Automated Segmentation of Lymph Nodes on Neck CT Scans Using Deep Learning.

Early and accurate detection of cervical lymph nodes is essential for the optimal management and staging of patients with head and neck malignancies. Pilot studies have demonstrated the potential for radiomic and artificial intelligence (AI) approaches in increasing diagnostic accuracy for the detection and classification of lymph nodes, but implementation of many of these approaches in real-world clinical settings would necessitate an automated lymph node segmentation pipeline as a first step. In this study, we aim to develop a non-invasive deep learning (DL) algorithm for detecting and automatically segmenting cervical lymph nodes in 25,119 CT slices from 221 normal neck contrast-enhanced CT scans from patients without head and neck cancer. We focused on the most challenging task of segmentation of small lymph nodes, evaluated multiple architectures, and employed U-Net and our adapted spatial context network to detect and segment small lymph nodes measuring 5-10 mm. The developed algorithm achieved a Dice score of 0.8084, indicating its effectiveness in detecting and segmenting cervical lymph nodes despite their small size. A segmentation framework successful in this task could represent an essential initial block for future algorithms aiming to evaluate small objects such as lymph nodes in different body parts, including small lymph nodes looking normal to the naked human eye but harboring early nodal metastases.

An assessment of radiology resident competency in identifying suppurative retropharyngeal lymphadenitis: an examination using the WIDI SIM platform.

BACKGROUND AND PURPOSE: Suppurative retropharyngeal lymphadenitis is a retropharyngeal space infection almost exclusively seen in the young (4-8 years old) pediatric population. It can be misdiagnosed as a retropharyngeal abscess, leading to unnecessary invasive treatment procedures. This retrospective study aims to assess radiology residents' ability to independently identify CT imaging findings and make a definitive diagnosis of suppurative retropharyngeal lymphadenitis in a simulated call environment. MATERIALS AND METHODS: The Wisdom in Diagnostic Imaging Emergent/Critical Care Radiology Simulation (WIDI SIM) is a computer-aided emergency imaging simulation proven to be a reliable method for assessing resident preparedness for independent radiology call. The simulation included 65 cases across various imaging modalities of varying complexity, including normal studies, with one case specifically targeting suppurative retropharyngeal adenitis identification. Residents' free text responses were manually scored by faculty members using a standardized grading rubric, with errors subsequently classified by type. RESULTS: A total of 543 radiology residents were tested in three separate years on the imaging findings of suppurative retropharyngeal lymphadenitis using the Wisdom in Diagnostic Imaging simulation web-based testing platform. Suppurative retropharyngeal lymphadenitis was consistently underdiagnosed by radiology residents being tested for call readiness irrespective of the numbers of years in training. On average, only 3.5% of radiology residents were able to correctly identify suppurative retropharyngeal lymphadenitis on a contrast-enhanced computed tomography (CT). CONCLUSIONS: Our findings underscore a potential gap in radiology residency training related to the accurate identification of suppurative retropharyngeal lymphadenitis, highlighting the potential need for enhanced educational efforts in this area.

Radiology resident competency in detecting basilar artery occlusion: a simulation-based assessment.

PURPOSE: Basilar artery strokes are rare but can have characteristic imaging findings that can often be overlooked. This retrospective study aims to assess radiology residents' ability to identify CT imaging findings of basilar artery occlusion in a simulated call environment. METHODS: The Wisdom in Diagnostic Imaging Emergent/Critical Care Radiology Simulation (WIDI SIM)-a tested and reliable computer-aided emergency imaging simulation-was employed to assess resident readiness for independent radiology call. The simulations include 65 cases of varying complexity, including normal studies, with one case specifically assessing basilar artery stroke. Residents were presented with a single, unique case of basilar artery occlusion in two separate years of testing and were only provided with non-contrast CT images. Residents' free text responses were manually scored by faculty members using a standardized grading rubric, with errors subsequently classified by type. RESULTS: A total of 454 radiology residents were tested in two separate years on the imaging findings of basilar artery occlusion using the Wisdom in Diagnostic Imaging simulation web-based testing platform. Basilar artery occlusion was consistently underdiagnosed by radiology residents being tested for call readiness irrespective of the numbers of years in training. On average, only 14% of radiology residents were able to correctly identify basilar artery occlusion on non-contrast CT. CONCLUSIONS: Our findings underscore a potential gap in radiology residency training related to the detection of basilar artery occlusion, highlighting the potential need for increased educational efforts in this area.

