Interpretable vertebral fracture quantification via anchor-free landmarks localization.

Vertebral body compression fractures are early signs of osteoporosis. Though these fractures are visible on Computed Tomography (CT) images, they are frequently missed by radiologists in clinical settings. Prior research on automatic methods of vertebral fracture classification proves its reliable quality; however, existing methods provide hard-to-interpret outputs and sometimes fail to process cases with severe abnormalities such as highly pathological vertebrae or scoliosis. We propose a new two-step algorithm to localize the vertebral column in 3D CT images and then detect individual vertebrae and quantify fractures in 2D simultaneously. We train neural networks for both steps using a simple 6-keypoints based annotation scheme, which corresponds precisely to the current clinical recommendation. Our algorithm has no exclusion criteria, processes 3D CT in 2 seconds on a single GPU, and provides an interpretable and verifiable output. The method approaches expert-level performance and demonstrates state-of-the-art results in vertebrae 3D localization (the average error is 1mm), vertebrae 2D detection (precision and recall are 0.99), and fracture identification (ROC AUC at the patient level is up to 0.96). Our anchor-free vertebra detection network shows excellent generalizability on a new domain by achieving ROC AUC 0.95, sensitivity 0.85, specificity 0.9 on a challenging VerSe dataset with many unseen vertebra types.

Outstanding negative prediction performance of solid pulmonary nodule volume AI for ultra-LDCT baseline lung cancer screening risk stratification.

OBJECTIVE: To evaluate performance of AI as a standalone reader in ultra-low-dose CT lung cancer baseline screening, and compare it to that of experienced radiologists. METHODS: 283 participants who underwent a baseline ultra-LDCT scan in Moscow Lung Cancer Screening, between February 2017-2018, and had at least one solid lung nodule, were included. Volumetric nodule measurements were performed by five experienced blinded radiologists, and independently assessed using an AI lung cancer screening prototype (AVIEW LCS, v1.0.34, Coreline Soft, Co. ltd, Seoul, Korea) to automatically detect, measure, and classify solid nodules. Discrepancies were stratified into two groups: positive-misclassification (PM); nodule classified by the reader as a NELSON-plus /EUPS-indeterminate/positive nodule, which at the reference consensus read was < 100 mm3, and negative-misclassification (NM); nodule classified as a NELSON-plus /EUPS-negative nodule, which at consensus read was ≥ 100 mm3. RESULTS: 1149 nodules with a solid-component were detected, of which 878 were classified as solid nodules. For the largest solid nodule per participant (n = 283); 61 [21.6 %; 53 PM, 8 NM] discrepancies were reported for AI as a standalone reader, compared to 43 [15.1 %; 22 PM, 21 NM], 36 [12.7 %; 25 PM, 11 NM], 29 [10.2 %; 25 PM, 4 NM], 28 [9.9 %; 6 PM, 22 NM], and 50 [17.7 %; 15 PM, 35 NM] discrepancies for readers 1, 2, 3, 4, and 5 respectively. CONCLUSION: Our results suggest that through the use of AI as an impartial reader in baseline lung cancer screening, negative-misclassification results could exceed that of four out of five experienced radiologists, and radiologists' workload could be drastically diminished by up to 86.7%.

Inter-Observer Agreement between Low-Dose and Standard-Dose CT with Soft and Sharp Convolution Kernels in COVID-19 Pneumonia.

Computed tomography (CT) has been an essential diagnostic tool during the COVID-19 pandemic. The study aimed to develop an optimal CT protocol in terms of safety and reliability. For this, we assessed the inter-observer agreement between CT and low-dose CT (LDCT) with soft and sharp kernels using a semi-quantitative severity scale in a prospective study (Moscow, Russia). Two consecutive scans with CT and LDCT were performed in a single visit. Reading was performed by ten radiologists with 3-25 years' experience. The study included 230 patients, and statistical analysis showed LDCT with a sharp kernel as the most reliable protocol (percentage agreement 74.35 ± 43.77%), but its advantage was marginal. There was no significant correlation between radiologists' experience and average percentage agreement for all four evaluated protocols. Regarding the radiation exposure, CTDIvol was 3.6 ± 0.64 times lower for LDCT. In conclusion, CT and LDCT with soft and sharp reconstructions are equally reliable for COVID-19 reporting using the "CT 0-4" scale. The LDCT protocol allows for a significant decrease in radiation exposure but may be restricted by body mass index.

Accelerating 3D Medical Image Segmentation by Adaptive Small-Scale Target Localization.

