Multiple Readers Research Articles

Abstract TP84: Ring-like Pattern On Post-EVT MRI Is A Marker Of Perihematomal Edema

Introduction: Perihematomal edema is a consequence of primary intracerebral hemorrhage. On diffusion MRI, a corresponding heterogenous pattern of hypo- and hyper-intensity is visible on apparent diffusion coefficient (ADC), suggesting both cytotoxic and vasogenic edema. Analogously, endovascular therapy (EVT) for ischemic stroke can result in hemorrhagic transformation (HT) and a concomitant ring-like pattern, or “Rings” of mixed intensity on ADC. The objective of this study is to describe this pattern of secondary injury early after EVT. Methods: Patients were included if they consented to the prospective Natural History of Stroke GUARDS study from April 2018 to February 2022, were treated with EVT for anterior circulation LVO, achieved TICI 2b/3 reperfusion, and had early MRI, ~2h post-EVT. Two independent readers measured early and 24h DWI lesion volumes and 5d FLAIR lesion volumes using a semi-automated validated approach. Presence of any HT on GRE and Rings on ADC (Figure) on early, 24h, and 5d MRI were read independently by multiple readers who then reached consensus. The relationship of Rings with HT and lesion volume was explored. Results: Eighty-nine patients were included with median age 64y, 55% female, median admit NIHSS 18, 64% M1 LVO, and 47% IV tPA. On early MRI, 14 (16%) patients had Rings, 12 (86%) of whom also had HT (p<0.001). At 24h, 42 (50%) patients had Rings with 33 (79%) also having HT (p<0.001). There was a strong association between presence of Rings at 24h and 5d-FLAIR lesion volume, 56mL vs. 24mL in patients with vs. without Rings (p=0.002). Conclusions: Rings on ADC was seen in half of patients 24h post-EVT and appears to reflect peri-lesion edema associated with hemorrhagic transformation. This imaging marker was associated with larger lesion volumes at 5 days and may be able to serve as a target for adjunctive therapy post-EVT to mitigate edema formation.

Comparing ARIA‐E severity scales and effects of treatment management thresholds

AbstractBackgroundVasogenic edema and sulcal effusions detected with Magnetic Resonance Imaging (MRI), known as Amyloid‐Related Imaging Abnormalities‐Edema (ARIA‐E), are manageable side effects of anti‐amyloid immunotherapies. Periodic MRI scans with severity assessment of ARIA‐E are used in clinical trials or to guide practice with these therapies.Several scales with varying complexity measure radiological ARIA‐E severity to guide dosing. The ongoing GRADUATE studies use the 60‐point Barkhof Grand Total Scale (BGTS)1, continuing treatment for asymptomatic ARIA‐E ≤ 3, and temporarily suspending dosing for BGTS > 3. Other studies have used Simple ARIA‐E Severity Scales2 (SAESS‐3, SAESS‐5), continuing treatment for asymptomatic SAESS‐3 ≤ 1, but temporarily suspending dose for SAESS‐3 > 13. We previously reported good correlation between BGTS, SAESS‐3, and SAESS‐5 by a single reader. Here we report correlations among these scales scored by multiple readers and assess the impact of these various scores on ARIA management.MethodsT2‐FLAIR MRI scans were selected from a representative set of incident ARIA‐E cases (70 with previously detected ARIA‐E, 5 without ARIA‐E from original BGTS assessment) in the SCarlet RoAD (NCT01224106) and Marguerite RoAD (NCT02051608) Phase III studies. Three experienced neuroradiologists scored ARIA‐E severity using the 3 scales from pairs of baseline/follow‐up scans, blinded to prior read results. Median results of the 3 readers were computed for further analysis.ResultInter‐reader agreement for all 3 scales was high (ICC=0.866 [0.811, 0.908], 0.888 [0.841, 0.924], and 0.858 [0.776, 0.910] for the SAESS‐3, SAESS‐5, and BGTS scales respectively). Spearman’s rank correlation with BGTS scale was 0.88 for the SAESS‐3 scale and 0.90 for SAESS‐5. Median interquartile range (IQR) BGTS scores for the two simplified scales are shown in Figure 1. Best agreement for the SAESS‐3 > 1 treatment management threshold was achieved with BGTS > 3, with an accuracy of 93.3% (Figure 2).ConclusionsSAESS‐3 and SAESS‐5, which may be more suitable for routine clinical practice, are well correlated to the BGTS scale. Treatment management based on the SAESS‐3 > 1 threshold shows high agreement with the BGTS > 3 threshold.1. Barkhof et al., 2013; AJNR 34:8.2. Bracoud et al., 2017; AAIC. P1‐047.3. ADUHELM (aducanumab‐avwa) FDA PI.

