Commercial Platforms Research Articles

Short-Term Load Forecasting of the Greek Electricity System

Short-term load forecasting is an essential instrument in power system planning, operation, and control. It is involved in the scheduling of capacity dispatch, system reliability analysis, and maintenance planning for turbines and generators. Despite the high level of development of advanced types of machine learning models in commercial codes and platforms, the prediction accuracy needs further improvement, especially in certain short, problematic time periods. To this end, this paper employs public domain electric load data and typical climatic data to make 24-hour-ahead hourly electricity load forecasts of the Greek system based on two types of robust, standard feed-forward artificial neural networks. The accuracy and stability of the prediction performance are measured by means of the modeling error values. The current prediction accuracy levels of mean absolute percentage error, mean value μ = 2.61% with σ = 0.33% of the Greek system operator for 2022, attained with noon correction, are closely matched with a simple feed-forward artificial neural network, attaining mean value μ = 3.66% with σ = 0.30% with true 24-hour-ahead prediction. Specific instances of prediction failure in cases of unexpectedly high or low energy demand are analyzed and discussed. The role of the structure and quality of input data of the training datasets is demonstrated to be the most critical factor in further increasing the accuracy and reliability of forecasting.

Social Media Use Among Parents and Women of Childbearing Age in the US.

Many parents and pregnant women in the US use social media to access health-related information. Estimates of current use of different platforms among these populations are needed. We used data from a 2021 Pew Research Center survey to describe use of commercial social media platforms by US parents and US women aged 18 to 39 years. Most US parents and women of childbearing age use YouTube, Facebook, and Instagram, with most engaging daily. Understanding social media use patterns can help public health professionals, health care systems, and researchers reach selected populations with evidence-based health information and health promotion programs.

Moving intra-individual variability (IIV) towards clinical utility: IIV measured using a commercial testing platform

ObjectivesIntra-individual variability (IIV), measured across repeated response times (RT) during continuous psychomotor tasks, is an early marker of cognitive change in the context of neurodegeneration. To advance IIV towards broader application in clinical research, we evaluated IIV from a commercial cognitive testing platform and compared it to the calculation approaches used in experimental cognitive studies. MethodsCognitive assessment was administered in participants with multiple sclerosis (MS) during the baseline of an unrelated study. Cogstate was used for computer-based measures providing three timed-trial tasks measuring simple (Detection; DET) and choice (Identification; IDN) RT and working memory (One-Back; ONB). IIV for each task was automatically output by the program (calculated as a log10-transformed standard deviation or “LSD”). We calculated IIV from the raw RTs using coefficient of variation (CoV), regression-based, and ex-Gaussian methods. The IIV from each calculation was then compared by rank across participants. ResultsA total of n = 120 participants with MS aged 20–72 (Mean ± SD, 48.99 ± 12.09) completed the baseline cognitive measures. For each task, the interclass correlation coefficient was generated. Each ICC showed that LSD, CoV, ex-Gaussian, and regression methods clustered strongly (Average ICC for DET: 0.95 with 95% CI [0.93, 0.96]; Average ICC for IDN: 0.92 with 95% CI [0.88 to 0.93]; Average ICC for ONB: 0.93 with 95% CI [0.90 to 0.94]). Correlational analyses indicated the strongest correlation between LSD and CoV for all tasks (rs ≥ 0.94). ConclusionThe LSD was consistent with research-based methods for IIV calculations. These findings support the use of LSD for the future measurement of IIV for clinical studies.

THE PHENOMENON OF POST-INTERNET ART: ARTISTIC STRATEGIES AND PLACE IN THE FIELD OF CONTEMPORARY ART

The study is devoted to the phenomenon of post-Internet art, its main representatives and key works, and a description of their characteristics. In contrast to the practices of net art, post-Internet art is a “return to the material”: artists create objects inspired by the network and often bearing the “mark” of network aesthetics. However, these objects have physical characteristics; they evoke a tactile sensation in the viewer, serving as a spontaneous affirmation of materiality in our increasingly virtualised world. Post-Internet art is a little-studied area of contemporary art; the emergence of this trend is associated with the development of net-art practices (network art), which have offered new forms of interaction with viewers who got the opportunity to join the artistic process as full participants. The term “post-Internet” should be understood not as a situation “after” the Internet (as if the Internet “ended”) but as a situation after some significant event that changed the usual course of things. “Post-Internet” is a situation after the Internet has ceased to be perceived in a romantic way, as a space of anonymity and permissiveness, “undeveloped land” and is now understood as part of an online-offline “bundle” - with appropriate communication channels, creation of commercial platforms, digital currencies and etc., that is, as part of vernacular practices (everyday experience). From the point of view of its orientation, post-Internet art deals with understanding the consequences of the influence of Internet practices on contemporary art and culture. Net art artists created such works that were intended for the Internet, that is, used the possibilities of the Internet as a means of expression (they studied the structure of the Internet, the features of the code, the specifics of certain browsers, programs, etc.), whereas post-Internet artists seem to “bring” digital objects from the digital environment into the material world and use both online and offline formats for reflection on the effects of almost ubiquitous networking. In the 2010s, the Internet was no longer perceived as a “new” medium; it became more difficult to talk about post-Internet art as a specific movement in art since more and more aspects of our daily life are structurally determined by the World Wide Web, including art as a whole. Of course, the Internet continues to interest artists as a kind of socio-cultural state; however, discourses related to the Internet start to be increasingly woven into a wider range of issues - digital capitalism, ecology, social issues, etc. Therefore, the “marking” of art created under the influence of the Internet or referring to it in one way or another as “post-Internet art” is becoming less and less common, which allows us to speak about the specific temporal boundaries of post-Internet art as an art movement.