Enhancing Radiology Education With a Case-Based Intro to Radiology on the UF WIDI e-Learning Platform.

RATIONALE AND OBJECTIVES: This study explores the implementation and efficacy of an online, interactive, case-based radiology education tool, Wisdom in Diagnostic Imaging (WIDI) Case-Based Intro to Radiology (CBIR). We hypothesize that the WIDI CBIR platform would enhance radiology teaching, foster critical thinking, and provide a comprehensive curriculum in imaging interpretation and utilization. MATERIALS AND METHODS: A focus group consisting of 1 undergraduate, 7 medical students, 9 physician assistant students, and 3 PhD students participated in this study. We tested 3 different teaching methods: a didactic approach without WIDI, a proctored didactic approach using WIDI, and a flipped classroom approach using WIDI. An online survey was conducted to assess student preference and feedback on these methods and the use of WIDI in their curriculum. RESULTS: Most students preferred the proctored didactic approach with WIDI. They reported that the platform complemented their curriculum and encouraged critical thinking. The modules covered adequate clinical and imaging details and enhanced their skills in imaging interpretation. Despite the limitations of a small sample size and reliance on self-reported outcomes, this study indicates that the WIDI platform could be integrated into PA and medical school curricula throughout the US, offering a standardized radiology curriculum. CONCLUSION: The UF WIDI appears to be a promising tool for modernizing radiology education, improving imaging interpretation skills, and enhancing appropriate imaging selection among nonradiologist medical learners. WIDI offers case-based education in imaging use, workflow, search-pattern selection, and interpretation of common radiological findings, potentially bridging the gap in radiology education. Further research and larger studies are required to assess the long-term impact on performance and clinical practice.

Medical imaging and multimodal artificial intelligence models for streamlining and enhancing cancer care: opportunities and challenges.

INTRODUCTION: Artificial intelligence (AI) has the potential to transform oncologic care. There have been significant developments in AI applications in medical imaging and increasing interest in multimodal models. These are likely to enable improved oncologic care through more precise diagnosis, increasingly in a more personalized and less invasive manner. In this review, we provide an overview of the current state and challenges that clinicians, administrative personnel and policy makers need to be aware of and mitigate for the technology to reach its full potential. AREAS COVERED: The article provides a brief targeted overview of AI, a high-level review of the current state and future potential AI applications in diagnostic radiology and to a lesser extent digital pathology, focusing on oncologic applications. This is followed by a discussion of emerging approaches, including multimodal models. The article concludes with a discussion of technical, regulatory challenges and infrastructure needs for AI to realize its full potential. EXPERT OPINION: There is a large volume of promising research, and steadily increasing commercially available tools using AI. For the most advanced and promising precision diagnostic applications of AI to be used clinically, robust and comprehensive quality monitoring systems and informatics platforms will likely be required.

Introduction to Radiomics and Artificial Intelligence: A Primer for Radiologists.

Health informatics and artificial intelligence (AI) are expected to transform the healthcare enterprise and the future practice of radiology. There is an increasing body of literature on radiomics and deep learning/AI applications in medical imaging. There are also a steadily increasing number of FDA cleared AI applications in radiology. It is therefore essential for radiologists to have a basic understanding of these approaches, whether in academia or private practice. In this article, we will provide an overview of the field and familiarize the readers with the fundamental concepts behind these approaches.

Applications of Artificial Intelligence in the Radiology Roundtrip: Process Streamlining, Workflow Optimization, and Beyond.

