AI Efficacy as a Function of Trainee Interpreter Proficiency: Lessons from a Randomized Controlled Trial.

BACKGROUND AND PURPOSE: Recently, AI tools have been deployed with increasing speed in educational and clinical settings. However, the use of AI by trainees across different levels of experience has not been well studied. This study investigates the impact of AI assistance on diagnostic accuracy for intracranial hemorrhage (ICH) and large vessel occlusion (LVO) by medical students (MS) and resident trainees (RT). MATERIALS AND METHODS: This prospective study was conducted between March 2023 and October 2023. MS and RT were asked to identify ICH and LVO in 100 non-contrast head CTs and 100 head CTAs, respectively. One group received diagnostic aid simulating AI for ICH only (n = 26), the other for LVO only (n = 28). Primary outcomes included accuracy, sensitivity, and specificity for ICH / LVO detection without and with aid. Study interpretation time was a secondary outcome. Individual responses were pooled and analyzed with chi-square; differences in continuous variables were assessed with ANOVA. RESULTS: 48 participants completed the study, generating 10,779 ICH or LVO interpretations. With diagnostic aid, MS accuracy improved 11.0 points (P < .001) and RT accuracy showed no significant change. ICH interpretation time increased with diagnostic aid for both groups (P < .001) while LVO interpretation time decreased for MS (P < .001). Despite worse performance in detection of the smallest vs. the largest hemorrhages at baseline, MS were not more likely to accept a true positive AI result for these more difficult tasks. Both groups were considerably less accurate when disagreeing with the AI or when supplied with an incorrect AI result. CONCLUSIONS: This study demonstrated greater improvement in diagnostic accuracy with AI for MS compared to RT. However, MS were less likely than RT to overrule incorrect AI interpretations and were less accurate, even with diagnostic aid, than the AI was by itself. ABBREVIATIONS: ICH = intracranial hemorrhage; LVO = large vessel occlusion; MS = medical students; RT = resident trainees.

Posttreatment Surveillance Imaging After Radiation for Head and Neck Cancer.

Importance: The role of surveillance imaging after treatment for head and neck cancer is controversial and evidence to support decision-making is limited. Objective: To determine the use of surveillance imaging in asymptomatic patients with head and neck cancer in remission after completion of chemoradiation. Design, Setting, and Participants: This was a retrospective, comparative effectiveness research review of adult patients who had achieved a complete metabolic response to initial treatment for head and neck cancer as defined by having an unequivocally negative positron emission tomography (PET) scan using the PET response criteria in solid tumors (PERCIST) scale within the first 6 months of completing therapy. The medical records of 501 consecutive patients who completed definitive radiation therapy (with or without chemotherapy) for newly diagnosed squamous cell carcinoma of the head and neck between January 2014 and June 2022 were reviewed. Exposure: Surveillance imaging was defined as the acquisition of a PET with computed tomography (CT), magnetic resonance imaging (MRI), or CT of the head and neck region in the absence of any clinically suspicious symptoms and/or examination findings. For remaining patients, subsequent surveillance after the achievement of a complete metabolic response to initial therapy was performed on an observational basis in the setting of routine follow-up using history-taking and physical examination, including endoscopy. This expectant approach led to imaging only in the presence of clinically suspicious symptoms and/or physical examination findings. Main Outcome and Measures: Local-regional control, overall survival, and progression-free survival based on assignment to either the surveillance imaging or expectant management cohort. Results: This study included 340 patients (mean [SD] age, 59 [10] years; 201 males [59%]; 88 Latino patients [26%]; 145 White patients [43%]) who achieved a complete metabolic response during this period. There was no difference in 3-year local-regional control, overall survival, progression-free survival, or freedom from distant metastasis between patients treated with surveillance imaging vs those treated expectantly. Conclusions and Relevance: In this comparative effectiveness research, imaging-based surveillance failed to improve outcomes compared with expectant management for patients who were seemingly in remission after completion of primary radiation therapy for head and neck cancer.