Evaluation of pooling strategy of SARS-CoV-2 RT-PCR in limited resources setting in Egypt at low prevalence.

Sample pooling testing for SARS-COV-2 can be an effective tool in COVID-19 screening when resources are limited, yet it is important to assess the performance before implementation as pooling has its limitations. Our objective was to assess the efficacy of pooling samples for coronavirus 2019 (COVID-19) compared to an individual analysis by using commercial platforms for nucleic acid testing. A total of 2200 nasopharyngeal swabs for SARS-COV-2 were tested individually and in pools of 4, 8, and 10. The cycle threshold (Ct) values of the positive pooled samples were compared to their corresponding individual positive samples. In pool size 10 samples, an estimated increase of 3-Ct was obtained, which led to false negative results in low viral load positive samples. Pooling SARS COV-2 samples is an effective strategy of screening to increase laboratories' capacity and reduce costs without affecting diagnostic performance. A pool size of 8 is recommended.

Deep Learning Recommendations of E-Education Based on Clustering and Sequence

Commercial e-learning platforms have to overcome the challenge of resource overload and find the most suitable material for educators using a recommendation system (RS) when an exponential increase occurs in the amount of available online educational resources. Therefore, we propose a novel DNN method that combines synchronous sequences and heterogeneous features to more accurately generate candidates in e-learning platforms that face an exponential increase in the number of available online educational courses and learners. Mitigating the learners’ cold-start problem was also taken into consideration during the modeling. Grouping learners in the first phase, and combining sequence and heterogeneous data as embeddings into recommendations using deep neural networks, are the main concepts of the proposed approach. Empirical results confirmed the proposed solution’s potential. In particular, the precision rates were equal to 0.626 and 0.492 in the cases of Top-1 and Top-5 courses, respectively. Learners’ cold-start errors were 0.618 and 0.697 for 25 and 50 new learners.

Measurement of LAG-3 Expression Across Multiple Staining Platforms With the 17B4 Antibody Clone.

An immunohistochemistry (IHC) assay developed to detect lymphocyte-activation gene 3 (LAG-3), a novel immune checkpoint inhibitor target, has demonstrated high analytic precision and interlaboratory reproducibility using a Leica staining platform, but it has not been investigated on other IHC staining platforms. To evaluate the performance of LAG-3 IHC assays using the 17B4 antibody clone across widely used IHC staining platforms: Agilent/Dako Autostainer Link 48 and VENTANA BenchMark ULTRA compared to Leica BOND-RX (BOND-RX). Eighty formalin-fixed, paraffin-embedded melanoma tissue blocks were cut into consecutive sections and evaluated using staining platform-specific IHC assays with the 17B4 antibody clone. Duplicate testing was performed on the BOND-RX platform to assess intraplatform agreement. LAG-3 expression using a numeric score was evaluated by a pathologist and with a digital scoring algorithm. LAG-3 positivity was determined from manual scores using a 1% or greater cutoff. LAG-3 IHC staining patterns and intensities were visually similar across all 3 staining platforms. Spearman and Pearson correlations were 0.75 or greater for interplatform and BOND-RX intraplatform concordance when LAG-3 expression was evaluated with a numeric score determined by a pathologist. Correlation increased with a numeric score determined with a digital scoring algorithm (Spearman and Pearson correlations ≥0.88 for all comparisons). Overall percentage agreement was 77.5% or greater for interplatform and BOND-RX intraplatform comparisons when LAG-3 positivity was determined using a 1% or greater cutoff. Data presented here demonstrate that LAG-3 expression can be robustly and reproducibly assessed across 3 major commercial IHC staining platforms using the 17B4 antibody clone.