Joint Interference Cancellation and Symbol Detection for Multistatic Backscatter System

Backscatter communication is a promising passive radio technique for the green Internet of Things (IoT) due to its low cost and low power consumption. However, the limited communication range is a challenging issue because of the double fading of the backscatter link. This paper proposes a multi-static backscatter system in which multiple readers illuminate a tag simultaneously to strengthen the received backscatter signals. In this system, direct links greatly impact the backscatter links, making it difficult to recover the tag information. The interference cancellation (IC) is exploited to cancel out the interference of the direct link and improve the detection probability. Then the expectation maximization (EM) algorithm is used for symbol detection, in which latent variable is introduced to form simpler distribution, and parameters are updated based on their probability. The numerical results verify the better performance of our proposed method.

Using Machine Learning Methods Incorporating Individual Reader Annotations to Classify Paediatric Chest Radiographs in Epidemiological Studies.

Introduction: Epidemiological studies that involve interpretation of chest radiographs (CXRs) suffer from inter-reader and intra-reader variability. Inter-reader and intra-reader variability hinder comparison of results from different studies or centres, which negatively affects efforts to track the burden of chest diseases or evaluate the efficacy of interventions such as vaccines. This study explores machine learning models that could standardize interpretation of CXR across studies and the utility of incorporating individual reader annotations when training models using CXR data sets annotated by multiple readers. Methods: Convolutional neural networks were used to classify CXRs from seven low to middle-income countries into five categories according to the World Health Organization's standardized methodology for interpreting paediatric CXRs. We compared models trained to predict the final/aggregate classification with models trained to predict how each reader would classify an image and then aggregate predictions for all readers using unweighted mean. Results: Incorporating individual reader's annotations during model training improved classification accuracy by 3.4% (multi-class accuracy 61% vs 59%). Model accuracy was higher for children above 12 months of age (68% vs 58%). The accuracy of the models in different countries ranged between 45% and 71%. Conclusions: Machine learning models can annotate CXRs in epidemiological studies reducing inter-reader and intra-reader variability. In addition, incorporating individual reader annotations can improve the performance of machine learning models trained using CXRs annotated by multiple readers.

Using Machine Learning Methods Incorporating Individual Reader Annotations to Classify Paediatric Chest Radiographs in Epidemiological Studies

Introduction: Epidemiological studies that involve interpretation of chest radiographs (CXRs) suffer from inter-reader and intra-reader variability. Inter-reader and intra-reader variability hinder comparison of results from different studies or centres, which negatively affects efforts to track the burden of chest diseases or evaluate the efficacy of interventions such as vaccines. This study explores machine learning models that could standardize interpretation of CXR across studies and the utility of incorporating individual reader annotations when training models using CXR data sets annotated by multiple readers. Methods: Convolutional neural networks were used to classify CXRs from seven low to middle-income countries into five categories according to the World Health Organization's standardized methodology for interpreting paediatric CXRs. We compared models trained to predict the final/aggregate classification with models trained to predict how each reader would classify an image and then aggregate predictions for all readers using unweighted mean. Results: Incorporating individual reader's annotations during model training improved classification accuracy by 3.4% (multi-class accuracy 61% vs 59%). Model accuracy was higher for children above 12 months of age (68% vs 58%). The accuracy of the models in different countries ranged between 45% and 71%. Conclusions: Machine learning models can annotate CXRs in epidemiological studies reducing inter-reader and intra-reader variability. In addition, incorporating individual reader annotations can improve the performance of machine learning models trained using CXRs annotated by multiple readers.

Myeloma Spine and Bone Damage Score (MSBDS) on Whole-Body Computed Tomography (WBCT): Multiple Reader Agreement in a Multicenter Reliability Study.

