Head CT Images Research Articles

Pediatric severe traumatic brain injury mortality prediction determined with machine learning-based modeling

IntroductionSevere traumatic brain injury (sTBI) is a leading cause of mortality in children. As clinical prognostication is important in guiding optimal care and decision making, our goal was to create a highly discriminative sTBI outcome prediction model for mortality. MethodsMachine learning and advanced analytics were applied to the patient admission variables obtained from a comprehensive pediatric sTBI database. Demographic and clinical data, head CT imaging abnormalities and blood biochemical data from 196 children and adolescents admitted to a tertiary pediatric intensive care unit (PICU) with sTBI were integrated using feature ranking by way of a forest of randomized decision trees, and a model was generated from a reduced number of admission variables with maximal ability to discriminate outcome. ResultsIn total, 36 admission variables were analyzed using feature ranking with variable weighting to determine their predictive importance for mortality following sTBI. Reduction analysis utilizing Borata feature selection resulted in a parsimonious six-variable model with a mortality classification accuracy of 82%. The final admission variables that predicted mortality were: partial thromboplastin time (22%); motor Glasgow Coma Scale (21%); serum glucose (16%); fixed pupil(s) (16%); platelet count (13%) and creatinine (12%). Using only these six admission variables, a t-distributed stochastic nearest neighbor embedding algorithm plot demonstrated visual separation of sTBI patients that lived or died, with high mortality predictive ability of this model on the validation dataset (AUC = 0.90) which was confirmed with a conventional area-under-the-curve statistical approach on the total dataset (AUC = 0.91; P < 0.001). ConclusionsMachine learning-based modeling identified the most clinically important prognostic factors resulting in a pragmatic, high performing prognostic tool for pediatric sTBI with excellent discriminative ability to predict mortality risk with 82% classification accuracy (AUC = 0.90). After external multicenter validation, our prognostic model might help to guide treatment decisions, aggressiveness of therapy and prepare family members and caregivers for timely end-of-life discussions and decision making. Level of evidenceIII; Prognostic.

Contemporary national trends and disparities for head CT use in emergency department settings: Insights from National Hospital Ambulatory Medical Care Survey (NHAMCS) 2007–2017

BackgroundThe exponential growth in CT utilization in emergency department (ED) until 2008 raised concerns regarding cost and radiation exposure. Head CT was one of the commonest studies. This led to mitigating efforts such as appropriate use guidelines, policy and payment reforms. The impact of these efforts is not fully understood. In addition, disparities in outcomes of acute conditions presenting to the ED is well known however recent trends in imaging utilization patterns and disparities are not well understood. In this study, we describe nationwide trends and disparities associated with head CT in ED settings between 2007 and 2014. MethodsWe analyzed 2007–2017 National Hospital Ambulatory Medical Care Survey (NHAMCS) with the primary goal to assess the rate and patterns of head CT imaging in ED. ResultsThere were an estimated 117 million in 2007 and 139 million ED visits in 2017. There was a 4% increase in the any CT use in 2017 compared to 2007. No significant change in head CT utilization rate was seen. The 2007 head CT rate was 6.7% (95% CI: 6.1–7.3) compared to 7.7% (95% CI: 6.8–8.6) in 2017. Trauma, Headache and Dizziness are the top three indications for head CT use in the ED respectively. On adjusted analyses, significantly higher head CT utilization was seen in elderly, (age>65 yrs) and significantly lower utilization rate was seen in Non-Hispanic Black and Medicaid patients, and patients in rural locations. ConclusionsPreviously reported exponential growth of CT use in ED is no longer seen. In particular, there was no significant change in ED head CT use between 2007 and 2017. Headache and Dizziness remain commonly used indications despite limited utility in most clinical scenarios, indicating continued need for appropriate use of imaging. There is significantly lower CT utilization in Non-Hispanic Black, Medicaid patients and those in rural locations, suggesting disparities in diagnostic work-up in marginalized and rural populations. This underscores the need for standardizing care regardless of race, insurance status and location.

Adaptation of Deep Learning Image Reconstruction for Pediatric Head CT: A Focus on the Image Quality.