The prevailing approach for three-dimensional (3D) medical image segmentation is to use convolutional networks. Recently, deep learning methods have achieved human-level performance in several important applied problems, such as volumetry for lung-cancer diagnosis or delineation for radiation therapy planning. However, state-of-the-art architectures, such as U-Net and DeepMedic, are computationally heavy and require workstations accelerated with graphics processing units for fast inference. However, scarce research has been conducted concerning enabling fast central processing unit computations for such networks. Our paper fills this gap. We propose a new segmentation method with a human-like technique to segment a 3D study. First, we analyze the image at a small scale to identify areas of interest and then process only relevant feature-map patches. Our method not only reduces the inference time from 10 min to 15 s but also preserves state-of-the-art segmentation quality, as we illustrate in the set of experiments with two large datasets.

Braincase anatomy of extant Crocodylia, with new insights into the development and evolution of the neurocranium in crocodylomorphs.

Present-day crocodylians exhibit a remarkably akinetic skull with a highly modified braincase. We present a comprehensive description of the neurocranial osteology of extant crocodylians, with notes on the development of individual skeletal elements and a discussion of the terminology used for this project. The quadrate is rigidly fixed by multiple contacts with most braincase elements. The parabasisphenoid is sutured to the pterygoids (palate) and the quadrate (suspensorium); as a result, the basipterygoid joint is completely immobilized. The prootic is reduced and externally concealed by the quadrate. It has a verticalized buttress that participates in the canal for the temporal vasculature. The ventrolateral processes of the otoccipitals completely cover the posteroventral region of the braincase, enclose the occipital nerves and blood vessels in narrow bony canals and also provide additional sutural contacts between the braincase elements and further consolidate the posterior portion of the crocodylian skull. The otic capsule of crocodylians has a characteristic cochlear prominence that corresponds to the lateral route of the perilymphatic sac. Complex internal structures of the otoccipital (extracapsular buttress) additionally arrange the neurovascular structures of the periotic space of the cranium. Most of the braincase elements of crocodylians are excavated by the paratympanic pneumatic sinuses. The braincase in various extant crocodylians has an overall similar structure with some consistent variation between taxa. Several newly observed features of the braincase are present in Gavialis gangeticus and extant members of Crocodylidae to the exclusion of alligatorids: the reduced exposure of the prootic buttress on the floor of the temporal canal, the sagittal nuchal crest of the supraoccipital projecting posteriorly beyond the postoccipital processes and the reduced paratympanic pneumaticity. The most distinctive features of the crocodylian braincase (fixed quadrate and basipterygoid joint, consolidated occiput) evolved relatively rapidly at the base of Crocodylomorpha and accompanied the initial diversification of this clade during the Late Triassic and Early Jurassic. We hypothesize that profound rearrangements in the individual development of the braincases of basal crocodylomorphs underlie these rapid evolutionary modifications. These rearrangements are likely reflected in the embryonic development of extant crocodylians and include the involvement of neomorphic dermal anlagen in different portions of the developing chondrocranium, the extensive ossification of the palatoquadrate cartilage as a single expanded quadrate and the anteromedial inclination of the quadrate.

A phantom study to optimise the automatic tube current modulation for chest CT in COVID-19.

On March 11, 2020, the World Health Organization declared the coronavirus disease 2019 (COVID-19) pandemic. The expert organisations recommend more cautious use of thoracic computed tomography (CT), opting for low-dose protocols. We aimed at determining a threshold value of automatic tube current modulation noise index below which there is a chance to miss an onset of ground-glass opacities (GGO) in COVID-19. A team of radiologists and medical physicists performed 25 phantom CT studies using different automatic tube current modulation settings (SUREExposure3D technology). We then conducted a retrospective evaluation of the chest CT images from 22 patients with COVID-19 and calculated the density difference between the GGO and unaffected tissue. Finally, the results were matched to the phantom study results to determine the minimum noise index threshold value. The minimum density difference at the onset of COVID-19 was 252 HU (p < 0.001). This was found to correspond to the SUREExposure 3D noise index of 36. We established the noise index threshold of 36 for the Canon scanner without iterative reconstructions, allowing for a decrease in the dose-length product by 80%. The proposed protocol needs to be validated in a prospective study.

CT-Based COVID-19 triage: Deep multitask learning improves joint identification and severity quantification.