Objective: To assess the reliability of the myeloma spine and bone damage score (MSBDS) across multiple readers with different levels of expertise and from different institutions. Methods: A reliability exercise, including 104 data sets of static images and complete CT examinations of patients affected by multiple myeloma (MM), was performed. A complementary imaging atlas provided detailed examples of the MSBDS scores, including low-risk and high-risk lesions. A total of 15 readers testing the MSBDS were evaluated. ICC estimates and their 95% confidence intervals were calculated based on mean rating (k = 15), absolute agreement, a two-way random-effects model and Cronbach’s alpha. Results: Overall, the ICC correlation coefficient was 0.87 (95% confidence interval: 0.79–0.92), and the Cronbach’s alpha was 0.93 (95% confidence interval: 0.94–0.97). Global inter- and intra-observer agreement among the 15 readers with scores below or equal to 6 points and scores above 6 points were 0.81 (95% C.I.: 0.72–0.86) and 0.94 (95% C.I.:0.91–0.98), respectively. Conclusion: We present a consensus-based semiquantitative scoring systems for CT in MM with a complementary CT imaging atlas including detailed examples of relevant scoring techniques. We found substantial agreement among readers with different levels of experience, thereby supporting the role of the MSBDS for possible large-scale applications. Significance and Innovations • Based on previous work and definitions of the MSBDS, we present real-life reliability data for quantitative bone damage assessment in multiple myeloma (MM) patients on CT. • In this study, reliability for the MSBDS, which was tested on 15 readers with different levels of expertise and from different institutions, was shown to be moderate to excellent. • The complementary CT imaging atlas is expected to enhance unified interpretations of the MSBDS between different professionals dealing with MM patients in their routine clinical practice.

Human versus Artificial Intelligence–Based Echocardiographic Analysis as a Predictor of Outcomes: An Analysis from the World Alliance Societies of Echocardiography COVID Study

BackgroundTransthoracic echocardiography is the leading cardiac imaging modality for patients admitted with COVID-19, a condition of high short-term mortality. The aim of this study was to test the hypothesis that artificial intelligence (AI)–based analysis of echocardiographic images could predict mortality more accurately than conventional analysis by a human expert.MethodsPatients admitted to 13 hospitals for acute COVID-19 who underwent transthoracic echocardiography were included. Left ventricular ejection fraction (LVEF) and left ventricular longitudinal strain (LVLS) were obtained manually by multiple expert readers and by automated AI software. The ability of the manual and AI analyses to predict all-cause mortality was compared.ResultsIn total, 870 patients were enrolled. The mortality rate was 27.4% after a mean follow-up period of 230 ± 115 days. AI analysis had lower variability than manual analysis for both LVEF (P = .003) and LVLS (P = .005). AI-derived LVEF and LVLS were predictors of mortality in univariable and multivariable regression analysis (odds ratio, 0.974 [95% CI, 0.956-0.991; P = .003] for LVEF; odds ratio, 1.060 [95% CI, 1.019-1.105; P = .004] for LVLS), but LVEF and LVLS obtained by manual analysis were not. Direct comparison of the predictive value of AI versus manual measurements of LVEF and LVLS showed that AI was significantly better (P = .005 and P = .003, respectively). In addition, AI-derived LVEF and LVLS had more significant and stronger correlations to other objective biomarkers of acute disease than manual reads.ConclusionsAI-based analysis of LVEF and LVLS had similar feasibility as manual analysis, minimized variability, and consequently increased the statistical power to predict mortality. AI-based, but not manual, analyses were a significant predictor of in-hospital and follow-up mortality.

MRI evaluation of meniscal anatomy: which parameters reach the best inter-observer concordance?

PurposeThe aim of the study is to evaluate which MRI parameters achieve the best degree of inter-individual concordance in the description of meniscal fibrocartilage, regarding its morphology, signal and position.Materials and methodsEighty-nine knee MRIs were included in the study, retrospectively re-evaluated by three radiologists who completed a binary report (normal/abnormal) describing the meniscus signal, position relative to the tibial plateau margin and morphology. The inter-individual concordance value was calculated using Cohen's test.ResultsWe obtained different inter-individual concordance values according to the parameters considered. The concordance was poor in the description of the meniscal position relative to the tibial plateau margin (average k = 0.6); the result was comparable in the description of the meniscal morphology (average k = 0.56). The best results were obtained with the meniscal signal analysis (average k = 0.8).ConclusionTo the best of our knowledge, there are no studies in the literature assessing the concordance between multiple readers in the description of the parameters we studied. The results we obtained suggest that the most reliable parameter for describing meniscal fibrocartilage is its signal intensity, whereas morphology and position may lead to different interpretations that are not always unequivocal.