There are many impactful applications of artificial intelligence (AI) in the electronic radiology roundtrip and the patient's journey through the healthcare system that go beyond diagnostic applications. These tools have the potential to improve quality and safety, optimize workflow, increase efficiency, and increase patient satisfaction. In this article, we review the role of AI for process improvement and workflow enhancement which includes applications beginning from the time of order entry, scan acquisition, applications supporting the image interpretation task, and applications supporting tasks after image interpretation such as result communication. These non-diagnostic workflow and process optimization tasks are an important part of the arsenal of potential AI tools that can streamline day to day clinical practice and patient care.

Osteoid Osteoma of the Coracoid Process Presenting as Adhesive Capsulitis in a 10-Year-Old Male: A Case Report.

A 10-year-old male presented with symptoms in his right shoulder indicative of adhesive capsulitis. Radiographic films did not demonstrate any osseous abnormalities. Magnetic resonance imaging demonstrated the presence of an eccentric lesion within the coracoid process consistent with an osteoid osteoma. Six months after surgical removal the patient is back to full activities. For the pediatric population, surgeons must always consider diagnoses that could alter a patient's growth or result in long-term disability. In particular, an atypical presentation of musculoskeletal disease in a pediatric patient presenting with a disease that typically is seen in the older population warrants further workup.

Bubbles in the belly: How well do radiology trainees recognize pneumatosis in pediatric patients on plain film?

OBJECTIVES: Pneumatosis intestinalis is a radiographic finding that refers to the presence of gas within the walls of the small or large bowel. This sign is diagnostic in the setting of premature infants with suspected necrotizing enterocolitis. Earlier detection of NEC on X-ray is vital to improve the overall management in these infants. The identification of pneumatosis intestinalis concerning for necrotizing enterocolitis by the "on-call" trainee is critical in the care of the preterm infant. Our objective was to study trainees' ability to identify pneumatosis on an emergent overnight film. METHODS: The Emergent/Critical Care Imaging SIMulation (WIDI SIM) is a strategically designed computer-aided simulation of an emergency imaging experience that has been rigorously tested and proven to be a reliable means for assessing radiology trainee preparedness to competently and independently cover radiology call. One test case each year included requires the trainee to accurately identify pneumatosis on a plain radiograph of the abdomen in a neonate. RESULTS: A total of 463 radiology trainees from 32 distinct training programs across the country were given a case of pneumatosis using the WIDI simulation web-based testing platform. On average only 28% of radiology trainees were able to correctly identify pneumatosis intestinalis on plain film. Although the sample sizes in the upper-level trainees were smaller, those with greater number years of training performed better. CONCLUSIONS: Further training must be given to radiology trainees to accurately recognize pneumatosis and report concerns for necrotizing enterocolitis. ADVANCES IN KNOWLEDGE: This paper is the first to describe and assess the ability of the "on-call" radiology trainee to accurately recognize pneumatosis and report concerns for necrotizing enterocolitis. Our paper includes the largest cohort of radiology trainees evaluated to this date.

Does shoulder stability differ with or without subscapularis repair after primary reverse total shoulder arthroplasty? A systematic review.

The use of reverse total shoulder arthroplasty (RTSA) has expanded from its original indication as a rotator cuff arthropathy treatment to include a large variety of pathologies. A frequently reported complication with this surgery is postoperative shoulder instability with reported incidence varying widely from 2.3 to 38%. The etiology for this instability is broad and includes prosthesis design, mechanical impingement, surgical technique, and axillary/deltoid function. A PROSPERO-registered systematic review was performed utilizing PRISMA guidelines using Cochrane, PUBMED, Embase, and Eline. Of the 1442 studies initially identified, 7 studies met all inclusion criteria, all of which were level III or IV evidence. All 7 studies evaluated postoperative instability, but no study reported a statistically significant difference in instability rates between the groups. Dislocations occurred in 5 patients (5/679, 0.7%) with subscapularis repair and 8 patients (8/527, 1.5%) without repair. A nonsignificant difference in the risk of instability for surgeries with repair compared to surgeries without repair was found (overall risk difference: 0.01, random effects 95% confidence interval: -0.00 to 0.02, P = .11). This review suggests no difference in postoperative shoulder instability rates between patients that underwent primary RTSA with or without subsequent repair of the subscapularis tendon. Interestingly, one study comparing implants with a medialized or nonlateralized implant showed a significantly increased rate of dislocation with the medialized group compared to the lateralized group. When these groups were then stratified based on subscapularis repair status, there was no increased risk with a nonrepaired tendon. This study suggests that implant design may have more influence on the stability of RTSA than subscapularis status. However, overall, there does appear to be a trend suggesting improved postoperative clinical outcomes and active range of motion for patients with a subscapularis repair vs. without a repair. Further research is needed to better elucidate the ideal combination of surgical technique and implant design to minimize postoperative glenohumeral joint instability while optimizing postoperative clinical outcomes and range of motion after primary RTSA.

