Search

Supplemental material is available for this article.

While multiple factors impact disease, artificial intelligence (AI) studies in medicine often use small, non-diverse patient cohorts due to data sharing and privacy issues. Federated learning (FL) has emerged as a solution, enabling training across hospitals without direct data sharing. Here, we present FL-PedBrain, an FL platform for pediatric posterior fossa brain tumors, and evaluate its performance on a diverse, realistic, multi-center cohort. Pediatric brain tumors were targeted due to the scarcity of such datasets, even in tertiary care hospitals. Our platform orchestrates federated training for joint tumor classification and segmentation across 19 international sites. FL-PedBrain exhibits less than a 1.5% decrease in classification and a 3% reduction in segmentation performance compared to centralized data training. FL boosts segmentation performance by 20 to 30% on three external, out-of-network sites. Finally, we explore the sources of data heterogeneity and examine FL robustness in real-world scenarios with data imbalances., (© 2024. The Author(s).)

Prostate cancer may recur several years after definitive treatment, such as prostatectomy or radiation therapy. A rise in serum prostate-specific antigen (PSA) level is the first sign of disease recurrence, and this is termed biochemical recurrence . Patients with biochemical recurrence have worse survival outcomes. Radiologic localization of recurrent disease helps in directing patient management, which may vary from active surveillance to salvage radiation therapy, androgen-deprivation therapy, or other forms of systemic and local therapy. The likelihood of detecting the site of recurrence increases with higher serum PSA level. MRI provides optimal diagnostic performance for evaluation of the prostatectomy bed. Prostate-specific membrane antigen (PSMA) PET radiotracers currently approved by the U.S. Food and Drug Administration demonstrate physiologic urinary excretion, which can obscure recurrence at the vesicourethral junction. However, MRI and PSMA PET/CT have comparable diagnostic performance for evaluation of local recurrence after external-beam radiation therapy or brachytherapy. PSMA PET/CT outperforms MRI in identifying recurrence involving the lymph nodes and bones. Caveats for use of both PSMA PET/CT and MRI do exist and may cause false-positive or false-negative results. Hence, these techniques have complementary roles and should be interpreted in conjunction with each other, taking the patient history and results of any additional prior imaging studies into account. Novel PSMA agents at various stages of investigation are being developed, and preliminary data show promising results; these agents may revolutionize the landscape of prostate cancer recurrence imaging in the future. © RSNA, 2023 Quiz questions for this article are available through the Online Learning Center. See the invited commentary by Turkbey in this issue. The slide presentation from the RSNA Annual Meeting is available for this article.

This dataset is composed of cervical spine CT images with annotations related to fractures; it is available at https://www.kaggle.com/competitions/rsna-2022-cervical-spine-fracture-detection/., Competing Interests: Disclosures of conflicts of interest: H.M.L. No relevant relationships. E.C. No relevant relationships. T.R. No relevant relationships. F.C.K. Consultant for MD.ai and GE HealthCare; member of the Radiological Society of North America (RSNA) Machine Learning Steering Committee member and the Society for Imaging Informatics in Medicine Machine Learning Education Subcommittee (both unpaid). L.M.P. Associate editor for Radiology: Artificial Intelligence; patents planned, issued, or pending: US-20220051060-A1, “Methods for creating privacy-protecting synthetic data leveraging a constrained generative ensemble model,” and US-20220051402-A1, “Systems for automated lesion detection and related methods.” J.T. Provided expert witness deposition for Phillips, Spallas, & Angstadt in October 2022, unrelated to this article. R.L.B. Support from RSNA to author. E.G. No relevant relationships. K.W.Y. No relevant relationships. M.H. No relevant relationships. A.L.S. Author’s lab receives funding from the National Institutes of Health (NIH); member of the advisory board for the National Cancer Institute Imaging Data Commons. J.S. No relevant relationships. D.B. No relevant relationships. S.A. No relevant relationships. A.P.L. No relevant relationships. M.I.G.A. No relevant relationships. J.O.J. No relevant relationships. J.J.P. Author’s lab receives funding from the NIH, which pays this author’s salary. M.L. No relevant relationships. H.D. No relevant relationships. E.A. No relevant relationships. A.Y. No relevant relationships. Y.M. No relevant relationships. J.K.C. Grants or contracts from GE HealthCare and Genentech; technology licensed to Boston AI; consulting fees from Siloam Vision; deputy editor of Radiology: Artificial Intelligence. A.E.F. Standing director, liaison for information technology, of RSNA board of directors; member of RSNA News editorial board., (© 2023 by the Radiological Society of North America, Inc.)