A Motion Artifact Correction Algorithm for Cone-Beam CT in Patients with Hepatic Malignancies Treated with Transarterial Chemoembolization

PurposeTo evaluate the effect of a motion artifact correction algorithm (MACA) on cone-beam computed tomography (CT) during transarterial chemoembolization for hepatic malignancies. Materials and MethodsFrom June 2020 to March 2021, 42 patients with mild-to-severe motion artifacts detected using single cone-beam CT scans were evaluated retrospectively. The image quality of native and motion-corrected data was compared. The maximum intensity, sharpness, and full width at half maximum (FWHM) of 5 segmental hepatic arteries were quantitatively measured. The overall quality of maximum intensity projection (MIP) images, conspicuity of tumor-supplying arteries, and need for selective angiography to ascertain the vascular anatomy were qualitatively evaluated by multiple readers. Paired t and Wilcoxon signed rank tests were used to compare the parameters. ResultsThe mean maximum intensity and sharpness increased from 2,792.01 HU ± 451.36 to 3,148.40 HU ± 594.46 and from 0.31 ± 0.02/mm to 0.34 ± 0.02/mm, respectively, using the MACA (both P < .001). The MACA decreased the mean FWHM from 2.02 mm ± 0.27 to 1.78 mm ± 0.26 (P < .001). The overall quality of the MIP images and the conspicuity of the tumor-supplying artery were enhanced from 2.5 to 3.0 points and from 3.0 to 4.0 points, respectively (both P < .001). Selective angiography was expected to be omitted in 7 cases (16.7%, 7/42) after using the MACA. ConclusionsThe MACA significantly improved both quantitative and qualitative image quality of cone-beam CT in selected patients with motion artifacts during transarterial chemoembolization for hepatic malignancies.

Sing-Ant: RFID Indoor Positioning System Using Single Antenna with Multiple Beams Based on LANDMARC Algorithm

RFID localization methods have been widely used in indoor positioning systems (IPS). Most localization techniques involve the use of multiple antennas and the placement of antennas and readers in order to ensure accurate positioning results. However, most localization techniques are complex and require high overhead costs in terms of needing multiple antennas and RFID readers. In this paper, we proposed a method to use a single antenna to perform all the reads and rely rather on the antenna beams to acquire multiple positioning data. A single array of antennas is configured to have multiple angles of operation and rely on different power levels as compared to regular antennas. By manipulating the beam pattern, direction and power, multiple sub-antennas can be conceived and the method utilizes antenna beams and relies mainly on one antenna to realize two-dimensional localization.

UHF-RFID-Based Real-Time Vehicle Localization in GPS-Less Environments

The vehicle localization, which aims to identify a vehicle and then position the vehicle with a high precision, can be used to facilitate various applications and services in vehicular networks. Unfortunately, conventional localization systems, e.g., global positioning system (GPS), hardly meet the accuracy requirements especially in certain specific scenarios, such as tunnels. At the same time, Ultrahigh frequency (UHF) radio frequency identification (RFID) has become an efficient booster for internet of things (IoT) due to the desirable advantages, such as low cost, battery-free, and unique identification. In this paper, based on the UHF-RFID, we propose a novel real-time vehicle localization scheme in GPS-Less Environments. Considering the practical implementation of multiple RFID reader antennas on a vehicle is constrained, we adopt single antenna multi-frequency ranging scheme, in which the integer ambiguity problem is solved by the maximum-likelihood estimation (MLE)-based robust Chinese remainder theorem (CRT). With the reconstructed distances between the tags and the reader, the coordinates of the vehicle then can be calculated with the Levenberg-Marquardt (LM) algorithm. Furthermore, the computational complexities of the algorithms and the time consumption of the proposed scheme are analyzed. The experimental results demonstrate that the proposed scheme can track vehicle’s location with error lower than 27 cm at the probability of 90%.

Reduction in Chest CT Severity and Improved Hospital Outcomes in SARS-CoV-2 Omicron Compared with Delta Variant Infection.

