RESUMEN
INTRODUCTION: This study investigates how quantitative texture analysis can be used to non-invasively identify novel radiogenomic correlations with clear cell renal cell carcinoma (ccRCC) biomarkers. METHODS: The Cancer Genome Atlas-Kidney Renal Clear Cell Carcinoma open-source database was used to identify 190 sets of patient genomic data that had corresponding multiphase contrast-enhanced CT images in The Cancer Imaging Archive. 2,824 radiomic features spanning fifteen texture families were extracted from CT images using a custom-built MATLAB software package. Robust radiomic features with strong inter-scanner reproducibility were selected. Random forest, AdaBoost, and elastic net machine learning (ML) algorithms evaluated the ability of the selected radiomic features to predict the presence of 12 clinically relevant molecular biomarkers identified from the literature. ML analysis was repeated with cases stratified by stage (I/II vs. III/IV) and grade (1/2 vs. 3/4). 10-fold cross validation was used to evaluate model performance. RESULTS: Before stratification by tumor grade and stage, radiomics predicted the presence of several biomarkers with weak discrimination (AUC 0.60-0.68). Once stratified, radiomics predicted KDM5C, SETD2, PBRM1, and mTOR mutation status with acceptable to excellent predictive discrimination (AUC ranges from 0.70 to 0.86). CONCLUSIONS: Radiomic texture analysis can potentially identify a variety of clinically relevant biomarkers in patients with ccRCC and may have a prognostic implication.
Asunto(s)
Carcinoma de Células Renales , Neoplasias Renales , Humanos , Carcinoma de Células Renales/diagnóstico por imagen , Carcinoma de Células Renales/genética , Carcinoma de Células Renales/patología , Neoplasias Renales/diagnóstico por imagen , Neoplasias Renales/genética , Neoplasias Renales/patología , Reproducibilidad de los Resultados , Tomografía Computarizada por Rayos X/métodos , Aprendizaje Automático , Estudios RetrospectivosRESUMEN
OBJECTIVES: To evaluate the utility of CT-based radiomics signatures in discriminating low-grade (grades 1-2) clear cell renal cell carcinomas (ccRCC) from high-grade (grades 3-4) and low TNM stage (stages I-II) ccRCC from high TNM stage (stages III-IV). METHODS: A total of 587 subjects (mean age 60.2 years ± 12.2; range 22-88.7 years) with ccRCC were included. A total of 255 tumors were high grade and 153 were high stage. For each subject, one dominant tumor was delineated as the region of interest (ROI). Our institutional radiomics pipeline was then used to extract 2824 radiomics features across 12 texture families from the manually segmented volumes of interest. Separate iterations of the machine learning models using all extracted features (full model) as well as only a subset of previously identified robust metrics (robust model) were developed. Variable of importance (VOI) analysis was performed using the out-of-bag Gini index to identify the top 10 radiomics metrics driving each classifier. Model performance was reported using area under the receiver operating curve (AUC). RESULTS: The highest AUC to distinguish between low- and high-grade ccRCC was 0.70 (95% CI 0.62-0.78) and the highest AUC to distinguish between low- and high-stage ccRCC was 0.80 (95% CI 0.74-0.86). Comparable AUCs of 0.73 (95% CI 0.65-0.8) and 0.77 (95% CI 0.7-0.84) were reported using the robust model for grade and stage classification, respectively. VOI analysis revealed the importance of neighborhood operation-based methods, including GLCM, GLDM, and GLRLM, in driving the performance of the robust models for both grade and stage classification. CONCLUSION: Post-validation, CT-based radiomics signatures may prove to be useful tools to assess ccRCC grade and stage and could potentially add to current prognostic models. Multiphase CT-based radiomics signatures have potential to serve as a non-invasive stratification schema for distinguishing between low- and high-grade as well as low- and high-stage ccRCC. KEY POINTS: ⢠Radiomics signatures derived from clinical multiphase CT images were able to stratify low- from high-grade ccRCC, with an AUC of 0.70 (95% CI 0.62-0.78). ⢠Radiomics signatures derived from multiphase CT images yielded discriminative power to stratify low from high TNM stage in ccRCC, with an AUC of 0.80 (95% CI 0.74-0.86). ⢠Models created using only robust radiomics features achieved comparable AUCs of 0.73 (95% CI 0.65-0.80) and 0.77 (95% CI 0.70-0.84) to the model with all radiomics features in classifying ccRCC grade and stage, respectively.