To assess the effect of deep learning image reconstruction (DLIR) for head CT in pediatric patients. We collected 126 pediatric head CT images, which were reconstructed using filtered back projection, iterative reconstruction using adaptive statistical iterative reconstruction (ASiR)-V, and all three levels of DLIR (TrueFidelity; GE Healthcare). Each image set group was divided into four subgroups according to the patients' ages. Clinical and dose-related data were reviewed. Quantitative parameters, including the signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR), and qualitative parameters, including noise, gray matter-white matter (GM-WM) differentiation, sharpness, artifact, acceptability, and unfamiliar texture change were evaluated and compared. The SNR and CNR of each level in each age group increased among strength levels of DLIR. High-level DLIR showed a significantly improved SNR and CNR (p < 0.05). Sequential reduction of noise, improvement of GM-WM differentiation, and improvement of sharpness was noted among strength levels of DLIR. Those of high-level DLIR showed a similar value as that with ASiR-V. Artifact and acceptability did not show a significant difference among the adapted levels of DLIR. Adaptation of high-level DLIR for the pediatric head CT can significantly reduce image noise. Modification is needed while processing artifacts.

P167 - Automatic head CT image quality quantification with deep learning: phantom study

P167 - Automatic head CT image quality quantification with deep learning: phantom study

Retrospective study of deep learning to reduce noise in non-contrast head CT images.

Presented herein is a novel CT denoising method uses a skip residual encoder-decoder framework with group convolutions and a novel loss function to improve the subjective and objective image quality for improved disease detection in patients with acute ischemic stroke (AIS). In this retrospective study, confirmed AIS patients with full-dose NCCT head scans were randomly selected from a stroke registry between 2016 and 2020. 325 patients (67±15 years, 176 men) were included. 18 patients each with 4-7 NCCTs performed within 5-day timeframe (83 total scans) were used for model training; 307 patients each with 1-4 NCCTs performed within 5-day timeframe (380 total scans) were used for hold-out testing. In the training group, a mean CT was created from the patient's co-registered scans for each input CT to train a rotation-reflection equivariant U-Net with skip and residual connections, as well as a group convolutional neural network (SRED-GCNN) using a custom loss function to remove image noise. Denoising performance was compared to the standard Block-matching and 3D filtering (BM3D) method and RED-CNN quantitatively and visually. Signal-to-noise ratio (SNR) and contrast-to-noise (CNR) were measured in manually drawn regions-of-interest in grey matter (GM), white matter (WM) and deep grey matter (DG). Visual comparison and impact on spatial resolution were assessed through phantom images. SRED-GCNN reduced the original CT image noise significantly better than BM3D, with SNR improvements in GM, WM, and DG by 2.47x, 2.83x, and 2.64x respectively and CNR improvements in DG/WM and GM/WM by 2.30x and 2.16x respectively. Compared to the proposed SRED-GCNN, RED-CNN reduces noise effectively though the results are visibly blurred. Scans denoised by the SRED-GCNN are shown to be visually clearer with preserved anatomy. The proposed SRED-GCNN model significantly reduces image noise and improves signal-to-noise and contrast-to-noise ratios in 380 unseen head NCCT cases.

Perioperative prophylactic middle meningeal artery embolization for chronic subdural hematoma: a series of 44 cases.