Impact of an automated large vessel occlusion detection tool on clinical workflow and patient outcomes.

Purpose: Automated large vessel occlusion (LVO) tools allow for prompt identification of positive LVO cases, but little is known about their role in acute stroke triage when implemented in a real-world setting. The purpose of this study was to evaluate the automated LVO detection tool's impact on acute stroke workflow and clinical outcomes. Materials and methods: Consecutive patients with a computed tomography angiography (CTA) presenting with suspected acute ischemic stroke were compared before and after the implementation of an AI tool, RAPID LVO (RAPID 4.9, iSchemaView, Menlo Park, CA). Radiology CTA report turnaround times (TAT), door-to-treatment times, and the NIH stroke scale (NIHSS) after treatment were evaluated. Results: A total of 439 cases in the pre-AI group and 321 cases in the post-AI group were included, with 62 (14.12%) and 43 (13.40%) cases, respectively, receiving acute therapies. The AI tool demonstrated a sensitivity of 0.96, a specificity of 0.85, a negative predictive value of 0.99, and a positive predictive value of 0.53. Radiology CTA report TAT significantly improved post-AI (mean 30.58 min for pre-AI vs. 22 min for post-AI, p < 0.0005), notably at the resident level (p < 0.0003) but not at higher levels of expertise. There were no differences in door-to-treatment times, but the NIHSS at discharge was improved for the pre-AI group adjusted for confounders (parameter estimate = 3.97, p < 0.01). Conclusion: Implementation of an automated LVO detection tool improved radiology TAT but did not translate to improved stroke metrics and outcomes in a real-world setting.

Deep Learning-Based Algorithm for Automatic Detection of Pulmonary Embolism in Chest CT Angiograms.

PURPOSE: Since the prompt recognition of acute pulmonary embolism (PE) and the immediate initiation of treatment can significantly reduce the risk of death, we developed a deep learning (DL)-based application aimed to automatically detect PEs on chest computed tomography angiograms (CTAs) and alert radiologists for an urgent interpretation. Convolutional neural networks (CNNs) were used to design the application. The associated algorithm used a hybrid 3D/2D UNet topology. The training phase was performed on datasets adequately distributed in terms of vendors, patient age, slice thickness, and kVp. The objective of this study was to validate the performance of the algorithm in detecting suspected PEs on CTAs. METHODS: The validation dataset included 387 anonymized real-world chest CTAs from multiple clinical sites (228 U.S. cities). The data were acquired on 41 different scanner models from five different scanner makers. The ground truth (presence or absence of PE on CTA images) was established by three independent U.S. board-certified radiologists. RESULTS: The algorithm correctly identified 170 of 186 exams positive for PE (sensitivity 91.4% [95% CI: 86.4-95.0%]) and 184 of 201 exams negative for PE (specificity 91.5% [95% CI: 86.8-95.0%]), leading to an accuracy of 91.5%. False negative cases were either chronic PEs or PEs at the limit of subsegmental arteries and close to partial volume effect artifacts. Most of the false positive findings were due to contrast agent-related fluid artifacts, pulmonary veins, and lymph nodes. CONCLUSIONS: The DL-based algorithm has a high degree of diagnostic accuracy with balanced sensitivity and specificity for the detection of PE on CTAs.

Impact of easing COVID-19 safety measures on trauma computed tomography imaging volumes.

