was 0.901 cm(2) which is less than the margin of error of 1 cm(2) in mesh trimming for orbital reconstruction. The average duration required for each measurement was 3 minutes 19 seconds (ranging from 1 minute 35 seconds to 5 minutes). Measurements performed with our novel protocol resulted in minimal interobserver variability. This protocol is effective and generated reproducible results, is easy to teach and utilize, and its findings can be interpreted easily.

Three-dimensional fracture visualisation of multidetector CT of the skull base in trauma patients: comparison of three reconstruction algorithms

The purpose of this study was to retrospectively assess the detection rate of skull-base fractures for three different three-dimensional (3D) reconstruction methods of cranial CT examinations in trauma patients. A total of 130 cranial CT examinations of patients with previous head trauma were subjected to 3D reconstruction of the skull base, using solid (SVR) and transparent (TVR) volume-rendering technique and maximum intensity projection (MIP). Three radiologists independently evaluated all reconstructions as well as standard high-resolution multiplanar reformations (HR-MPRs). Mean fracture detection rates for all readers reading rotating reconstructions were 39, 36, 61 and 64% for SVR, TVR, MIP and HR-MPR respectively. Although not significantly different from HR-MPR with respect to sensitivity (P = 0.9), MIP visualised 18% of fractures that were not reported in HR-MPR. Because of the relatively low detection rate using HR-MPRs alone, we recommend reading MIP reconstructions in addition to the obligatory HR-MPRs to improve fracture detection.