BackgroundThe SARS-Cov-2 Omicron variant demonstrates rapid spread but with reduced disease severity. Studies evaluating the lung imaging findings of Omicron infection versus non-Omicron variants remain lacking.PurposeTo compare Omicron and Delta variants of SARS-CoV-2 by their chest CT radiological pattern, biochemical parameters, clinical severity and hospital outcomes after adjusting for vaccination status.Materials and MethodsRetrospective study of hospitalized adult patients rt-PCR positive for SARS-CoV-2 with CT pulmonary angiography performed within 7 days of admission between December 1, 2021 and January 14, 2022. Blinded radiological analysis with multiple readers including RSNA CT classification, chest CT severity score (CT-SS, range 0 least severe to 25 most severe) and CT imaging features including bronchial wall thickening.Results106 patients (Delta n=66, Omicron n=40) were evaluated (mean age, 58 years ± 18, 58 men). In the Omicron group, 37% (15/40) of CT pulmonary angiograms were categorized as normal compared with 15% (10/66) in the Delta group (p=.016). Using a generalized linear model to control for confounding variables, including vaccination status, Omicron variant infection was associated with a CT-SS that was lower by 7.2 points compared to infection with Delta variant (β=-7.2, 95%CI: -9.9, -4.5; p <.001). Bronchial wall thickening was more common with Omicron than with the Delta variant (odds ratio [OR] 2.4, 95%CI: 1.01, 5.92, p=.04). Vaccination with a booster shot was associated with a protective effect on chest infection compared with the unvaccinated (CT-SS median 5 (IQR 0-11), CT-SS median 11 (IQR 7.5-14), respectively; p = .03). The Delta variant was associated with a higher odds ratio of severe disease (OR 4.6, 95%CI: 1.2, 26, p=.01) and critical care admission (OR 7.0, 95%CI: 1.5, 66, p=.004) than the Omicron variant.ConclusionThe SARS-COV-2 Omicron variant was associated with fewer and less severe changes on chest CT compared with the Delta variant. Patients with Omicron had greater frequency of bronchial wall thickening but lower clinical severity and improved hospital outcomes than those with Delta.

A High Time-Efficient Missing Tag Detection Scheme for Integrated RFID Systems.

Missing tag incidents are common in RFID-enabled supply-chain and warehousing scenarios due to cargo theft and employee error operations, which may lead to serious economic losses or potential safety hazards. On the premise of ensuring the accuracy of missing tag detection, this paper aims to improve the time efficiency in an integrated RFID system. Unlike prior work focusing on detecting missing items from a large number of homogeneous tags that are monitored by a single reader, one integrated RFID system possesses multiple readers to communicate with the heterogeneous tags, which have different categorical attributes. In addition, the prior work required repeating the execution several times to capture the missing tags in assorted categories, which is of low time efficiency. Thus, a protocol called Multi-reader Missing Tag Detection (MMTD) is proposed to capture the missing tag quickly and reliably, which can detect missing tags from different categories in a parallel manner and is much more time-efficient than previous work. MMTD has two major advantages compared to prior work: (i) It leverages the knowledge of the spatial distribution of tags to divide up a difficult detection task into several lightweight tasks, which are shared by multiple readers. (ii) It personalizes the time frame of the reader based on the tag population to optimize the utilization of the communication channel. The final simulation results reveal that MMTD is the best in time-efficiency among the comparison protocols, and MMTD outperforms the other missing tag detection protocols by at least in the Integrated RFID scenarios.

Reliability of Magnetic Resonance Imaging Interpretation of Lateral Discoid Meniscus: A Multicenter Study