Chronic subdural hematoma (cSDH) is a common and challenging pathology to treat due to both the historically high recurrence rate following surgical evacuation and the medical comorbidities inherent in the aging patient population that it primarily affects. Middle meningeal artery (MMA) embolization has shown promise in the treatment of cSDHs, most convincingly to avoid surgical evacuation in relatively asymptomatic patients. Symptomatic patients requiring surgical evacuation may also benefit from perioperative MMA embolization to prevent cSDH recurrence. The goal of this study was to determine the utility of perioperative MMA embolization for symptomatic cSDH requiring surgical evacuation and to assess if there is a decrease in the cSDH recurrence rate compared to historical recurrence rates following surgical evacuation alone. Symptomatic cSDHs were evacuated using a subdural evacuating port system (SEPS) with 5-mm twist-drill craniostomy in an intensive care unit or by performing a craniotomy in the operating room, using either a small (silver dollar, < 4 cm) or large (≥ 4 cm) craniotomy. MMA embolization was performed perioperatively using angiography, selective catheterization of the MMA, and infusion of polyvinyl particles. Outcomes were assessed clinically and radiographically with interval head CT imaging. There were 44 symptomatic cSDHs in 41 patients, with 3 patients presenting with bilateral symptomatic cSDH. All cSDHs were evacuated using an SEPS (n = 18), a silver-dollar craniotomy (n = 16), or a large craniotomy (n = 10). Prophylactic MMA embolization was performed successfully in all cSDHs soon after surgical evacuation. There were no deaths and no procedural complications. There was an overall reduction of greater than 50% or resolution of cSDH in 40/44 (90.9%) cases, regardless of the evacuation procedure used. Of the 44 prophylactic cases, there were 2 (4.5%) cases of cSDH recurrence that required repeat surgical evacuation at the 1-year follow-up. These 2 cSDHs were initially evacuated using an SEPS and subsequently required a craniotomy, thereby representing an overall 4.5% recurrence rate of treated cSDH requiring repeat evacuation. Most notably, of the 26 patients who underwent surgical evacuation with a craniotomy followed by MMA embolization, none had cSDH recurrence requiring repeat intervention. Perioperative prophylactic MMA embolization in the setting of surgical evacuation, via either craniotomy or SEPS, may help to lower the recurrence rate of cSDH.

Hybrid active shape and deep learning method for the accurate and robust segmentation of the intracochlear anatomy in clinical head CT and CBCT images.

Purpose: Robust and accurate segmentation methods for the intracochlear anatomy (ICA) are a critical step in the image-guided cochlear implant programming process. We have proposed an active shape model (ASM)-based method and a deep learning (DL)-based method for this task, and we have observed that the DL method tends to be more accurate than the ASM method while the ASM method tends to be more robust. Approach: We propose a DL-based U-Net-like architecture that incorporates ASM segmentation into the network. A quantitative analysis is performed on a dataset that consists of 11 cochlea specimens for which a segmentation ground truth is available. To qualitatively evaluate the robustness of the method, an experienced expert is asked to visually inspect and grade the segmentation results on a clinical dataset made of 138 image volumes acquired with conventional CT scanners and of 39 image volumes acquired with cone beam CT (CBCT) scanners. Finally, we compare training the network (1)first with the ASM results, and then fine-tuning it with the ground truth segmentation and (2)directly with the specimens with ground truth segmentation. Results: Quantitative and qualitative results show that the proposed method increases substantially the robustness of the DL method while having only a minor detrimental effect (though not significant) on its accuracy. Expert evaluation of the clinical dataset shows that by incorporating the ASM segmentation into the DL network, the proportion of good segmentation cases increases from 60/177 to 119/177 when training only with the specimens and increases from 129/177 to 151/177 when pretraining with the ASM results. Conclusions: A hybrid ASM and DL-based segmentation method is proposed to segment the ICA in CT and CBCT images. Our results show that combining DL and ASM methods leads to a solution that is both robust and accurate.

Optimising endotracheal length in adult cats: a retrospective CT study.

The aim of this study was to determine the maximal endotracheal insertion length by measuring the larynx to carina (L-C) distance by means of CT. An additional objective was to establish certain anatomical landmarks to optimise the process of endotracheal intubation (ETI). Head, neck and thoracic CT images from adult cats at a single referral hospital between 2013 and 2020 were retrospectively evaluated. After standardising and identifying key markers (larynx, carina and first rib) the L-C, larynx to first rib (L-1R) and first rib to carina (1R-C) distances were measured. Forty-five adult cats were enrolled in the study, from which a total of nine different breeds were identified. The L-C distance was 14.3 ± 1.1 cm. This was longer in male (14.7 ± 1.1 cm) than in female cats (13.5 ± 0.7 cm). The first rib (1R) was 8.8 ± 0.7 cm from the larynx and the mean 1R-C distance was 5.4 ± 0.7 cm. The carina was found within the fifth intercostal space in 93.3% (n = 42) of the cats. The process of ETI in adult cats may be guided by using the L-C and L-1R distance for a maximal and optimal endotracheal tube introduction, respectively. In addition, the maximal insertion length may be guided by estimating the position of the carina parallel to the fifth intercostal space.