Asunto(s)
Carcinoma de Células Renales , Neoplasias Renales , Adulto , Anciano , Anciano de 80 o más Años , Área Bajo la Curva , Carcinoma de Células Renales/diagnóstico por imagen , Carcinoma de Células Renales/patología , Humanos , Neoplasias Renales/diagnóstico por imagen , Neoplasias Renales/patología , Aprendizaje Automático , Persona de Mediana Edad , Estudios Retrospectivos , Tomografía Computarizada por Rayos X/métodos , Adulto JovenRESUMEN
OBJECTIVES: Our purpose was to differentiate between malignant from benign soft tissue neoplasms using a combination of MRI-based radiomics metrics and machine learning. METHODS: Our retrospective study identified 128 histologically diagnosed benign (n = 36) and malignant (n = 92) soft tissue lesions. 3D ROIs were manually drawn on 1 sequence of interest and co-registered to other sequences obtained during the same study. One thousand seven hundred eight radiomics features were extracted from each ROI. Univariate analyses with supportive ROC analyses were conducted to evaluate the discriminative power of predictive models constructed using Real Adaptive Boosting (Adaboost) and Random Forest (RF) machine learning approaches. RESULTS: Univariate analyses demonstrated that 36.89% of individual radiomics varied significantly between benign and malignant lesions at the p ≤ 0.05 level. Adaboost and RF performed similarly well, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81), respectively, after 10-fold cross-validation. Restricting the machine learning models to only sequences extracted from T2FS and STIR sequences maintained comparable performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84), respectively. CONCLUSION: Machine learning decision classifiers constructed from MRI-based radiomics features show promising ability to preoperatively discriminate between benign and malignant soft tissue masses. Our approach maintains applicability even when the dataset is restricted to T2FS and STIR fluid-sensitive sequences, which may bolster practicality in clinical application scenarios by eliminating the need for complex co-registrations for multisequence analysis. KEY POINTS: ⢠Predictive models constructed from MRI-based radiomics data and machine learning-augmented approaches yielded good discriminative power to correctly classify benign and malignant lesions on preoperative scans, with AUCs of 0.77 (95% CI 0.68-0.85) and 0.72 (95% CI 0.63-0.81) for Real Adaptive Boosting (Adaboost) and Random Forest (RF), respectively. ⢠Restricting the models to only use metrics extracted from T2 fat-saturated (T2FS) and Short-Tau Inversion Recovery (STIR) sequences yielded similar performance, with AUCs of 0.73 (95% CI 0.64-0.82) and 0.75 (95% CI 0.65-0.84) for Adaboost and RF, respectively. ⢠Radiomics-based machine learning decision classifiers constructed from multicentric data more closely mimic the real-world practice environment and warrant additional validation ahead of prospective implementation into clinical workflows.
Asunto(s)
Sarcoma , Neoplasias de los Tejidos Blandos , Humanos , Imagen por Resonancia Magnética , Estudios Prospectivos , Estudios Retrospectivos , Neoplasias de los Tejidos Blandos/diagnóstico por imagenRESUMEN
Objective: Increasing evidence over the last decade suggests that many cases of unexplained chronic cough (UCC) have a neurogenic etiology, with laryngeal hypersensitivity (LH) being identified as a key mechanism. Official guidelines since 2015 have adopted use of neuromodulators and adjuvant speech therapy as a result, but historically implementation of guidelines is slow. Our survey aimed to investigate gaps in diagnosis and management practices of otolaryngology providers in caring for patients with UCC. Study Design: Cross-sectional study. Setting: Survey. Methods: 12-item survey was distributed to 110 otolaryngology practitioners experienced in diagnosis and treatment of chronic cough at a regional otolaryngology continuing education conference. Statistical analysis included Kendall's Tau Rank Correlation to measure the ordinal association between responses to questions, and Fisher's exact test to determine if there were associations between responses and years of career experience. Results: Forty eligible respondents underwent subsequent analysis. There was no association between frequency of identifying LH as a primary etiology and use of neuromodulators (τ = 0.23, P = .10). However, there was a significant correlation between LH and referrals to speech therapy (τ = 0.27, P = .05). Fisher's exact test did not reveal any significant differences among any responses based on practitioner experience. Conclusion: Our results indicate a possible disparity in treatment of UCC with neuromodulators and the utilization of speech therapy despite guideline recommendations advocating for neuromodulators with adjuvant speech therapy. Further research with larger sample sizes and more specific inquiries is necessary to elucidate this association and control for any regional differences.