Radial Flow Perfusion Enables Real-Time Profiling of Cellular Metabolism at Low Oxygen Levels with Hyperpolarized 13C NMR Spectroscopy.

In this study, we describe new methods for studying cancer cell metabolism with hyperpolarized 13C magnetic resonance spectroscopy (HP 13C MRS) that will enable quantitative studies at low oxygen concentrations. Cultured hepatocellular carcinoma cells were grown on the surfaces of non-porous microcarriers inside an NMR spectrometer. They were perfused radially from a central distributer in a modified NMR tube (bioreactor). The oxygen level of the perfusate was continuously monitored and controlled externally. Hyperpolarized substrates were injected continuously into the perfusate stream with a newly designed system that prevented oxygen and temperature perturbations in the bioreactor. Computational and experimental results demonstrated that cell mass oxygen profiles with radial flow were much more uniform than with conventional axial flow. Further, the metabolism of HP [1-13C]pyruvate was markedly different between the two flow configurations, demonstrating the importance of avoiding large oxygen gradients in cell perfusion experiments.

Rate of resident recognition of nonaccidental trauma: how well do residents perform?

BACKGROUND: A resident working under the entrustable professional activity of a graduated independent coverage model must identify nonaccidental trauma and notify clinicians of this concern to facilitate potential removal of the child from harm. The resident's role in identifying child abuse has not previously been studied. OBJECTIVE: The purpose of this study was to assess radiology residents' ability to identify radiographic imaging findings of nonaccidental trauma in a simulated call environment. MATERIALS AND METHODS: The Wisdom in Diagnostic Imaging Emergent/Critical Care Radiology Simulation (WIDI SIM) is a strategically designed computer-aided simulation of an emergency imaging experience that has been rigorously tested and proven to be a reliable means for assessing resident preparedness to competently and independently cover radiology call. The residents are provided with 65 cases of varying difficulty, including normal studies. Those cases include ones that require the resident to accurately identify skeletal findings of nonaccidental trauma. The residents respond using free text that is then scored manually by faculty members utilizing a robust grading rubric. Missed cases are then categorized by observation error and/or interpretative errors. RESULTS: A total of 675 radiology residents were given a case of nonaccidental trauma using the WIDI simulation web-based test platform between 2014 and 2017. Child abuse was consistently underdiagnosed by radiology residents being tested for call readiness irrespective of the numbers of years in training. Correct diagnostic score averages per year ranged from 10% to 79%. In year 1, 71% of residents correctly diagnosed nonaccidental trauma with a total of 101 participants. In year 2, 130 residents participated with only 20% answering correctly. In year 3, there were 201 participants with 79% correctly diagnosing nonaccidental trauma. In year 4, only 10% of 243 residents accurately diagnosed nonaccidental trauma. Residents from various stages of training were tested. Cases used with leading histories provided, accounted for a higher correct call rate. CONCLUSION: Radiology residency programs may be deficient in teaching residents to accurately recognize and report child abuse. This deficiency has potential implications for programs where a radiology attending is not present after hours as it raises the concern that child abuse may be missed by residents taking independent call. These data further question whether graduating radiologists are competent to recognize nonaccidental trauma and demonstrates the need for additional mandatory training during radiology residency.