PURPOSE: The coronavirus disease 2019 (COVID-19) pandemic has led to substantial disruptions in healthcare staffing and operations. Stay-at-home (SAH) orders and limitations in social gathering implemented in spring 2020 were followed by initial decreases in healthcare and imaging utilization. This study aims to evaluate the impact of subsequent easing of SAH on trauma volumes, demand for, and turnaround times for trauma computed tomography (CT) exams, hypothesizing that after initial decreases, trauma volumes have increased as COVID safety measures have been reduced. METHODS: Patient characteristics, CT imaging volumes, and turnaround time were analyzed for all adult activated emergency department trauma patients requiring CT imaging at a single Level-I trauma center (1/2018-2/2022) located in the sixth most populous county in the USA. Based on COVID safety measures in place in the state of California, three time periods were compared: baseline (PRE, 1/1/2018-3/19/2020), COVID safety measures (COVID, 3/20/2020-1/25/2021), and POST (1/26/2021-2/28/2022). RESULTS: There were 16,984 trauma patients across the study (PRE = 8289, COVID = 3139, POST = 5556). The average daily trauma patient volumes increased significantly in the POST period compared to the PRE and COVID periods (13.9 vs. 10.3 vs. 10.1, p < 0.001), with increases in both blunt (p < 0.001) and penetrating (p = 0.002) trauma. The average daily number of trauma CT examinations performed increased significantly in the POST period compared to the PRE and COVID periods (56.7 vs. 48.3 vs. 47.6, p < 0.001), with significant increases in average turnaround time (47 min vs. 31 and 37, p < 0.001). CONCLUSION: After initial decreases in trauma radiology volumes following stay-at-home orders, subsequent easing of safety measures has coincided with increases in trauma imaging volumes above pre-pandemic levels and longer exam turnaround times.

Point-of-Care Brain MRI: Preliminary Results from a Single-Center Retrospective Study.

Background Point-of-care (POC) MRI is a bedside imaging technology with fewer than five units in clinical use in the United States and a paucity of scientific studies on clinical applications. Purpose To evaluate the clinical and operational impacts of deploying POC MRI in emergency department (ED) and intensive care unit (ICU) patient settings for bedside neuroimaging, including the turnaround time. Materials and Methods In this preliminary retrospective study, all patients in the ED and ICU at a single academic medical center who underwent noncontrast brain MRI from January 2021 to June 2021 were investigated to determine the number of patients who underwent bedside POC MRI. Turnaround time, examination limitations, relevant findings, and potential CT and fixed MRI findings were recorded for patients who underwent POC MRI. Descriptive statistics were used to describe clinical variables. The Mann-Whitney U test was used to compare the turnaround time between POC MRI and fixed MRI examinations. Results Of 638 noncontrast brain MRI examinations, 36 POC MRI examinations were performed in 35 patients (median age, 66 years [IQR, 57-77 years]; 21 women), with one patient undergoing two POC MRI examinations. Of the 36 POC MRI examinations, 13 (36%) occurred in the ED and 23 (64%) in the ICU. There were 12 of 36 (33%) POC MRI examinations interpreted as negative, 14 of 36 (39%) with clinically significant imaging findings, and 10 of 36 (28%) deemed nondiagnostic for reasons such as patient motion. Of 23 diagnostic POC MRI examinations with comparison CT available, three (13%) demonstrated acute infarctions not apparent on CT scans. Of seven diagnostic POC MRI examinations with subsequent fixed MRI examinations, two (29%) demonstrated missed versus interval subcentimeter infarctions, while the remaining demonstrated no change. The median turnaround time of POC MRI was 3.4 hours in the ED and 5.3 hours in the ICU. Conclusion Point-of-care (POC) MRI was performed rapidly in the emergency department and intensive care unit. A few POC MRI examinations demonstrated acute infarctions not apparent at standard-of-care CT examinations. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Anzai and Moy in this issue.

Neuroanatomical Correlates Underlying the Association Between Maternal Interleukin 6 Concentration During Pregnancy and Offspring Fluid Reasoning Performance in Early Childhood.