RESUMEN
OBJECTIVES: To evaluate the performance of machine learning-augmented MRI-based radiomics models for predicting response to neoadjuvant chemotherapy (NAC) in soft tissue sarcomas. METHODS: Forty-four subjects were identified retrospectively from patients who received NAC at our institution for pathologically proven soft tissue sarcomas. Only subjects who had both a baseline MRI prior to initiating chemotherapy and a post-treatment scan at least 2 months after initiating chemotherapy and prior to surgical resection were included. 3D ROIs were used to delineate whole-tumor volumes on pre- and post-treatment scans, from which 1708 radiomics features were extracted. Delta-radiomics features were calculated by subtraction of baseline from post-treatment values and used to distinguish treatment response through univariate analyses as well as machine learning-augmented radiomics analyses. RESULTS: Though only 4.74% of variables overall reached significance at p ≤ 0.05 in univariate analyses, Laws Texture Energy (LTE)-derived metrics represented 46.04% of all such features reaching statistical significance. ROC analyses similarly failed to predict NAC response, with AUCs of 0.40 (95% CI 0.22-0.58) and 0.44 (95% CI 0.26-0.62) for RF and AdaBoost, respectively. CONCLUSION: Overall, while our result was not able to separate NAC responders from non-responders, our analyses did identify a subset of LTE-derived metrics that show promise for further investigations. Future studies will likely benefit from larger sample size constructions so as to avoid the need for data filtering and feature selection techniques, which have the potential to significantly bias the machine learning procedures.
Asunto(s)
Terapia Neoadyuvante , Sarcoma , Humanos , Estudios Retrospectivos , Imagen por Resonancia Magnética/métodos , Sarcoma/diagnóstico por imagen , Sarcoma/tratamiento farmacológico , Aprendizaje AutomáticoRESUMEN
BACKGROUND: The coronavirus disease 2019 (COVID-19) pandemic altered education, exams, and residency applications for United States medical students. AIM: To determine the specific impact of the pandemic on US medical students and its correlation to their anxiety levels. METHODS: An 81-question survey was distributed via email, Facebook and social media groups using REDCapTM. To investigate risk factors associated with elevated anxiety level, we dichotomized the 1-10 anxiety score into low (≤ 5) and high (≥ 6). This cut point represents the 25th percentile. There were 90 (29%) shown as low anxiety and 219 (71%) as high anxiety. For descriptive analyses, we used contingency tables by anxiety categories for categorical measurements with chi square test, or mean ± STD for continuous measurements followed by t-test or Wilcoxson rank sum test depending on data normality. Least Absolute Shrinkage and Selection Operator was used to select important predictors for the final multivariate model. Hierarchical Poisson regression model was used to fit the final multivariate model by considering the nested data structure of students clustered within State. RESULTS: 397 medical students from 29 states were analyzed. Approximately half of respondents reported feeling depressed since the pandemic onset. 62% of participants rated 7 or higher out of 10 when asked about anxiety levels. Stressors correlated with higher anxiety scores included "concern about being unable to complete exams or rotations if contracting COVID-19" (RR 1.34; 95%CI: 1.05-1.72, P = 0.02) and the use of mental health services such as a "psychiatrist" (RR 1.18; 95%CI: 1.01-1.3, P = 0.04). However, those students living in cities that limited restaurant operations to exclusively takeout or delivery as the only measure of implementing social distancing (RR 0.64; 95%CI: 0.49-0.82, P < 0.01) and those who selected "does not apply" for financial assistance available if needed (RR 0.83; 95%CI: 0.66-0.98, P = 0.03) were less likely to have a high anxiety. CONCLUSION: COVID-19 significantly impacted medical students in numerous ways. Medical student education and clinical readiness were reduced, and anxiety levels increased. It is vital that medical students receive support as they become physicians. Further research should be conducted on training medical students in telemedicine to better prepare students in the future for pandemic planning and virtual healthcare.