The current COVID-19 pandemic overloads healthcare systems, including radiology departments. Though several deep learning approaches were developed to assist in CT analysis, nobody considered study triage directly as a computer science problem. We describe two basic setups: Identification of COVID-19 to prioritize studies of potentially infected patients to isolate them as early as possible; Severity quantification to highlight patients with severe COVID-19, thus direct them to a hospital or provide emergency medical care. We formalize these tasks as binary classification and estimation of affected lung percentage. Though similar problems were well-studied separately, we show that existing methods could provide reasonable quality only for one of these setups. We employ a multitask approach to consolidate both triage approaches and propose a convolutional neural network to leverage all available labels within a single model. In contrast with the related multitask approaches, we show the benefit from applying the classification layers to the most spatially detailed feature map at the upper part of U-Net instead of the less detailed latent representation at the bottom. We train our model on approximately 1500 publicly available CT studies and test it on the holdout dataset that consists of 123 chest CT studies of patients drawn from the same healthcare system, specifically 32 COVID-19 and 30 bacterial pneumonia cases, 30 cases with cancerous nodules, and 31 healthy controls. The proposed multitask model outperforms the other approaches and achieves ROC AUC scores of 0.87±0.01 vs. bacterial pneumonia, 0.93±0.01 vs. cancerous nodules, and 0.97±0.01 vs. healthy controls in Identification of COVID-19, and achieves 0.97±0.01 Spearman Correlation in Severity quantification. We have released our code and shared the annotated lesions masks for 32 CT images of patients with COVID-19 from the test dataset.

Differences in Temporal Volume between Males and Females and the Influence of Age and BMI: A Cross-Sectional CT-Imaging Study.

BACKGROUND: The temple has been identified as one of the most compelling facial regions in which to seek aesthetic improvement-both locally and in the entire face-when injecting soft tissue fillers. OBJECTIVE: The objective of this study is to identify influences of age, gender, and body mass index (BMI) on temporal parameters to better understand clinical observations and to identify optimal treatment strategies for treating temporal hollowing. METHODS: The sample consisted of 28 male and 30 female individuals with a median age of 53 (34) years and a median BMI of 27.00 (6.94) kg/m2. The surface area of temporal skin, the surface area of temporal bones, and the temporal soft tissue volume were measured utilizing postprocessed computed tomography (CT) images via the Hausdorff minimal distance algorithm. Differences between the investigated participants related to age, BMI, and gender were calculated. RESULTS: Median skin surface area was greater in males compared with females 5,100.5 (708) mm2 versus 4,208.5 (893) mm2 (p < 0.001) as was the median bone surface area 5,329 (690) mm2 versus 4,477 (888) mm2 (p < 0.001). Males had on average 11.04 mL greater temporal soft tissue volume compared with age and BMI-matched females with p < 0.001. Comparing the volume between premenopausal versus postmenopausal females, the median temporal soft tissue volume was 46.63 mL (11.94) versus 40.32 mL (5.69) (p = 0.014). CONCLUSION: The results of this cross-sectional CT imaging study confirmed previous clinical and anatomical observations and added numerical evidence to those observations for a better clinical integration of the data.

Three-Dimensional Description of the Angular Artery in the Nasolabial Fold.

BACKGROUND: Due to its arterial vasculature, the nasolabial sulcus is one of the most challenging facial regions to treat when trying to ameliorate the signs of facial aging. OBJECTIVES: The aim of the present study was to provide data on the 3-dimensional course of the angular artery within the nasolabial sulcus in relation to age, gender, and body mass index to increase safety during minimally invasive treatments. METHODS: Thee hundred nasolabial sulci from 75 males and 75 females of Russian Caucasian ethnic background (mean [standard deviation] age, 45.7 [18.7] years; mean body mass index, 25.14 [4.9] kg/m2) were analyzed. Bilateral multiplanar measurements were based on contrast-enhanced computed tomography cranial scans. RESULTS: Up to 3 arteries could be identified within the nasolabial sulcus: ~90% contained 1 arterial trunk, ~9% had 2 trunks, and ~1% had 3 trunks; females had more arteries than men. The artery is located at mean depths of 21.6 mm at the oral commissure and 8.9 mm at the nasal ala. The angular artery was lateral to the nasolabial sulcus in 100% of cases; the smallest distance between the artery and the nasolabial sulcus was at the oral commissure (11.91 [7.9] mm) and the greatest was at the nasal ala (13.73 [3.9] mm). CONCLUSIONS: In contrast to current concepts, the angular artery is not located strictly subdermal to the nasolabial sulcus but at a variable depth, and in 100% of the investigated cases lateral to the nasolabial sulcus. With increasing age, the depth and lateral distance between arteries and sulci reduces significantly, underscoring the need for special caution when injecting this site.