675 ICA Ligation and Malignant MCA Syndrome Secondary to A Gunshot Wound: A Case Presentation

Abstract Penetrating carotid injuries occur infrequently in the UK. Without surgical intervention, mortality rates are close to 100%. Whilst vessel repair is the optimal surgical choice, zone III neck injuries or haemodynamic instability in particular often require vessel ligation, which carries higher mortality rates, increased risk of stroke and poorer outcomes overall. We present a young patient who sustained a gunshot wound to the neck. Emergency exploration revealed a penetrating injury to zone III of the left side of the neck and a complete transection of the left internal carotid artery. Due to haemodynamic instability, damage control surgery was mandated, and the internal carotid artery was ligated. The patient was admitted to ITU but had a severe acute neurological deterioration post-operation. CT head imaging revealed a large left frontal-parietal infarct with rightward midline shift, in keeping with malignant MCA syndrome. He underwent an emergency decompressive craniectomy, with resolution of the midline shift seen on subsequent imaging. At 3 months, the patient walked independently out of hospital, with a reduction in his modified Rankin Scale score from 5 to 2 following intensive therapies input. Whilst rare, the importance of timely recognition and intervention of an infrequent complication of carotid ligation is paramount. Decompressive craniectomy is the mainstay of treatment for malignant MCA syndrome, yet outcomes remain poor in all ages, highlighting the patient’s noteworthy recovery. Though emerging evidence suggests novel endovascular repair techniques may provide favourable outcomes and reduced complications in carotid injuries, open surgery remains the treatment of choice.

Does cerebral oximetry always measure brain tissue oxygen saturation? An anatomical study utilizing computed tomography.

Background and Aims:to quantify the scalp-cortex distance and determine its variation among patients. We hypothesized that in a significant number of patients, this distance is greater than the maximum penetration depth of current cerebral oximeters.Material and Methods:A retrospective anatomic study using transverse head CT images selected randomly from 102 patients over the age of 18 years without brain swelling, intracranial mass effect, or brain hemorrhage. Scalp-cortex distances were determined at two separate locations along the craniocaudal axis; most cephalad to the frontal sinus (I0) and also 2 cm cephalad to that location (I2). Multiple measurements were obtained bilaterally at 1, 3, 5, 7, and 9 cm from midline.Results:The average scalp-cortex distance was 14.3 mm and 15 mm at I0 and I2 respectively. Distances varied more in I2 than in I0; from the measurements, 12.8% vs. 6.8% were over 20 mm, 4.4% vs. 2.2% over 25 mm, 1.1% vs. 0.6% over 35 mm and 0.6% vs. none over 40 mm at I2 and I0, respectively. 1.5% of the measurements at I2 were over 30 mm.Conclusion:Cerebral oximetry manufacturers all claim to measure cerebral tissue up to a depth of 20-25 mm; 20 mm with the EQUANOX and INVOS compared with 25 mm with the FORE-SIGHT. Scalp-cortex distance is within 25 mm in more than 95% of patients. However, even with the probe placed as per the manufacturer’s recommendations, in a small but significant subset of patients, this distance is greater than the maximum penetration depth of current cerebral oximeters and hence may not reflect actual brain tissue oxygen saturation.

Three‐Dimensional Distribution of Bone‐Resorption Lesions in Osteonecrosis of the Femoral Head Based on the Three‐Pillar Classification

ObjectiveTo investigate three‐dimensional distribution of bone‐resorptive lesions based on the three‐pillar classification and its effect on the disease progression of osteonecrosis of the femoral head (ONFH).MethodsA total of 194 femoral head CT images from 117 patients diagnosed with ARCO stage II and III ONFH were retrospectively reviewed from April 2014 to February 2019. Three‐dimensional structures of the femoral head and the bone‐resorptive lesions were reconstructed. Using the three‐pillar classification and coronal plane of the femoral head, we divided each femoral head into six regions to observe the location characteristics of bone‐resorption lesions, and explore the destruction of different areas of the femoral head by the bone‐resorptive lesions. Then the hips were divided into two groups based on whether they contained bone‐resorption lesions and compared the difference of stage II and stage III between the two groups.ResultsThe regional distribution revealed 39 (27.27%), 55 (38.46%), six (4.20%), 23 (16.08%), 17 (11.89%) and three (2.10%) bone‐resorptive lesions in regions I, II, III, IV, V and VI respectively. The lateral pillar, AL (I + IV), contained 44.76% of the lesions, central pillar, C (II + V), 48.95%, and medial pillar, M (III + VI), 6.29%. Moreover, there were 81.82% bone‐resorption lesions in anterolateral pillar, AL (I + II + IV), and 18.18% in posteromedial pillar, PM (III + V + VI). In all ONFH hips, the lateral pillar of 81(88.04%) femoral heads were affected, the central pillar of 84 (91.30%) femoral heads were affected, and the medical pillar of 29 (31.52%) femoral heads were affected. The ratio of ARCO stage III in the group with bone‐resorption lesions was significantly higher than that of the group without bone‐resorption lesions (76.09% vs 30.39%, P < 0.001).ConclusionsThis study demonstrated that the bone‐resorption lesions are mainly distributed in the lateral and central pillar of the femoral head, and the two pillars of the femoral head are usually involved by bone‐resorption lesions. Furthermore, the ratio of ARCO stage III in the group with bone‐resorption lesions was significantly higher than that of the group without bone‐resorption lesions, suggesting that the bone‐resorption lesions might accelerate the progression of ONFH.

