Evaluation of deep learning for detecting intraosseous jaw lesions in cone beam computed tomography volumes.

OBJECTIVE: The study aim was to develop and assess the performance of a deep learning (DL) algorithm in the detection of radiolucent intraosseous jaw lesions in cone beam computed tomography (CBCT) volumes. STUDY DESIGN: A total of 290 CBCT volumes from more than 12 different scanners were acquired. Fields of view ranged from 6 × 6 × 6 cm to 18 × 18 × 16 cm. CBCT volumes contained either zero or at least one biopsy-confirmed intraosseous lesion. 80 volumes with no intraosseous lesions were included as controls and were not annotated. 210 volumes with intraosseous lesions were manually annotated using ITK-Snap 3.8.0. 150 volumes (10 control, 140 positive) were presented to the DL software for training. Validation was performed using 60 volumes (30 control, 30 positive). Testing was performed using the remaining 80 volumes (40 control, 40 positive). RESULTS: The DL algorithm obtained an adjusted sensitivity by case, specificity by case, positive predictive value by case, and negative predictive value by case of 0.975, 0.825, 0.848, and 0.971, respectively. CONCLUSIONS: A DL algorithm showed moderate success at lesion detection in their correct locations, as well as recognition of lesion shape and extent. This study demonstrated the potential of DL methods for intraosseous lesion detection in CBCT volumes.

Incidental findings involving the temporomandibular joint on computed tomography and magnetic resonance imaging.

The temporomandibular joint (TMJ) is frequently imaged in head and neck computed tomography (CT) and magnetic resonance imaging (MRI) studies. Depending on the indication for the study, an abnormality of the TMJ may be an incidental finding. These findings encompass both intra- and extra-articular disorders. They may also be related to local, regional or systemic conditions. Familiarity with these findings along with pertinent clinical information helps narrow the list of differential diagnoses. While definitive diagnosis may not be immediately apparent, a systematic approach contributes to improved discussions between clinicians and radiologists and better patient management.

Comparison of radiographical characteristics and diagnostic accuracy of intraosseous jaw lesions on panoramic radiographs and CBCT.

OBJECTIVE: To investigate the differences in imaging characteristics and the diagnostic accuracy of 225 intraosseous jaw lesions on panoramic radiographs (PAN) versus cone beam computed tomography (CBCT). METHODS: 225 sets of PAN and CBCT images with biopsy-proven histopathological diagnoses were retrospectively compared in terms of radiographic features and diagnostic accuracy. The imaging characteristics of PAN and CBCT were independently evaluated by two oral and maxillofacial radiologists who were required to answer 12 questions and provided up to three differential diagnoses with their confidence scores. RESULTS: Odds ratios (ORs) were statistically significant for border cortication (OR = 1.521; p = .003) and border continuity (OR = 0.421; p = .001), involvement on neurovascular canals (OR = 2.424; p < .001), expansion (OR = 7.948; p < .001), cortical thinning (OR = 20.480; p < .001) as well as its destruction (OR = 25.022; p < .001) and root resorption (OR = 2.477; p < .001). Furthermore, imaging features in the posterior and mandibular regions showed better agreement than those in the anterior and maxillary regions, respectively. The diagnostic accuracy of the first differential diagnosis was higher on CBCT than on PAN (Observer 1:78.7 vs 64.4%; Observer 2: 78.7 vs 70.2% (p < .001)). The observers' confidence scores were also higher at CBCT interpretation compared with PAN. CONCLUSIONS: CBCT demonstrated a greater number of imaging characteristics of intraosseous jaw lesions compared with PAN, especially in the anterior regions of both jaws and in the maxilla. Diagnostic accuracy is improved with CBCT compared to PAN, especially for lesions in the maxilla. Radiologists have greater confidence when using CBCT.

Comparing panoramic radiographs and cone beam computed tomography: Impact on radiographic features and differential diagnoses.

OBJECTIVES: The aims of this study were to determine whether lesion features appear differently on panoramic radiography (PAN) and cone beam computed tomography (CBCT), and whether the use of CBCT affects diagnostic accuracy and observers' confidence in comparison with PAN. STUDY DESIGN: Three oral and maxillofacial radiologists reviewed 33 sets of PAN images and CBCT volumes of biopsy-proven lesions. They described 12 different lesion features and provided up to 3 ranked differential diagnoses, as well as their confidence with respect to those diagnoses. Their confidence was weighted by the rank at which the correct diagnosis was provided. RESULTS: Odds ratios (ORs) were statistically significant for border definition (OR = 5.45; P = .004), continuity of border cortication (OR = 0.34; P = .035), effect on neurovascular canals (OR = 6.38; P = .043), expansion (OR = 18.56; P < .001), cortical thinning (OR = 30.22; P < .001), and cortical destruction (OR = 9.80; P < .001). There was no association between the 2 modalities and the rank at which the correct differential diagnoses were made or the observers' weighted confidence. CONCLUSIONS: Before acquiring a CBCT scan to aid in the diagnosis of an intraosseous lesion, clinicians should consider the diagnostic information that is expected to be gained. In this study, although there were differences between PAN and CBCT with respect to some lesion features, CBCT did not help improve diagnostic accuracy.

Evaluation of the Sorbent Properties of Single- and Multiwalled Carbon Nanotubes for Volatile Organic Compounds through Thermal Desorption-Gas Chromatography/Mass Spectrometry.

Carbon nanotubes (CNTs) possess well-defined structural and chemical characteristics coupled with a large surface area that makes them ideal as sorbent materials for applications where adsorption processes are required. The adsorption properties of carboxylated derivatives of multiwalled carbon nanotubes (COOH-MWCNT) and singlewalled carbon nanotubes (COOH-SWCNT), together with their nonfunctionalized counterparts (MWCNT and SWCNT) for 48 common atmospheric volatile organic compounds (VOCs) were determined using thermal desorption-gas chromatography/mass spectrometry (TD-GCMS). The CNTs exhibited similar recoveries for many of the VOCs compared to the standard sorbent materials, Carbopack X and Tenax TA. However, VOCs with electron donor-acceptor (EDA) properties such as carbonyls, alkenes, and alcohols exhibited poorer recoveries on all CNTs compared to Carbopack X and Tenax TA. The poor recoveries of VOCs from the CNTs has important implications for the long term use and storage of CNTs, because it demonstrates that they will become progressively more contaminated with common atmospheric VOCs, therefore potentially affecting their surface-based properties.