The Course of the Angular Artery in the Midface: Implications for Surgical and Minimally Invasive Procedures.

BACKGROUND: Previous anatomic studies have provided valuable information on the 2-dimensional course of the angular segment of the facial artery in the midface and its arterial connections. The third dimension (ie, the depth of the artery) is less well characterized. OBJECTIVES: The objective of the present study was to describe the 3-dimensional pathway of the angular segment of the facial artery and its relationship to the muscles of facial expression. METHODS: The bilateral location and the depth of the midfacial segment of the facial artery was measured utilizing multi-planar computed tomographic image analyses obtained from contrast agent-enhanced cranial computed tomographic scans of 156 Caucasians aged a of 45.19 ± 18.7 years and with a mean body mass index of 25.05 ± 4.9 kg/m2. RESULTS: At the nasal ala, the mean depth of the main arterial trunk was 13.7 ± 3.7 mm (range, 2.7-25.0 mm), whereas at the medial canthus it was 1.02 ± 0.62 mm (range, 1.0-3.0 mm). This was reflected by the arteries' relationship to the midfacial muscles: at the nasal ala superficial to levator anguli oris in 62.0% but deep to the levator labii superioris alaeque nasi in 53.6%; at the medial canthus superficial to the levator labii superioris alaeque nasi in 83.1% and superficial to the orbicularis oculi in 82.7%. CONCLUSIONS: The results presented herein confirm the high variability in the course of the angular segment of the facial artery. Various arterial pathways have been identified providing evidence that, in the midface, there is no guaranteed safe location for minimally invasive procedures.

Age and Gender Differences of the Frontal Bone: A Computed Tomographic (CT)-Based Study.

BACKGROUND: Age-related changes of the frontal bone in both males and females have received limited attention, although understanding these changes is crucial to developing the best surgical and nonsurgical treatment plans for this area. OBJECTIVES: To investigate age-related and gender-related changes of the forehead. METHODS: Cranial computed tomographic images from 157 Caucasian individuals were investigated (10 males and 10 females from each of the following decades: 20-29 years, 30-39 years, 40-49 years, 50-59 years, 60-69 years, 70-79 years, 80-89 years, and of 8 males and 9 females aged 90-98 years). Frontal bone thickness and forehead distance measurements were carried out to analyze age and gender differences. RESULTS: With increasing age, the size of a male forehead reduces until no significant differences to a female forehead is present at old age (P = 0.307). The thickness of the frontal bone of the lower forehead (≤4 cm cranial to the nasal root) increased slightly in both genders with increasing age. In the upper forehead (≥4 cm cranial to the nasal root), frontal bone thickness decreased significantly (P = 0.002) in males but showed no statistically significant change in thickness in females (P = 0.165). CONCLUSIONS: The shape of the frontal bone varies in young individuals of different genders and undergoes complex changes with age because of bone remodeling. Understanding these bony changes, in addition to those in the soft tissues, helps physicians choose the best surgical and nonsurgical treatment options for the forehead.

The Relationship between Bone Remodeling and the Clockwise Rotation of the Facial Skeleton: A Computed Tomographic Imaging-Based Evaluation.

BACKGROUND: Information on the onset and gender differences of midfacial skeletal changes, including the complete understanding of the theory behind the clockwise rotational theory, remains elusive. METHODS: One hundred fifty-seven Caucasian individuals (10 men and 10 women aged 20 to 29 years, 30 to 39 years, 40 to 49 years, 50 to 59 years, 60 to 69 years, 70 to 79 years, and 80 to 89 years, and eight men and nine women aged 90 to 98 years) were investigated. Multiplanar computed tomographic scans with standardized angle and distance measurements in all three anatomical axes and in alignment to the sella-nasion (horizontal) line were conducted. RESULTS: Both men and women displayed an increase in orbital floor angle (p < 0.001, maximum at 60 to 69 years), decrease in maxillary angle (p = 0.035, 40 to 49 years), increase in palate angle (p < 0.001, 50 to 59 years), increase in vomer angle (p = 0.022, 30 to 39 years), but a decrease in the pterygoid angle (p = 0.002, 80 to 89 years). Orbital width decreased (p < 0.001, 60 to 69 years), pyriform aperture width increased (p = 0.015, 60 to 69 years), and midfacial height decreased with aging (p < 0.001, 60 to 69 years). CONCLUSIONS: Age-related changes of the midfacial skeleton occurred independently of gender, but at various time points in different locations. The observed changes seem to be driven by a bone resorption center located in the posterior maxilla, rather than by a rotational movement of the facial skeleton.