RESUMEN
Soon after reports of a novel coronavirus capable of causing severe pneumonia surfaced in late 2019, expeditious global spread of the Severe Acute Respiratory Distress Syndrome Coronavirus 2 (SARS-CoV-2) forced the World Health Organization to declare an international state of emergency. Although best known for causing symptoms of upper respiratory tract infection in mild cases and fulminant pneumonia in severe disease, Coronavirus Disease 2019 (COVID-19) has also been associated with gastrointestinal, neurologic, cardiac, and hematologic presentations. Despite concerns over poor specificity and undue radiation exposure, chest imaging nonetheless remains central to the initial diagnosis and monitoring of COVID-19 progression, as well as to the evaluation of complications. Classic features on chest CT include ground-glass and reticular opacities with or without superimposed consolidations, frequently presenting in a bilateral, peripheral, and posterior distribution. More recently, studies conducted with MRI have shown excellent concordance with chest CT in visualizing typical features of COVID-19 pneumonia. For patients in whom exposure to ionizing radiation should be avoided, particularly pregnant patients and children, pulmonary MRI may represent a suitable alternative to chest CT. Although PET imaging is not typically considered among first-line investigative modalities for the diagnosis of lower respiratory tract infections, numerous reports have noted incidental localization of radiotracer in parenchymal regions of COVID-19-associated pulmonary lesions. These findings are consistent with data from Middle East Respiratory Syndrome-CoV cohorts which suggested an ability for 18F-FDG PET to detect subclinical infection and lymphadenitis in subjects without overt clinical signs of infection. Though highly sensitive, use of PET/CT for primary detection of COVID-19 is constrained by poor specificity, as well as considerations of cost, radiation burden, and prolonged exposure times for imaging staff. Even still, decontamination of scanner bays is a time-consuming process, and proper ventilation of scanner suites may additionally require up to an hour of downtime to allow for sufficient air exchange. Yet, in patients who require nuclear medicine investigations for other clinical indications, PET imaging may yield the earliest detection of nascent infection in otherwise asymptomatic individuals. Especially for patients with concomitant malignancies and other states of immunocompromise, prompt recognition of infection and early initiation of supportive care is crucial to maximizing outcomes and improving survivability.
Asunto(s)
COVID-19/diagnóstico por imagen , Imagen por Resonancia Magnética , Tomografía de Emisión de Positrones , Tomografía Computarizada por Rayos X , Humanos , Sensibilidad y EspecificidadRESUMEN
PURPOSE: Our purpose was to conduct a comprehensive systematic review of all existing literature regarding imaging findings on chest CT and associated clinical features in pregnant patients diagnosed with COVID-19. MATERIALS & METHODS: A literature search was conducted on April 21, 2020 and updated on July 24, 2020 using PubMed, Embase, World Health Organization, and Google Scholar databases. Only studies which described chest CT findings of COVID-19 in pregnant patients were included for analysis. RESULTS: A total of 67 articles and 427 pregnant patients diagnosed with COVID-19 were analyzed. The most frequently encountered pulmonary findings on chest CT were ground-glass opacities (77.2%, 250/324), posterior lung involvement (72.5%, 50/69), multilobar involvement (71.8%, 239/333), bilateral lung involvement (69.4%, 231/333), peripheral distribution (68.1%, 98/144), and consolidation (40.9%, 94/230). Pregnant patients were also found to present more frequently with consolidation (40.9% vs. 21.0-31.8%) and pleural effusion (30.0% vs. 5.0%) in comparison to the general population. Associated clinical features included antepartum fever (198 cases), lymphopenia (128 cases), and neutrophilia (97 cases). Of the 251 neonates delivered, 96.8% had negative RT-PCR and/or IgG antibody testing for COVID-19. In the eight cases (3.2%) of reported neonatal infection, tests were either conducted on samples collected up to 72 h after birth or were found negative on all subsequent RT-PCR tests. CONCLUSION: Pregnant patients appear to present more commonly with more advanced COVID-19 CT findings compared to the general adult population. Furthermore, characteristic laboratory abnormalities found in pregnant patients tended to mirror those found in the general patient population. Lastly, results from neonatal testing suggest a low risk of vertical transmission.