BACKGROUND: Maternal inflammation during pregnancy can alter offspring brain development and influence risk for disorders commonly accompanied by deficits in cognitive functioning. We therefore examined associations between maternal interleukin 6 (IL-6) concentrations during pregnancy and offspring cognitive ability and concurrent magnetic resonance imaging-based measures of brain anatomy in early childhood. We further examined newborn brain anatomy in secondary analyses to consider whether effects are evident soon after birth and to increase capacity to differentiate effects of pre- versus postnatal exposures. METHODS: IL-6 concentrations were quantified in early (12.6 ± 2.8 weeks), mid (20.4 ± 1.5 weeks), and late (30.3 ± 1.3 weeks) pregnancy. Offspring nonverbal fluid intelligence (Gf) was assessed at 5.2 ± 0.6 years using a spatial reasoning task (Wechsler Preschool and Primary Scale of Intelligence-Matrix) (n = 49). T1-weighted magnetic resonance imaging scans were acquired at birth (n = 89, postmenstrual age = 42.9 ± 2.0 weeks) and in early childhood (n = 42, scan age = 5.1 ± 1.0 years). Regional cortical volumes were examined for a joint association between maternal IL-6 and offspring Gf performance. RESULTS: Average maternal IL-6 concentration during pregnancy was inversely associated with offspring Gf performance after adjusting for socioeconomic status and the quality of the caregiving and learning environment (R2 = 13%; p = .02). Early-childhood pars triangularis volume was jointly associated with maternal IL-6 and childhood Gf (pcorrected < .001). An association also was observed between maternal IL-6 and newborn pars triangularis volume (R2 = 6%; p = .02). CONCLUSIONS: These findings suggest that the origins of variation in child cognitive ability can, in part, trace back to maternal conditions during the intrauterine period of life and support the role of inflammation as an important component of this putative biological pathway.

Validation of a Deep Learning Tool in the Detection of Intracranial Hemorrhage and Large Vessel Occlusion.

Purpose: Recently developed machine-learning algorithms have demonstrated strong performance in the detection of intracranial hemorrhage (ICH) and large vessel occlusion (LVO). However, their generalizability is often limited by geographic bias of studies. The aim of this study was to validate a commercially available deep learning-based tool in the detection of both ICH and LVO across multiple hospital sites and vendors throughout the U.S. Materials and Methods: This was a retrospective and multicenter study using anonymized data from two institutions. Eight hundred fourteen non-contrast CT cases and 378 CT angiography cases were analyzed to evaluate ICH and LVO, respectively. The tool's ability to detect and quantify ICH, LVO, and their various subtypes was assessed among multiple CT vendors and hospitals across the United States. Ground truth was based off imaging interpretations from two board-certified neuroradiologists. Results: There were 255 positive and 559 negative ICH cases. Accuracy was 95.6%, sensitivity was 91.4%, and specificity was 97.5% for the ICH tool. ICH was further stratified into the following subtypes: intraparenchymal, intraventricular, epidural/subdural, and subarachnoid with true positive rates of 92.9, 100, 94.3, and 89.9%, respectively. ICH true positive rates by volume [small (<5 mL), medium (5-25 mL), and large (>25 mL)] were 71.8, 100, and 100%, respectively. There were 156 positive and 222 negative LVO cases. The LVO tool demonstrated an accuracy of 98.1%, sensitivity of 98.1%, and specificity of 98.2%. A subset of 55 randomly selected cases were also assessed for LVO detection at various sites, including the distal internal carotid artery, middle cerebral artery M1 segment, proximal middle cerebral artery M2 segment, and distal middle cerebral artery M2 segment with an accuracy of 97.0%, sensitivity of 94.3%, and specificity of 97.4%. Conclusion: Deep learning tools can be effective in the detection of both ICH and LVO across a wide variety of hospital systems. While some limitations were identified, specifically in the detection of small ICH and distal M2 occlusion, this study highlights a deep learning tool that can assist radiologists in the detection of emergent findings in a variety of practice settings.

Outcomes of Artificial Intelligence Volumetric Assessment of Kidneys and Renal Tumors for Preoperative Assessment of Nephron-Sparing Interventions.