Background:Lateral discoid meniscus (LDM) often present with complex morphology that can be challenging to assess and treat arthroscopically. Preoperative magnetic resonance imaging (MRI) is frequently used for diagnosis and surgical planning, however it is not known whether surgeons are reliable and consistent in their interpretation of MRI findings.Hypothesis/Purpose:We hypothesize that surgeons experienced in treating LDM are able to reliably interpret discoid pathology using MRI, with the exception of evaluating dynamic factors (i.e. instability).Methods:Forty five discoid meniscus MRI selected from a pool of surgical discoid meniscus cases were included in this review. Five reviewers (experienced pediatric sports medicine surgeons) who were not involved in the MRI selection process performed independent review of each MRI to determine discoid meniscus classification. More than 4 weeks later, a second reading was performed by 3 of the 5 reviewers. Interobserver and intraobserver reliability of the primary (width, presence of instability or tear) and secondary (location of instability or tear, tear type) rating factors was assessed using the Fleiss κ coefficient, designed for multiple readers with nominal variables (reliability: fair, 0.21–0.40; moderate, 0.41–0.60; substantial, 0.61–0.80; excellent, 0.81–1.00). Reliability is reported as κ (95% CI).Results:Interobserver reliability of assessment of meniscal width was substantial 0.67 (0.58-0.76), and intraobserver was moderate 0.52 (0.35-0.69). Assessment of presence of peripheral instability had moderate interobserver reliability 0.54 (0.45-0.63) and substantial intraobserver 0.61 (0.44-0.78). Presence of tear had fair interobserver reliability 0.39 (0.29–0.48) and substantial intraobserver 0.68 (0.51-0.85). When identifying location of instability or tear, interobserver reliability was moderate while intraobserver agreement was substantial for assessment of anterior instability (0.47 (0.37-0.56) and 0.68 (0.51-0.85), respectively); posterior instability (0.56 (0.47-0.65) and 0.62 (0.45-0.79), respectively); and posterior tear (0.41 (0.32-0.50) and 0.69 (0.52-0.86), respectively). The exception was identifying anterior tear, with fair interobserver reliability 0.33 (0.23-0.42) and moderate intraobserver 0.56 (0.39-0.73); Interobserver reliability was fair for tear type 0.34 (0.28-0.41) and intraobserver reliability moderate 0.55 (0.43-0.67).Conclusion:Orthopaedic surgeons experienced in the treatment of LDM vary from each other in their classification using MRI, especially with regard to assessment of meniscus tears. MRI evaluation may be helpful to diagnose discoid by width and identify presence of instability, two major factors in the decision to proceed with surgery. However, definitive treatment should be guided by a comprehensive arthroscopic evaluation.

A Novel Spatio-Temporal Data Redundancy Elimination Approach for RFID Systems with Multiple Readers

In this paper, we propose a Voronoi diagram based spatio-temporal data redundancy elimination approach for RFID systems having multiple readers so that every RFID tag will be read by only one reader depending on the distance between the tag and the center of the Minimum Enclosing Circle (MEC) of the Voronoi cell which the reader belongs to. Thus, the spatial redundancy in the gathered data is completely eliminated. Temporal redundancy is also minimized by reading the RFID tags at regular time intervals larger than an appropriately chosen threshold value. Existing methods of reducing data redundancy in RFID systems are often associated with loss of data, false positive errors and false negative errors. In contrast to this, our proposed technique is free from any false positive and false negative errors with no loss of data and every tag being read by only one single reader. Simulation of our proposed approach also establishes its superiority to the existing techniques in terms of these performance parameters.

Individualized and Generalized Learner Models for Predicting Missed Hepatic Metastases.

The diagnostic performance of radiologist readers exhibits substantial variation that cannot be explained by CT acquisition protocol differences. Studying reader detectability from CT images may help identify why certain types of lesions are missed by multiple or specific readers. Ten subspecialized abdominal radiologists marked all suspected metastases in a multi-reader-multi-case study of 102 deidentified contrast-enhanced CT liver scans at multiple radiation dose levels. A reference reader marked ground truth metastatic and benign lesions with the aid of histopathology or tumor progression on later scans. Multi-slice image patches and 3D radiomic features were extracted from the CT images. We trained deep convolutional neural networks (CNN) to predict whether an average (generalized) or individual radiologist reader would detect or miss a specific metastasis from an image patch containing it. The individualized CNN showed higher performance with an area under the receiver operating characteristic curve (AUC) of 0.82 compared to a generalized one (AUC = 0.78) in predicting reader-specific detectability. Random forests were used to build the respective versions from radiomic features. Both the individualized (AUC = 0.64) and generalized (AUC = 0.59) predictors from radiomic features showed limited ability to differentiate detected from missed lesions. This shows that CNN can identify and learn automated features that are better predictors of reader detectability of lesions than radiomic features. Individualized prediction of difficult lesions may allow targeted training of idiosyncratic weaknesses but requires substantial training data for each reader.

MATLAB toolbox for ROC analysis of multi-reader multi-case diagnostic imaging studies.