Asunto(s)
COVID-19 , Enfermedades Pulmonares , Adulto , Prueba de COVID-19 , Femenino , Humanos , Recién Nacido , Pulmón , Embarazo , SARS-CoV-2 , Tomografía Computarizada por Rayos XRESUMEN
Since first report of a novel coronavirus in December of 2019, the Coronavirus Disease 2019 (COVID-19) pandemic has crippled healthcare systems around the world. While many initial screening protocols centered around laboratory detection of the virus, early testing assays were thought to be poorly sensitive in comparison to chest computed tomography, especially in asymptomatic disease. Coupled with shortages of reverse transcription polymerase chain reaction (RT-PCR) testing kits in many parts of the world, these regions instead turned to the use of advanced imaging as a first-line screening modality. However, in contrast to previous Severe Acute Respiratory Syndrome and Middle East Respiratory Syndrome coronavirus epidemics, chest X-ray has not demonstrated optimal sensitivity to be of much utility in first-line screening protocols. Though current national and international guidelines recommend for the use of RT-PCR as the primary screening tool for suspected cases of COVID-19, institutional and regional protocols must consider local availability of resources when issuing universal recommendations. Successful containment and social mitigation strategies worldwide have been thus far predicated on unified governmental responses, though the underlying ideologies of these practices may not be widely applicable in many Western nations. As the strain on the radiology workforce continues to mount, early results indicate a promising role for the use of machine-learning algorithms as risk stratification schema in the months to come.
Asunto(s)
Técnicas de Laboratorio Clínico/métodos , Control de Enfermedades Transmisibles/métodos , Infecciones por Coronavirus/diagnóstico , Salud Global , Tamizaje Masivo/métodos , Pandemias , Neumonía Viral/diagnóstico , Asia , Betacoronavirus , COVID-19 , Prueba de COVID-19 , Coronavirus , Infecciones por Coronavirus/epidemiología , Infecciones por Coronavirus/prevención & control , Infecciones por Coronavirus/virología , Europa (Continente) , Humanos , América del Norte , Pandemias/prevención & control , Neumonía Viral/epidemiología , Neumonía Viral/prevención & control , Neumonía Viral/virología , Reacción en Cadena de la Polimerasa/métodos , Estudios Retrospectivos , SARS-CoV-2 , Tomografía Computarizada por Rayos X/métodosRESUMEN
Background: The novel Severe Acute Respiratory Syndrome Coronavirus-2 has led to a global pandemic in which case fatality rate (CFR) has varied from country to country. This study aims to identify factors that may explain the variation in CFR across countries. Methods: We identified 24 potential risk factors affecting CFR. For all countries with over 5000 reported COVID-19 cases, we used country-specific datasets from the WHO, the OECD, and the United Nations to quantify each of these factors. We examined univariable relationships of each variable with CFR, as well as correlations among predictors and potential interaction terms. Our final multivariable negative binomial model included univariable predictors of significance and all significant interaction terms. Results: Across the 39 countries under consideration, our model shows COVID-19 case fatality rate was best predicted by time to implementation of social distancing measures, hospital beds per 1000 individuals, percent population over 70 years, CT scanners per 1 million individuals, and (in countries with high population density) smoking prevalence. Conclusion: Our model predicted an increased CFR for countries that waited over 14 days to implement social distancing interventions after the 100th reported case. Smoking prevalence and percentage population over the age of 70 years were also associated with higher CFR. Hospital beds per 1000 and CT scanners per million were identified as possible protective factors associated with decreased CFR.