Background: Renal-cell carcinoma is the most common kidney cancer and the 13th most common cause of cancer death worldwide. Partial nephrectomy and percutaneous ablation, increasingly utilized to treat small renal masses and preserve renal parenchyma, require precise preoperative imaging interpretation. We sought to develop and evaluate a convolutional neural network (CNN), a type of deep learning (DL) artificial intelligence (AI), to act as a surgical planning aid by determining renal tumor and kidney volumes through segmentation on single-phase CT. Materials and Methods: After Institutional Review Board approval, the CT images of 319 patients were retrospectively analyzed. Two distinct CNNs were developed for (1) bounding cube localization of the right and left hemiabdomen and (2) segmentation of the renal parenchyma and tumor within each bounding cube. Training was performed on a randomly selected cohort of 269 patients. CNN performance was evaluated on a separate cohort of 50 patients using Sorensen-Dice coefficients (which measures the spatial overlap between the manually segmented and neural network-derived segmentations) and Pearson correlation coefficients. Experiments were run on a graphics processing unit-optimized workstation with a single NVIDIA GeForce GTX Titan X (12GB, Maxwell Architecture). Results: Median Dice coefficients for kidney and tumor segmentation were 0.970 and 0.816, respectively; Pearson correlation coefficients between CNN-generated and human-annotated estimates for kidney and tumor volume were 0.998 and 0.993 (p < 0.001), respectively. End-to-end trained CNNs were able to perform renal parenchyma and tumor segmentation on a new test case in an average of 5.6 seconds. Conclusions: Initial experience with automated DL AI demonstrates that it is capable of rapidly and accurately segmenting kidneys and renal tumors on single-phase contrast-enhanced CT scans and calculating tumor and renal volumes.

Predictive Values of Location and Volumetric MRI Injury Patterns for Neurodevelopmental Outcomes in Hypoxic-Ischemic Encephalopathy Neonates.

Hypoxic-ischemic encephalopathy (HIE) is a severe neonatal complication with up to 40-60% long-term morbidity. This study evaluates the distribution and burden of MRI changes as a prognostic indicator of neurodevelopmental (ND) outcomes at 18-24 months in HIE infants who were treated with therapeutic hypothermia (TH). Term or late preterm infants who were treated with TH for HIE were analyzed between June 2012 and March 2016. Brain MRI scans were obtained from 107 TH treated infants. For each infant, diffusion weighted brain image (DWI) sequences from a 3T Siemens scanner were obtained for analysis. Of the 107 infants, 36 of the 107 infants (33.6%) had normal brain MR images, and 71 of the 107 infants (66.4%) had abnormal MRI findings. The number of clinical seizures was significantly higher in the abnormal MRI group (p < 0.001) than in the normal MRI group. At 18-24 months, 76 of the 107 infants (70.0%) showed normal ND stages, and 31 of the 107 infants (29.0%) exhibited abnormal ND stages. A lesion size count >500 was significantly associated with abnormal ND. Similarly, the total lesion count was larger in the abnormal ND group (14.16 vs. 5.29). More lesions in the basal ganglia (BG) and thalamus areas and a trend towards more abnormal MRI scans were significantly associated with abnormal ND at 18-24 months. In addition to clinical seizure, a larger total lesion count and lesion size as well as lesion involvement of the basal ganglia and thalamus were significantly associated with abnormal neurodevelopment at 18-24 months.

Development and external validation of a prognostic tool for COVID-19 critical disease.