Full Resolution Simulation for Evaluation of Critical Care Imaging Interpretation; Part 2: Random Effects Reveal the Interplay Between Case Difficulty, Resident Competence, and the Training Environment.

RATIONALE AND OBJECTIVES: To further characterize empirical data from a full-resolution simulation of critical care imaging coupled with post hoc grading of resident's interpretations by senior radiologists. To present results from estimating the random effects terms in a comprehensive mixed (hierarchical) regression model. MATERIALS AND METHODS: After accounting for 9 fixed effects detailed in Part 1 of this paper, we estimated normally distributed random effects, expressed in terms of score offsets for each case, resident, program, and grader. RESULTS: The fixed effects alone explained 8.8% of score variation and adding the random effects increased explanatory power of the model to account for 36% of score variation. As quantified by intraclass correlation coefficient (ICC = 28.5%; CI: 25.1-31.6) the majority of score variation is directly attributable to the case at hand. This "case difficulty" measure has reliability of 95%. Individual residents accounted for much of the remaining score variation (ICC = 5.3%; CI: 4.6-5.9) after adjusting for all other effects including level of training. The reliability of this "resident competence" measure is 82%. Residency training program influence on scores was small (ICC = 1.1%; CI: 0.42-1.7). Although a few significantly high and low ones can be identified, reliability of 73% militates for caution. At the same time, low intraprogram variation is very encouraging. Variation attributable to differences between graders was minimal (ICC = 0.58%; CI: 0.0-1.2) which reassures us that the method of scoring is reliable, consistent, and likely extensible. CONCLUSION: Full resolution simulation based evaluation of critical care radiology interpretation is being conducted remotely and efficiently at large scale. A comprehensive mixed model of the resulting scores reliably quantifies case difficulty and resident competence.

Full Resolution Simulation for Evaluation of Critical Care Imaging Interpretation; Part 1: Fixed Effects Identify Influences of Exam, Specialty, Fatigue, and Training on Resident Performance.

RATIONALE AND OBJECTIVES: To describe our full-resolution simulation of critical care imaging coupled with posthoc grading of resident's interpretations and present results from the fixed effects terms in a comprehensive mixed regression model of the resulting scores. MATERIALS AND METHODS: The system delivered full resolution DICOM studies via clinical-grade viewing software integrated with a custom built web-based workflow and reporting system. The interpretations submitted by participating residents from 47 different programs were graded (scores of 0-10) on a case by case basis by a cadre of faculty members from our department. The data from 5 yearly (2014-2018) cycles consisting of 992 separate 65 case, 8 hour simulation sessions were collated from the transaction records. We used a mixed (hierarchical) statistical model with nine fixed and four random independent variables. In this paper, we present the results from the nine fixed effects. RESULTS: There were 19,916/63,839 (27.0%, CI 26.7%-27.4%) scores in the 0-2 range (i.e., clinically significant miss). Neurological cases were more difficult with adjusted scores 2.3 (CI 1.9-3.2) lower than body/musculoskeletal cases. There was a small (0.3, CI 0.20-0.38 points) but highly significant (p<0.0001) decrease in score for the final 13/65 cases (fifth quintile) as evidence of fatigue during the last hour of an 8 hour shift. By comparing adjusted scores from mid-R1 (quarter 3) to late-R3 (quarter 12) we estimate the training effect as an increase of 2.2 (CI 1.90-2.50) points. CONCLUSION: Full resolution simulation based evaluation of critical care radiology interpretation is being conducted remotely and efficiently at large scale. Analysis of the resulting scores yields multiple insights into the interpretative process.

Hyperpolarized Metabolic Imaging Detects Latent Hepatocellular Carcinoma Domains Surviving Locoregional Therapy.