Asunto(s)
Infecciones por Coronavirus/mortalidad , Modelos Estadísticos , Neumonía Viral/mortalidad , Distribución por Edad , Betacoronavirus , COVID-19 , Control de Enfermedades Transmisibles/tendencias , Capacidad de Camas en Hospitales , Humanos , Internacionalidad , Pandemias , SARS-CoV-2 , Fumar , Tomógrafos Computarizados por Rayos X/provisión & distribuciónRESUMEN
Since the spread of the coronavirus disease 2019 (COVID-19) was designated as a pandemic by the World Health Organization, health care systems have been forced to adapt rapidly to defer less urgent care during the crisis. The United States (U.S.) has adopted a four-phase approach to decreasing and then resuming non-essential work. Through strong restrictive measures, Phase I slowed the spread of disease, allowing states to safely diagnose, isolate, and treat patients with COVID-19. In support of social distancing measures, non-urgent studies were postponed, and this created a backlog. Now, as states transition to Phase II, restrictions on non-essential activities will ease, and radiology departments must re-establish care while continuing to mitigate the risk of COVID-19 transmission all while accommodating this backlog. In this article, we propose a roadmap that incorporates the current practice guidelines and subject matter consensus statements for the phased reopening of non-urgent and elective radiology services. This roadmap will focus on operationalizing these recommendations for patient care and workforce management. Tiered systems are proposed for the prioritization of elective procedures, with physician-to-physician communication encouraged. Infection control methods, provision of personal protective equipment (PPE), and physical distancing measures are highlighted. Finally, changes in hours of operation, hiring strategies, and remote reading services are discussed for their potential to ease the transition to normal operations.
Asunto(s)
Infecciones por Coronavirus , Asignación de Recursos para la Atención de Salud , Accesibilidad a los Servicios de Salud , Control de Infecciones , Pandemias , Neumonía Viral , Guías de Práctica Clínica como Asunto , Radiografía , Betacoronavirus , COVID-19 , Coronavirus , Infecciones por Coronavirus/epidemiología , Infecciones por Coronavirus/prevención & control , Infecciones por Coronavirus/virología , Humanos , Pandemias/prevención & control , Atención al Paciente , Equipo de Protección Personal , Neumonía Viral/epidemiología , Neumonía Viral/prevención & control , Neumonía Viral/virología , Políticas , Radiología , SARS-CoV-2 , Estados Unidos/epidemiologíaRESUMEN
BACKGROUND: Efforts to reduce nosocomial spread of COVID-19 have resulted in unprecedented disruptions in clinical workflows and numerous unexpected stressors for imaging departments across the country. Our purpose was to more precisely evaluate these impacts on radiologists through a nationwide survey. METHODS: A 43-item anonymous questionnaire was adapted from the AO Spine Foundation's survey and distributed to 1521 unique email addresses using REDCap™ (Research Electronic Data Capture). Additional invitations were sent out to American Society of Emergency Radiology (ASER) and Association of University Radiologists (AUR) members. Responses were collected over a period of 8 days. Descriptive analyses and multivariate modeling were performed using SAS v9.4 software. RESULTS: A total of 689 responses from radiologists across 44 different states met the criteria for inclusion in the analysis. As many as 61% of respondents rated their level of anxiety with regard to COVID-19 to be a 7 out of 10 or greater, and higher scores were positively correlated the standardized number of COVID-19 cases in a respondent's state (RR = 1.11, 95% CI: 1.02-1.21, p = 0.01). Citing the stressor of "personal health" was a strong predictor of higher anxiety scores (RR 1.23; 95% CI: 1.13-1.34, p < 0.01). By contrast, participants who reported needing no coping methods were more likely to self-report lower anxiety scores (RR 0.4; 95% CI: 0.3-0.53, p < 0.01). CONCLUSION: COVID-19 has had a significant impact on radiologists across the nation. As these unique stressors continue to evolve, further attention must be paid to the ways in which we may continue to support radiologists working in drastically altered practice environments and in remote settings.