Deep learning from MRI-derived labels enables automatic brain tissue classification on human brain CT

Automatic methods for feature extraction, volumetry, and morphometric analysis in clinical neuroscience typically operate on images obtained with magnetic resonance (MR) imaging equipment. Although CT scans are less expensive to acquire and more widely available than MR scans, their application is currently limited to the visual assessment of brain integrity and the exclusion of co-pathologies. CT has rarely been used for tissue classification because the contrast between grey matter and white matter was considered insufficient. In this study, we propose an automatic method for segmenting grey matter (GM), white matter (WM), cerebrospinal fluid (CSF), and intracranial volume (ICV) from head CT images. A U-Net deep learning model was trained and validated on CT images with MRI-derived segmentation labels. We used data from 744 participants of the Gothenburg H70 Birth Cohort Studies for whom CT and T1-weighted MR images had been acquired on the same day. Our proposed model predicted brain tissue classes accurately from unseen CT images (Dice coefficients of 0.79, 0.82, 0.75, 0.93 and 0.98 for GM, WM, CSF, brain volume and ICV, respectively). To contextualize these results, we generated benchmarks based on established MR-based methods and intentional image degradation. Our findings demonstrate that CT-derived segmentations can be used to delineate and quantify brain tissues, opening new possibilities for the use of CT in clinical practice and research.

An optimal deep learning framework for multi-type hemorrhagic lesions detection and quantification in head CT images for traumatic brain injury.

Traumatic Brain Injury (TBI) could lead to intracranial hemorrhage (ICH), which has now been identified as a major cause of death after trauma if it is not adequately diagnosed and properly treated within the first 24 hours. CT examination is widely preferred for urgent ICH diagnosis, which enables the fast identification and detection of ICH regions. However, the use of it requires the clinical interpretation by experts to identify the subtypes of ICH. Besides, it is unable to provide the details needed to conduct quantitative assessment, such as the volume and thickness of hemorrhagic lesions, which may have prognostic importance to the decision-making on emergency treatment. In this paper, an optimal deep learning framework is proposed to assist the quantitative assessment for ICH diagnosis and the accurate detection of different subtypes of ICH through head CT scan. Firstly, the format of raw input data is converted from 3D DICOM to NIfTI. Secondly, a pre-trained multi-class semantic segmentation model is applied to each slice of CT images, so as to obtain a precise 3D mask of the whole ICH region. Thirdly, a fine-tuned classification neural network is employed to extract the key features from the raw input data and identify the subtypes of ICH. Finally, a quantitative assessment algorithm is adopted to automatically measure both thickness and volume via the 3D shape mask combined with the output probabilities of the classification network. The results of our extensive experiments demonstrate the effectiveness of the proposed framework where the average accuracy of 96.21 percent is achieved for three types of hemorrhage. The capability of our optimal classification model to distinguish between different types of lesion plays a significant role in reducing the false-positive rate in the existing work. Furthermore, the results suggest that our automatic quantitative assessment algorithm is effective in providing clinically relevant quantification in terms of volume and thickness. It is more important than the qualitative assessment conducted through visual inspection to the decision-making on emergency surgical treatment.