BACKGROUND: The rapid spread of coronavirus disease 2019 (COVID-19) revealed significant constraints in critical care capacity. In anticipation of subsequent waves, reliable prediction of disease severity is essential for critical care capacity management and may enable earlier targeted interventions to improve patient outcomes. The purpose of this study is to develop and externally validate a prognostic model/clinical tool for predicting COVID-19 critical disease at presentation to medical care. METHODS: This is a retrospective study of a prognostic model for the prediction of COVID-19 critical disease where critical disease was defined as ICU admission, ventilation, and/or death. The derivation cohort was used to develop a multivariable logistic regression model. Covariates included patient comorbidities, presenting vital signs, and laboratory values. Model performance was assessed on the validation cohort by concordance statistics. The model was developed with consecutive patients with COVID-19 who presented to University of California Irvine Medical Center in Orange County, California. External validation was performed with a random sample of patients with COVID-19 at Emory Healthcare in Atlanta, Georgia. RESULTS: Of a total 3208 patients tested in the derivation cohort, 9% (299/3028) were positive for COVID-19. Clinical data including past medical history and presenting laboratory values were available for 29% (87/299) of patients (median age, 48 years [range, 21-88 years]; 64% [36/55] male). The most common comorbidities included obesity (37%, 31/87), hypertension (37%, 32/87), and diabetes (24%, 24/87). Critical disease was present in 24% (21/87). After backward stepwise selection, the following factors were associated with greatest increased risk of critical disease: number of comorbidities, body mass index, respiratory rate, white blood cell count, % lymphocytes, serum creatinine, lactate dehydrogenase, high sensitivity troponin I, ferritin, procalcitonin, and C-reactive protein. Of a total of 40 patients in the validation cohort (median age, 60 years [range, 27-88 years]; 55% [22/40] male), critical disease was present in 65% (26/40). Model discrimination in the validation cohort was high (concordance statistic: 0.94, 95% confidence interval 0.87-1.01). A web-based tool was developed to enable clinicians to input patient data and view likelihood of critical disease. CONCLUSIONS AND RELEVANCE: We present a model which accurately predicted COVID-19 critical disease risk using comorbidities and presenting vital signs and laboratory values, on derivation and validation cohorts from two different institutions. If further validated on additional cohorts of patients, this model/clinical tool may provide useful prognostication of critical care needs.

Updates on Deep Learning and Glioma: Use of Convolutional Neural Networks to Image Glioma Heterogeneity.

Deep learning represents end-to-end machine learning in which feature selection from images and classification happen concurrently. This articles provides updates on how deep learning is being applied to the study of glioma and its genetic heterogeneity. Deep learning algorithms can detect patterns in routine and advanced MR imaging that elude the eyes of neuroradiologists and make predictions about glioma genetics, which impact diagnosis, treatment response, patient management, and long-term survival. The success of these deep learning initiatives may enhance the performance of neuroradiologists and add greater value to patient care by expediting treatment.

Impact of COVID-19 on Acute Stroke Presentation at a Comprehensive Stroke Center.

Background: COVID-19 has impacted healthcare in many ways, including presentation of acute stroke. Since time-sensitive thrombolysis is essential for reducing morbidity and mortality in acute stroke, any delays due to the pandemic can have serious consequences. Methods: We retrospectively reviewed the electronic medical records for patients presenting with acute ischemic stroke at a comprehensive stroke center in March-April 2020 (the early months of COVID-19) and compared to the same time period in 2019. Stroke metrics such as incidence, time to arrival, and immediate outcomes were assessed. Results: There were 48 acute ischemic strokes (of which 7 were transfers) in March-April 2020 compared to 64 (of which 12 were transfers) in 2019. The average last known well to arrival time (±SD) for stroke codes was 1,041 (±1682.1) min in 2020 and 554 (±604.9) min in 2019. Of the patients presenting directly to the ED with a known last known well time, 27.8% (10/36) presented in the first 4.5 h in 2020, in contrast to 40.5% (15/37) in 2019. Patients who died comprised 10.4% of the stroke cohort in 2020 (5/48) compared to 6.3% in 2019 (4/64). Conclusions: During the first 2 months of COVID-19, there were fewer overall stroke cases who presented to our hospital, and of these cases, there was delayed presentation in comparison to the same time period in 2019. Recognizing how stroke presentation may be affected by COVID-19 would allow for optimization of established stroke triage algorithms in order to ensure safe and timely delivery of stroke care during a pandemic.