Purpose: To assess whether size, diameter, and large vein involvement of MR-detected extramural venous invasion (MR-EMVI) have an impact on neoadjuvant therapy response in rectal adenocarcinoma., Methods: 57 patients with locally advanced rectal adenocarcinoma scanned with MRI before and after neoadjuvant therapy were included. Two abdominal radiologists evaluated the images with special emphasis on EMVI, on initial staging and after neoadjuvant treatment. The sensitivity and specificity of MRI for detection of rest EMVI were determined. The association of various MR-EMVI characteristics including number, size, and main vein involvement with treatment response was investigated. In subjects with discordance of radiology and pathology, elastin stain was performed, and images and slides were re-evaluated on site with a multidisciplinary approach., Results: At initial evaluation, 17 patients were MR-EMVI negative (29.8%) and 40 were MR-EMVI positive (70.2%). Complete/near-complete responders had less number (mean 1.45) and smaller diameter of MR-EMVI (mean 1.8 mm), when compared with partial responders (2.54 and 3.3 mm; p < 0.005). The sensitivity of MRI for rest EMVI detection was high, specificity was moderate, and in one patient elastin stain changed the final decision. In five patients with rest MR-EMVI positivity, carcinoma histopathologically had a distinctive serpiginous perivascular spread, growing along the track of vascular bundle, although it did not appear in intravascular spaces., Conclusion: This study demonstrates that not only the presence, but also size and number of EMVI that may be significant clinically and thus these parameters also ought to be incorporated to the MRI evaluation and prognostication of treatment response. From pathology perspective, tumors growing alongside major vessels may also reflect EMVI even if they are not demonstrably "intravascular.", (© 2022. The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.)

Aim: To investigate the relationship between pancreatic steatosis (PS) assessed on computed tomography (CT) and COVID-19 prognosis. Materials & methods: This retrospective study covers 396 patients with COVID-19 (mean age: 52.50 ± 16.86 years), who underwent unenhanced chest CT. PS was compared with clinical findings, comorbidities, requirements for hospitalization, intubation and intensive care unit (ICU), length of hospitalization and death. Results: PS was found to be strongly correlated with the severity of clinical findings and hospitalization rates (p < 0.001). In hospitalized patients, length of hospitalization (p = 0.002) and rates of ICU requirement (p = 0.003) were higher in patients with PS. Conclusion: PS, correlated with clinical severity and hospitalization requirement, is an independent risk factor for COVID-19., (© 2022 Future Medicine Ltd.)

BACKGROUND Metabolic syndrome affects 35% of the adult population in developed countries associated with non-alcoholic steatohepatitis, insulin resistance, and cardiovascular events. Fatty infiltration of the pancreas, or pancreatic steatosis, is a risk factor for acute pancreatitis, pancreatic malignancies, and diabetes mellitus, yet its relationship with metabolic syndrome is not well defined. METHODS We performed a single-centered retrospective observational study of 322 healthy subjects (subjects volunteering to be kidney transplant donors, mean age=46.3±13.5, 163 men and 159 women) in the last 2 years (July 2018-February 2020) from our institution. Pancreatic steatosis and hepatosteatosis were confirmed by computed tomography. RESULTS Pancreatic steatosis was present in 26.3% (85/322) of the subjects, and this finding correlated with age, body mass index (BMI), male sex, a family history of diabetes, creatinine, cystatin C, uric acid, low-density lipoprotein (LDL) cholesterol, triglycerides, glycemia, hemoglobin, transverse body diameter, and subcutaneous fat thickness levels by univariable logistic regression. On multiple linear regression only age (95% CI 1.01, 1.06), BMI (95% CI 1.01, 1.19), male sex (95% CI 1.49-5.99), uric acid (95% CI 1.01, 1.76), and subcutaneous fat thickness levels (95% CI 1.21-2.36) remained independently associated with pancreatic steatosis. CONCLUSION Pancreatic steatosis is common and associated with obesity, elevated serum uric acid, subcutaneous fat thickness, and male sex. Future studies are needed to evaluate if there are specific clinical consequences to the presence of pancreatic steatosis., Competing Interests: CONFLICT OF INTEREST The authors declare no conflict of interest related to this work., (© 2021 The Author(s).)