Dental radiographic identification using ante-mortem CT, cone-beam CT, and MRI head and neck assessments

Radiographs of the maxilla and mandible obtained for examining oral diseases include various anatomical features. Similarly, radiographs obtained in head and neck assessments may also include dental findings. In this study, we aimed to identify individuals using head and neck CT and MR imaging to obtain identifying information, including dental findings. We examined morphological similarities by superimposition and compared dental findings and tooth arrangement using head and neck images. In addition, 3D superimposition by ante-mortem (AM) cone-beam (CB) CT and post-mortem (PM) CT was also performed. In the evaluation of the superimposition of PM-CT, AM-CBCT and AM-MR images, all cases matched in 3D. In comparisons and collations of AM and PM images, the dental findings were similar. The findings suggest that dental findings can be assessed even on head and neck images, even if there are no AM dental images. Moreover, our attempts to compare and collate dental 3D images have indicated a new direction for future dental personal identification.

Application of a deep learning image reconstruction (DLIR) algorithm in head CT imaging for children to improve image quality and lesion detection

BackgroundTo evaluate the performance of a Deep Learning Image Reconstruction (DLIR) algorithm in pediatric head CT for improving image quality and lesion detection with 0.625 mm thin-slice images.MethodsLow-dose axial head CT scans of 50 children with 120 kV, 0.8 s rotation and age-dependent 150–220 mA tube current were selected. Images were reconstructed at 5 mm and 0.625 mm slice thickness using Filtered back projection (FBP), Adaptive statistical iterative reconstruction-v at 50% strength (50%ASIR-V) (as reference standard), 100%ASIR-V and DLIR-high (DL-H). The CT attenuation and standard deviation values of the gray and white matters in the basal ganglia were measured. The clarity of sulci/cisterns, boundary between white and gray matters, and overall image quality was subjectively evaluated. The number of lesions in each reconstruction group was counted.ResultsThe 5 mm FBP, 50%ASIR-V, 100%ASIR-V and DL-H images had a subjective score of 2.25 ± 0.44, 3.05 ± 0.23, 2.87 ± 0.39 and 3.64 ± 0.49 in a 5-point scale, respectively with DL-H having the lowest image noise of white matter at 2.00 ± 0.34 HU; For the 0.625 mm images, only DL-H images met the diagnostic requirement. The 0.625 mm DL-H images had similar image noise (3.11 ± 0.58 HU) of the white matter and overall image quality score (3.04 ± 0.33) as the 5 mm 50% ASIR-V images (3.16 ± 0.60 HU and 3.05 ± 0.23). Sixty-five lesions were recognized in 5 mm 50%ASIR-V images and 69 were detected in 0.625 mm DL-H images.ConclusionDL-H improves the head CT image quality for children compared with ASIR-V images. The 0.625 mm DL-H images improve lesion detection and produce similar image noise as the 5 mm 50%ASIR-V images, indicating a potential 85% dose reduction if current image quality and slice thickness are desired.

P-22 Artifact reduction in head and neck CT images using polyetheretherketone (PEEK), polyetherketoneketone (PEKK), polyphenylsulfone (PPSU) and polyether (PE) polymer implants in mandible reconstruction

P-22 Artifact reduction in head and neck CT images using polyetheretherketone (PEEK), polyetherketoneketone (PEKK), polyphenylsulfone (PPSU) and polyether (PE) polymer implants in mandible reconstruction

Technical Note: Dose prediction for head and neck radiotherapy using a three-dimensional dense dilated U-net architecture.

Radiation therapy treatment planning is a time-consuming and iterative manual process. Consequently, plan quality varies greatly between and within institutions. Artificial intelligence shows great promise in improving plan quality and reducing planning times. This technical note describes our participation in the American Association of Physicists in Medicine Open Knowledge-Based Planning Challenge (OpenKBP), a competition to accurately predict radiation therapy dose distributions. A three-dimensional (3D) densely connected U-Net with dilated convolutions was developed to predict 3D dose distributions given contoured CT images of head and neck patients as input. While traditional augmentation techniques such as rotations and translations were explored, it was found that training on random patches alone resulted in the greatest model performance. A custom-weighted mean squared error loss function was employed. Finally, an ensemble of best-performing networks was used to generate the final challenge predictions. Our team (SuperPod) placed second in the dose stream of the OpenKBP challenge. The average mean absolute difference between the predicted and clinical dose distributions of the testing dataset was 2.56Gy. On average, the predicted normalized target DVH metrics were within 3% of the clinical plans, and the predicted organ at risk DVH metrics were within 2Gy of the clinical plans. The developed 3D dense dilated U-Net architecture can accurately predict 3D radiotherapy dose distributions and can be used as part of a fully automated radiation therapy planning pipeline.