Nanopore sequencing for N1-methylpseudouridine in RNA reveals sequence-dependent discrimination of the modified nucleotide triphosphate during transcription.

Direct RNA sequencing with a commercial nanopore platform was used to sequence RNA containing uridine (U), pseudouridine (Ψ)or N1-methylpseudouridine (m1Ψ) in>100 different 5-nucleotide contexts. The base calling data for Ψ or m1Ψ were similar but different from U allowing their detection. Understanding the nanopore signatures for Ψ and m1Ψ enabled a running start T7 RNA polymerase assay to study the selection of UTP versus ΨTP or m1ΨTP competing mixtures in all possible adjacent sequence contexts. A significant sequence context dependency was observed for T7 RNA polymerase with insertion yields for ΨTP versus UTP spanning a range of 20-65%, and m1ΨTP versus UTP producing variable yields that differ by 15-70%. Experiments with SP6 RNA polymerase, as well as chemically-modified triphosphates and DNA templates provide insight to explain the observations. The SP6 polymerase introduced m1ΨTP when competed with UTP with a smaller window of yields (15-30%) across all sequence contexts studied. These results may aid in future efforts that employ RNA polymerases to make therapeutic mRNAs with sub-stoichiometric amounts of m1Ψ.

Gear Shift Quality Enhancement Parameters in Light Commercial Vehicles

In the present scenario enhancements of human ergonomics in Light commercial vehicles are imperative to implement customer comfort in vehicle level. Thus the gear shift mechanism plays an important role in the interface between the driver and vehicle. An exact and smooth running gear shift feel typically depends on the combination of both internal and external shifting elements of the gearbox and Change shift operation(CSO) system. The objective of this paper is to explain the parameters that impacts the gear shift quality and how it is improved by taking multiple design iterations in Light commercial vehicle platform through both objective and subjective evaluations. These iterations are given better results on gear shift feel in dynamic as well as static condition. This lead us to facilitate better shift comfort to end users. The content of the paper is limited to the scope of gear shift quality improvements at manual transmission and change shift operation components.

Structuring Cyber-Physical Systems for Distributed Control with IEC 61499 Standard

With the constant growth in size and complexity of modern industrial applications, it is not suitable to use a unique and centralized controller device for the whole process. Thus, automation solutions need to be designed by distributing control execution across different devices. Cyber-Physical Systems (CPSs) are designed to solve networked embedded systems complexities. The IEC 61499 standard fosters developing distributed control applications by defining a platform-independent application model that can be mapped to numerous hardware components. Therefore, a relevant research topic is how to solve the design complexities of industrial CPSs by applying modeling techniques. In that context, this paper proposes an IEC 61499-based model for CPSs to distribute the complexity of control software over numerous small devices. The approach enables the creation of a comprehensive structure that can combine design, simulation, and distributed deployment of automation software. The proposed scheme is validated through a packet sorting system implemented in a commercial IDE platform.

ADAMTS13 activity testing: evaluation of commercial platforms for diagnosis and monitoring of thrombotic thrombocytopenic purpura

BackgroundADAMTS13 activity is one of the key investigations needed to diagnose thrombotic thrombocytopenic purpura, and there are a number of different assays available to measure it. HemosIL AcuStar, a chemiluminescent immunoassay, was developed and used as a quicker, automated test. In clinical practice, discrepancies between AcuStar and the gold standard FRETS-VWF73 have been documented in a manner that would affect diagnosis and treatment. ObjectivesWe aimed to identify and highlight clinical situations where this discrepancy occurs and to attempt to determine the cause. MethodTherefore, we undertook a study to compare the FRETS-VWF73 assay with AcuStar, the Technozym Activity ELISA, and Ceveron FRET assays using a mixture of 94 retrospective and prospective patient samples. ResultsWe found that although the concordance between FRETS-VWF73 and the other methods was generally very good, discrepancies were found in a small number tested on AcuStar affecting diagnosis (5 of 32) and follow-up (7 of 51). A Wilcoxon test comparing FRETS-VWF73 to the AcuStar results suggested that the AcuStar results were significantly lower in 42 samples tested on all 4 platforms. We investigated potential causes for this difference by testing the impact of high vWF levels and addition of a monoclonal ADAMTS13 autoantibody (3H9) to samples. We found no impact of high vWF levels on interassay variability but found that 3H9 reduced ADAMTS13 activity levels much more in AcuStar and ELISA assays than in FRETS assays. ConclusionBased on our findings, we would suggest that when AcuStar is used upfront to guide management, a second testing method should be used in patients with an atypical thrombotic thrombocytopenic purpura presentation or unexpectedly slow ADAMTS13 recovery.

MaxATAC: Genome-scale transcription-factor binding prediction from ATAC-seq with deep neural networks.