A common study design for comparing the performances of diagnostic imaging tests is to obtain ratings from multiple readers of multiple cases whose true statuses are known. Typically, there is overlap between the tests, readers, and/or cases for which special analytical methods are needed to perform statistical comparisons. We present our new MATLAB MRMCaov toolbox, which is designed for multi-reader multi-case comparisons of two or more diagnostic tests. The toolbox allows for statistical comparison of reader performance metrics, such as area under the receiver operating characteristic curve (ROC AUC), with analysis of variance methods originally proposed by Obuchowski and Rockette (1995) and later unified and improved by Hillis and colleagues (2005, 2007, 2008, 2018). MRMCaov is open-source software with an integrated command-line interface for performing multi-reader multi-case statistical analysis, plotting, and presenting results. Its features (1) ROC AUC, likelihood ratios of positive or negative ratings, sensitivity, specificity, and expected utility reader performance metrics; (2) reader-specific ROC curves; (3) user-definable performance metrics; (4) test-specific estimates of mean performance along with confidence intervals and p-values for statistical comparisons; (5) support for factorial, nested, or partially paired study designs; (6) inference for random or fixed readers and cases; (7) DeLong, jackknife, or unbiased covariance estimation; and (8) compatibility with Microsoft Windows, Mac OS, and Linux.

CNN-Based Machine Learning Channel on TDMR Drive Data

In this article, we demonstrate a convolutional neural network (CNN)-based data detection channel on real data from a commercial hard disk drive (HDD) employed with two-dimensional magnetic recording (TDMR) dual-reader technology. The neural network has two sets of input neurons and is trained using 485 4-kB sectors from the drive. Data detection is performed using the trained network and a one-to-one comparison of bit-error-rate (BER) performance with the HDD channel is systematically conducted. The trained neural network demonstrates comparable or slightly better performance than the state-of-the-art HDD channel. In addition, comparing to detection with a single reader, CNN-machine learning (ML) channel autonomously enables ~6% recording density gain when the second reader is available, and there is no physical modeling needed to manually combine or “fuse” signals from the multiple readers. We believe that the understanding elucidated in this article is crucial for developing more capable, robust, and effective ML-based data detection channels with the potential to realize even greater areal density gain in TDMR systems and higher capacity HDD.

The Effects of Ibuprofen, Naproxen and Diclofenac on cell Apoptosis, Cell Proliferation and Histology Changes in Human Cholangiocarcinoma Cell Lines.

Objective:To examine the effects of ibuprofen, naproxen and diclofenac, non-steroidal anti-inflammatory drugs (NSAIDs) on cell proliferation activity of the human CCA cell lines. Methods:KKU-M139 and KKU-213B cell lines were used in this study. The cell viability was assessed by the MTT assay. Lipid synthesis determined by Oil red O staining and colorimetric assay. An inverted phase-contrast light microscope was used to investigate the histological change of the cells. Caspases 3/7 activity and AnnexinV/PI were used to assess apoptosis by multiple microplate reader. Results:The results showed that ibuprofen, naproxen and diclofenac suppressed the viability of the KKU-M139 and KKU-213B cells in a dose-dependent manner, as measured especially diclofenac. However, these three NSAIDs slightly decreased lipid synthesis determined by Oil red O staining and colorimetric assay. The histological change observations showed the shrinking cell and become star-shaped in high dose treated groups. Interestingly, these NSAIDs exhibited in both of KKU-M139 and KKU-213B cell lines, the diclofenac-treated cells had the most injury cells. The cells exhibited cell injury features. In addition, the detection of caspase 3/7 and AnnexinV/PI in this investigation revealed early cell apoptotic characteristics. Conclusion:These finding suggest that ibuprofen, naproxen and diclofenac suppress cell viability. The results reveal that ibuprofen, naproxen and diclofenac, which induce the histological change and apoptosis. This study indicates that these NSAIDs may be used as an anti-proliferation agent for the treatment of CCA in the future.

Dialogic interaction with diversified audiences in Twitter for Research Dissemination Purposes

International research groups are expected to ensure global dissemination and visibility of their knowledge production, for which Twitter is effectively employed to reach diversified audiences. This paper analyses the dialogic dimension of tweets published in accounts of Horizon2020 research projects, where group’s productivity and work are promoted, and multiple readers addressed. Our study focuses on the use, in these Twitter accounts, of interactional pragmatic strategies, their verbal realisation through engagement markers, as well as on medium affordances and non-verbal markers. A sample of 1.454 tweets from 10 accounts of the EUROPROtweets corpus were coded and analysed through NVivo. The data-driven pragmatic analysis triggered the identification of 8 interactional strategies. We then quantitatively analysed the use of engagement makers and qualitatively studied the characteristic non-verbal markers with a dialogic function within each of these. Our findings will help understand the complexities of current digital academic professional practices, especially as regards the dynamics of dialogic interaction in social media.