Applications of Artificial Intelligence to Prostate Multiparametric MRI (mpMRI): Current and Emerging Trends.

Prostate carcinoma is one of the most prevalent cancers worldwide. Multiparametric magnetic resonance imaging (mpMRI) is a non-invasive tool that can improve prostate lesion detection, classification, and volume quantification. Machine learning (ML), a branch of artificial intelligence, can rapidly and accurately analyze mpMRI images. ML could provide better standardization and consistency in identifying prostate lesions and enhance prostate carcinoma management. This review summarizes ML applications to prostate mpMRI and focuses on prostate organ segmentation, lesion detection and segmentation, and lesion characterization. A literature search was conducted to find studies that have applied ML methods to prostate mpMRI. To date, prostate organ segmentation and volume approximation have been well executed using various ML techniques. Prostate lesion detection and segmentation are much more challenging tasks for ML and were attempted in several studies. They largely remain unsolved problems due to data scarcity and the limitations of current ML algorithms. By contrast, prostate lesion characterization has been successfully completed in several studies because of better data availability. Overall, ML is well situated to become a tool that enhances radiologists' accuracy and speed.

Glioma-Induced Alterations in Neuronal Activity and Neurovascular Coupling during Disease Progression.

Diffusely infiltrating gliomas are known to cause alterations in cortical function, vascular disruption, and seizures. These neurological complications present major clinical challenges, yet their underlying mechanisms and causal relationships to disease progression are poorly characterized. Here, we follow glioma progression in awake Thy1-GCaMP6f mice using in vivo wide-field optical mapping to monitor alterations in both neuronal activity and functional hemodynamics. The bilateral synchrony of spontaneous neuronal activity gradually decreases in glioma-infiltrated cortical regions, while neurovascular coupling becomes progressively disrupted compared to uninvolved cortex. Over time, mice develop diverse patterns of high amplitude discharges and eventually generalized seizures that appear to originate at the tumors' infiltrative margins. Interictal and seizure events exhibit positive neurovascular coupling in uninfiltrated cortex; however, glioma-infiltrated regions exhibit disrupted hemodynamic responses driving seizure-evoked hypoxia. These results reveal a landscape of complex physiological interactions occurring during glioma progression and present new opportunities for exploring novel biomarkers and therapeutic targets.

Deep Learning AI Applications in the Imaging of Glioma.

This manuscript will review emerging applications of artificial intelligence, specifically deep learning, and its application to glioblastoma multiforme (GBM), the most common primary malignant brain tumor. Current deep learning approaches, commonly convolutional neural networks (CNNs), that take input data from MR images to grade gliomas (high grade from low grade) and predict overall survival will be shown. There will be more in-depth review of recent articles that have applied different CNNs to predict the genetics of glioma on pre-operative MR images, specifically 1p19q codeletion, MGMT promoter, and IDH mutations, which are important criteria for the diagnosis, treatment management, and prognostication of patients with GBM. Finally, there will be a brief mention of current challenges with DL techniques and their application to image analysis in GBM.

Optimizing Neuro-Oncology Imaging: A Review of Deep Learning Approaches for Glioma Imaging.

Radiographic assessment with magnetic resonance imaging (MRI) is widely used to characterize gliomas, which represent 80% of all primary malignant brain tumors. Unfortunately, glioma biology is marked by heterogeneous angiogenesis, cellular proliferation, cellular invasion, and apoptosis. This translates into varying degrees of enhancement, edema, and necrosis, making reliable imaging assessment challenging. Deep learning, a subset of machine learning artificial intelligence, has gained traction as a method, which has seen effective employment in solving image-based problems, including those in medical imaging. This review seeks to summarize current deep learning applications used in the field of glioma detection and outcome prediction and will focus on (1) pre- and post-operative tumor segmentation, (2) genetic characterization of tissue, and (3) prognostication. We demonstrate that deep learning methods of segmenting, characterizing, grading, and predicting survival in gliomas are promising opportunities that may enhance both research and clinical activities.