BACKGROUND AND AIMS: Advances in cancer treatment have improved survival; however, local recurrence and metastatic disease-the principal causes of cancer mortality-have limited the ability to achieve durable remissions. Local recurrences arise from latent tumor cells that survive therapy and are often not detectable by conventional clinical imaging techniques. Local recurrence after transarterial embolization (TAE) of hepatocellular carcinoma (HCC) provides a compelling clinical correlate of this phenomenon. In response to TAE-induced ischemia, HCC cells adapt their growth program to effect a latent phenotype that precedes local recurrence. APPROACH AND RESULTS: In this study, we characterized and leveraged the metabolic reprogramming demonstrated by latent HCC cells in response to TAE-induced ischemia to enable their detection in vivo using dynamic nuclear polarization (DNP) magnetic resonance spectroscopic imaging (MRSI) of 13 carbon-labeled substrates. Under TAE-induced ischemia, latent HCC cells demonstrated reduced metabolism and developed a dependence on glycolytic flux to lactate. Despite the hypometabolic state of these cells, DNP-MRSI of 1-13 C-pyruvate and its downstream metabolites, 1-13 C-lactate and 1-13 C-alanine, predicted histological viability. CONCLUSIONS: These studies provide a paradigm for imaging latent, treatment-refractory cancer cells, suggesting that DNP-MRSI provides a technology for this application.

A Review of Perceptual Expertise in Radiology-How it develops, How we can test it, and Why humans still matter in the era of Artificial Intelligence.

As the first step in image interpretation is detection, an error in perception can prematurely end the diagnostic process leading to missed diagnoses. Because perceptual errors of this sort-"failure to detect"-are the most common interpretive error (and cause of litigation) in radiology, understanding the nature of perceptual expertise is essential in decreasing radiology's long-standing error rates. In this article, we review what constitutes a perceptual error, the existing models of radiologic image perception, the development of perceptual expertise and how it can be tested, perceptual learning methods in training radiologists, and why understanding perceptual expertise is still relevant in the era of artificial intelligence. Adding targeted interventions, such as perceptual learning, to existing teaching practices, has the potential to enhance expertise and reduce medical error.

Fructose-1,6-Bisphosphatase 2 Inhibits Sarcoma Progression by Restraining Mitochondrial Biogenesis.

The remarkable cellular and genetic heterogeneity of soft tissue sarcomas (STSs) limits the clinical benefit of targeted therapies. Here, we show that expression of the gluconeogenic isozyme fructose-1,6-bisphosphatase 2 (FBP2) is silenced in a broad spectrum of sarcoma subtypes, revealing an apparent common metabolic feature shared by diverse STSs. Enforced FBP2 expression inhibits sarcoma cell and tumor growth through two distinct mechanisms. First, cytosolic FBP2 antagonizes elevated glycolysis associated with the "Warburg effect," thereby inhibiting sarcoma cell proliferation. Second, nuclear-localized FBP2 restrains mitochondrial biogenesis and respiration in a catalytic-activity-independent manner by inhibiting the expression of nuclear respiratory factor and mitochondrial transcription factor A (TFAM). Specifically, nuclear FBP2 colocalizes with the c-Myc transcription factor at the TFAM locus and represses c-Myc-dependent TFAM expression. This unique dual function of FBP2 provides a rationale for its selective suppression in STSs, identifying a potential metabolic vulnerability of this malignancy and possible therapeutic target.

Misregulation of Drosophila Myc Disrupts Circadian Behavior and Metabolism.

Drosophila Myc (dMyc) is highly conserved and functions as a transcription factor similar to mammalian Myc. We previously found that oncogenic Myc disrupts the molecular clock in cancer cells. Here, we demonstrate that misregulation of dMyc expression affects Drosophila circadian behavior. dMyc overexpression results in a high percentage of arrhythmic flies, concomitant with increases in the expression of clock genes cyc, tim, cry, and cwo. Conversely, flies with hypomorphic mutations in dMyc exhibit considerable arrhythmia, which can be rescued by loss of dMnt, a suppressor of dMyc activity. Metabolic profiling of fly heads revealed that loss of dMyc and its overexpression alter steady-state metabolite levels and have opposing effects on histidine, the histamine precursor, which is rescued in dMyc mutants by ablation of dMnt and could contribute to effects of dMyc on locomotor behavior. Our results demonstrate a role of dMyc in modulating Drosophila circadian clock, behavior, and metabolism.