Head CT: Toward Making Full Use of the Information the X-Rays Give.

Although clinical head CT images are typically interpreted qualitatively, automated methods applied to routine clinical head CTs enable quantitative assessment of brain volume, brain parenchymal fraction, brain radiodensity, and brain radiomass. These metrics gain clinical meaning when viewed relative to a reference database and expressed as quantile regression values. Quantitative imaging data can aid in objective reporting and in the identification of outliers, with possible diagnostic implications. The comparison to a reference database necessitates standardization of head CT imaging parameters and protocols. Future research is needed to learn the effects of virtual monochromatic imaging on the quantitative characteristics of head CT images.

Persistent postconcussive symptoms in children and adolescents with mild traumatic brain injury receiving initial head computed tomography.

The aim of this paper was to evaluate the prevalence of postconcussive symptoms and their relation to health-related quality of life (HRQOL) in pediatric and adolescent patients with mild traumatic brain injury (mTBI) who received head CT imaging during initial assessment. Patients aged between 5 and 21 years with mTBI (Glasgow Coma Scale scores 13-15) and available Rivermead Post Concussion Questionnaire (RPQ) at 6 months of follow-up in the multicenter, prospectively collected CENTER-TBI (Collaborative European NeuroTrauma Effectiveness Research in TBI) study were included. The prevalence of postconcussive symptoms was assessed, and the occurrence of postconcussive syndrome (PSC) based on the ICD-10 criteria, was analyzed. HRQOL was compared in patients with and without PCS using the Quality of Life after Brain Injury (QOLIBRI) questionnaire. A total of 196 adolescent or pediatric mTBI patients requiring head CT imaging were included. High-energy trauma was prevalent in more than half of cases (54%), abnormalities on head CT scans were detected in 41%, and admission to the regular ward or intensive care unit was necessary in 78%. Six months postinjury, 36% of included patients had experienced at least one moderate or severe symptom on the RPQ. PCS was present in 13% of adolescents and children when considering symptoms of at least moderate severity, and those patients had significantly lower QOLIBRI total scores, indicating lower HRQOL, compared with young patients without PCS (57 vs 83 points, p < 0.001). Adolescent and pediatric mTBI patients requiring head CT imaging show signs of increased trauma severity. Postconcussive symptoms are present in up to one-third of those patients, and PCS can be diagnosed in 13% 6 months after injury. Moreover, PCS is significantly associated with decreased HRQOL.

Blind spots on CT imaging of the head: Insights from 5 years of report addenda at a single institution

BackgroundErrors of detection (“misses”) are the major source of error in radiology. There is sparse prior literature describing patterns of detection error on CT head imaging. PurposeThe objective of this study was to gain insight to areas on CT head imaging where radiologists are most likely to miss clinically relevant findings. MethodsWe performed a cross-sectional study of consecutive reports of CT imaging of the head at a single institution spanning 5/1/2013–5/1/2018 (5 years). Detection errors described in addenda were categorized according to anatomic location, type of pathology, and potential impact on management. Blind spots were defined by the most common sites of missed findings. ResultsA total of 165,943 reports for CT head imaging were obtained. Addenda were found in 1658 (~1%) of reports, of which 359 (21.7%) described errors of detection. Within the extracranial soft tissues (n = 73) the most common “misses” were at incidentally imaged parotid glands and the frontal scalp. Within osseous structures (n = 149), blind spots included the nasal and occipital bones. Vascular lesions (n = 47) which passed detection were most common at the distal MCA, carotid terminus and sigmoid sinus/jugular bulb. No predisposition was seen for anatomic subsites within the CSF space (n = 60) and brain parenchyma (n = 65). ConclusionsConsistent patterns of blind spots are revealed. Radiologic teaching and search patterns to account for these sites of error may accelerate trainee competence and improve accuracy in the practice of radiology.