Extent of BOLD Vascular Dysregulation Is Greater in Diffuse Gliomas without Isocitrate Dehydrogenase 1 R132H Mutation.

Purpose To determine the effect that R132H mutation status of diffuse glioma has on extent of vascular dysregulation and extent of residual blood oxygen level-dependent (BOLD) abnormality after surgical resection. Materials and Methods This study was an institutional review board-approved retrospective analysis of an institutional database of patients, and informed consent was waived. From 2010 to 2017, 39 treatment-naïve patients with diffuse glioma underwent preoperative echo-planar imaging and BOLD functional magnetic resonance imaging. BOLD vascular dysregulation maps were made by identifying voxels with time series similar to tumor and dissimilar to healthy brain. The spatial overlap between tumor and vascular dysregulation was characterized by using the Dice coefficient, and areas of BOLD abnormality outside the tumor margins were quantified as BOLD-only fraction (BOF). Linear regression was used to assess effects of R132H status on the Dice coefficient, BOF, and residual BOLD abnormality after surgical resection. Results When compared with R132H wild-type (R132H-) gliomas, R132H-mutated (R132H+) gliomas showed greater spatial overlap between BOLD abnormality and tumor (mean Dice coefficient, 0.659 ± 0.02 [standard error] for R132H+ and 0.327 ± 0.04 for R132H-; P < .001), less BOLD abnormality beyond the tumor margin (mean BOF, 0.255 ± 0.03 for R132H+ and 0.728 ± 0.04 for R132H-; P < .001), and less postoperative BOLD abnormality (residual fraction, 0.046 ± 0.0047 for R132H+ and 0.397 ± 0.045 for R132H-; P < .001). Receiver operating characteristic curve analysis showed high sensitivity and specificity in the discrimination of R132H+ tumors from R132H- tumors with calculation of both Dice coefficient and BOF (area under the receiver operating characteristic curve, 0.967 and 0.977, respectively). Conclusion R132H mutation status is an important variable affecting the extent of tumor-associated vascular dysregulation and the residual vascular dysregulation after surgical resection. © RSNA, 2018 Online supplemental material is available for this article.

Beyond the embolus: "do not miss" diffusion abnormalities of ischaemic and non-ischaemic neurological disease.

Given the rapid evolution and technological advances in the diagnosis and treatment of acute ischaemic stroke (AIS), including the proliferation of comprehensive stroke centres and increasing emphasis on interventional stroke therapies, the need for prompt recognition of stroke due to acute large vessel occlusion has received significant attention in the recent literature. Diffusion-weighted imaging (DWI) is the gold standard for the diagnosis of acute ischaemic stroke, as images appear positive within minutes of ischaemic injury, and a high signal-to-noise ratio enables even punctate infarcts to be readily detected. DWI lesions resulting from a single arterial embolic occlusion or steno-occlusive lesion classically lateralise and conform to a specific arterial territory. When there is a central embolic source (e.g. left atrial thrombus), embolic infarcts are often found in multiple vascular territories. However, ischaemic disease arising from aetiologies other than arterial occlusion will often not conform to an arterial territory. Furthermore, there are several important entities unrelated to ischaemic disease that can present with abnormal DWI and which should not be confused with infarct. This pictorial review explores the scope and typical DWI findings of select neurologic conditions beyond acute arterial occlusion, which should not be missed or misinterpreted. TEACHING POINTS: • DWI abnormalities due to acute arterial occlusion must be promptly identified. • DWI abnormalities not due to arterial occlusion will often not conform to an arterial territory. • Several important non-ischaemic entities can present on DWI and should not be confused with infarct.