The need to predict acoustic propagation through marine sediments that contain gas bubbles has become increasingly important for civil engineering and climate studies. There are relatively few in situ acoustic wave propagation studies of muddy intertidal sediments, in which bubbles of biogenic gas (generally methane, a potent greenhouse gas) are commonly found. We used a single experimental rig to conduct two in situ intertidal acoustical experiments to improve understanding of acoustic remote sensing of gassy sediments, eventually including gas bubble size distributions. In the first experiment, we measured sediment sound speed and attenuation between four aligned hydrophones for a quasi-plane wave propagating along the array. The second experiment involved a focused insonified sediment volume created by two transducers emitting coincident sound beams at different frequencies that generated bubble-mediated acoustic signals at combination frequencies. The results from sediment core analyses, and comparison of in situ acoustic velocity and attenuation values with those of water-saturated sediments, together provide ample evidence for the presence of in situ gas bubbles in the insonified volumes of sediments. These datasets are suitable for linear and non-linear inversion studies that estimate in situ greenhouse gas bubble populations, needed for future acoustical remote sensing applications.

We investigate the acoustic wave propagation in bubbly liquid inside a pilot sonochemical reactor which aims to produce antibacterial medical textile fabrics by coating the textile with ZnO or CuO nanoparticles. Computational models on acoustic propagation are developed in order to aid the design procedures. The acoustic pressure wave propagation in the sonoreactor is simulated by solving the Helmholtz equation using a meshless numerical method. The paper implements both the state-of-the-art linear model and a nonlinear wave propagation model recently introduced by Louisnard (2012), and presents a novel iterative solution procedure for the nonlinear propagation model which can be implemented using any numerical method and/or programming tool. Comparative results regarding both the linear and the nonlinear wave propagation are shown. Effects of bubble size distribution and bubble volume fraction on the acoustic wave propagation are discussed in detail. The simulations demonstrate that the nonlinear model successfully captures the realistic spatial distribution of the cavitation zones and the associated acoustic pressure amplitudes., (Copyright © 2015 Elsevier B.V. All rights reserved.)

The type and the complexity of regimen aimed at achieving better glycemic control may impact patient's health-related quality of life (HRQoL) in diabetic patients. But, the relationship between HbA1c levels of diabetic patients and their HRQoL is not clear. Our study aims to determine whether metabolic control variables can predict HRQoL or not and also the impact of hypertension (HT) on HRQoL in type II diabetic patients. A total of 469 patients with type II diabetes and 134 control subjects were studied. Medical Outcomes Study Short-Form-General Health Survey (SF-36) questionnaire was used as a health survey tool to measure the QoL of patients in the study. SF-36 includes 8 individual subscales and two summary scales (physical component summary [PCS] and mental component summary [MCS]). Age, gender, fasting blood glucose, postprandial blood glucose, HbA1c, high-density lipoprotein-cholesterol (HDL-C), low-density lipoprotein-cholesterol (LDL-C), triglyceride, total cholesterol, Apolipoprotein B (apoB), non-HDL-C, and body mass index values of the subjects were recorded. For statistical evaluation, SPSS (Statistical Package for the Social Sciences) 15 under Windows 7 was used. MCS values of patients group were statistically lower than control group (P < .05). There was no significant difference in PCS values between groups (P > .05). Diabetic patients with HT had significantly lower PCS and MCS values than those without HT. In addition, there was a negative correlation between HbA1c level and PCS and MCS values (P < .05). Hypertensive diabetic patients had significantly higher fasting blood glucose, postprandial blood glucose, HbA1c, HDL-C, LDL-C, total cholesterol, and body mass index values than hypertensive control subjects (P < .05). Normotensive diabetic patients also had significantly lower PCS value than normotensive control subjects (P < .05). But, MCS value was not different between groups (P > .05). PCS values in diabetic male patients were significantly higher than in diabetic female patients (P < .05). MCS value did not differ by gender in diabetic patients (P > .05). In our study, it is clear that diabetes affected the patients' HRQoL. In addition, we showed negative correlations between HbA1c levels and PCS and MCS values. There was a significant difference in PCS scores between genders in patients with diabetes. But, there was no significant difference in PCS and MCS values by age in diabetic patients. And having concomitant HT in diabetic patients causes a decrease in both MCS and PCS scores. Thus, HT is an important factor that should be considered in QoL of the diabetic patients., (Copyright © 2016 American Society of Hypertension. Published by Elsevier Inc. All rights reserved.)

We described the recent spatial distribution of rheumatoid arthritis in Turkey and assessed the role of environmental variables in this distribution. We developed an observed rheumatoid arthritis (RA) incidence grid map by using georeferenced rheumatoid arthritis case data (2011) from the centres of 81 provinces and the kriging method with a spherical variogram model in geographic information systems (GIS). We also modelled rheumatoid arthritis incidence in GIS by using complementary spatial database including the grid map layers of 14 environmental variables of Turkey. We conducted principle component analysis and multiple regression to investigate the relationships among variables and develop a model, respectively. The produced model was run in GIS to obtain a predicted (model) RA map. We tested the reliability of the model map by residual statistics and found the model map dependable. Observed and model incidence maps revealed the geographic distribution of rheumatoid arthritis cases in Turkey. The mean temperature, minimum temperature, maximum temperature, water vapour pressure, elevation, potential evapotranspiration, latitude, distance to seas, sunshine fraction, precipitation, longitude and aspect variables were found to have significant impacts on rheumatoid arthritis. Consequently, the model incidence map established a good background to predict rheumatoid arthritis cases following environmental changes.

We described the spatiotemporal change of malaria (Plasmodium vivax) in Turkey over 34 years (1975-2008), and assessed the role of environmental variables in this change. We developed seven 5-year-period raster maps by using geo-referenced malaria case data from the centres of 81 provinces and the kriging method with a spherical variogram model in a geographic information systems (GIS) model. We also modelled malaria incidence in GIS by using our average malaria incidence raster map, and complementary spatial database including the raster map layers of 14 environmental variables. We chose linear regression analysis with backward method to investigate relationships among variables and develop a model. The model was run in GIS to obtain a model incidence raster map. We tested the reliability of the model map by residual statistics, and found the model map dependable. Five-year-period maps revealed that the distribution of malaria cases moved from the East Mediterranean region to the Southeast Anatolia region due to changing human activities. The latitude, minimum temperature, distance to seas and elevation variables were found to have significant impacts on malaria. Consequently, the model incidence map established a good background for early warning systems to predict epidemics of malaria following environmental changes., (Copyright © 2010 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.)

Unlabelled: In this study we sought to determine the propofol requirement and hemodynamic effects as guided by bispectral index (BIS) analysis during induction of anesthesia. Sixty patients were enrolled in this study. Propofol, 2 mg/kg, was given to Group I for induction. Propofol was administered for induction until loss of response to verbal commands and until BIS values were around 50 to Groups II and III. After induction, the smallest BIS value was different in Group I. Decreases in total propofol dose were 36% and 43% in Groups II and III respectively as compared with Group I. The dose of propofol assessed by BIS analysis results in an important reduction of propofol requirement without side effects., Implications: Hypotension during induction of anesthesia with propofol is common. This study has shown that propofol requirement assessed by bispectral index analysis during anesthesia induction may decrease the dose and side effects and provide for satisfactory depth of anesthesia.