Transcription factors read the genome, fundamentally connecting DNA sequence to gene expression across diverse cell types. Determining how, where, and when TFs bind chromatin will advance our understanding of gene regulatory networks and cellular behavior. The 2017 ENCODE-DREAM in vivo Transcription-Factor Binding Site (TFBS) Prediction Challenge highlighted the value of chromatin accessibility data to TFBS prediction, establishing state-of-the-art methods for TFBS prediction from DNase-seq. However, the more recent Assay-for-Transposase-Accessible-Chromatin (ATAC)-seq has surpassed DNase-seq as the most widely-used chromatin accessibility profiling method. Furthermore, ATAC-seq is the only such technique available at single-cell resolution from standard commercial platforms. While ATAC-seq datasets grow exponentially, suboptimal motif scanning is unfortunately the most common method for TFBS prediction from ATAC-seq. To enable community access to state-of-the-art TFBS prediction from ATAC-seq, we (1) curated an extensive benchmark dataset (127 TFs) for ATAC-seq model training and (2) built "maxATAC", a suite of user-friendly, deep neural network models for genome-wide TFBS prediction from ATAC-seq in any cell type. With models available for 127 human TFs, maxATAC is the largest collection of high-performance TFBS prediction models for ATAC-seq. maxATAC performance extends to primary cells and single-cell ATAC-seq, enabling improved TFBS prediction in vivo. We demonstrate maxATAC's capabilities by identifying TFBS associated with allele-dependent chromatin accessibility at atopic dermatitis genetic risk loci.

Extract-Shaped Immune Repertoires as Source for Nanobody-Based Human IgE in Grass Pollen Allergy.

The presence of allergen-specific IgE in serum is a biomarker for allergic disease. Specific IgE antibodies for research and diagnostics, however, remain scarce. In contrast to prototypic antibodies, camelid species have evolved single domains as moiety for antigen recognition. These so-called nanobodies represent a versatile platform for the development of diagnostic and therapeutic approaches. In this study, we aimed for generating nanobodies and derived IgE formats from an extract-shaped immune repertoire. Timothy grass pollen represents a complex, but well-defined mixture of individual allergens. Therefore, a repertoire library from a timothy grass pollen extract immunised llama was established. The selection by phage display yielded 3 nanobodies with immunoreactivity to the extract. IgE-like nanobody-based human IgE (nb-hIgE) antibodies were produced in mammalian cells and assessed in different immunoassays and commercial platforms. Immunoblotting and diagnostic ImmunoCap analysis of single timothy grass pollen allergens identified the major allergens Phl p 6 and Phl p 4 as targets. Assessment of immunoreactivity further documented significant molecular cross-reactivity with pollen extract of different grass species and variant presence of allergens within extracts of Pooideae grasses. In summary, our study shows that extract-based immunisation enables the generation of allergen-specific nanobodies and derived nb-hIgE formats linking nanobody technologies with allergological applications.

Migration-Enhanced Metal–Organic Chemical Vapor Deposition of Wafer-Scale Fully Coalesced WS2 and WSe2 Monolayers

Metal–organic chemical vapor deposition (MOCVD) is widely employed for the wafer-scale synthesis of transition metal dichalcogenide (TMDC) monolayers (MLs). Despite large efforts devoted to understanding the intricate nucleation and lateral growth mechanisms of TMDCs, little attention has been paid to the migration of adatoms on the top of an ML and its influence on parasitic/premature bilayer (BL) nucleation. In this work, using a commercial multi-wafer MOCVD platform, a novel two-stage migration-enhanced MOCVD process is introduced to realize the deposition of wafer-scale fully coalesced tungsten disulfide (WS2) and tungsten diselenide (WSe2) MLs with only sparse BL nucleation in a reasonable deposition time. With the WS2 ML coverage exceeding 99% on 2 in. sapphire substrates within 3 h, BL coverage is suppressed to ∼15%. Following the same migration enhancement approach, WSe2 MLs are synthesized in 90 min with <20% BL coverage. The migration of W adatoms on the already formed stable WS2 (or WSe2) ML domains is promoted by ramping down the delivery of the tungsten precursor. From the qualitative analysis of the nanomorphology, the migration length of W adatoms is estimated to be ≤100 nm. This approach can be seen as a reliable and solid basis for the development of future large-scale TMDC ML deposition techniques.

Comparison of Five Commercial Molecular Assays for Mycoplasma Testing of Cellular Therapy Products.

Testing of cellular therapy products for Mycoplasma is a regulatory requirement by the United States Food and Drug Administration (FDA) to ensure the sterility and safety of the product prior to release for patient infusion. The risk of Mycoplasma contamination in cell culture is high. Gold standard testing follows USP 63 which requires a 28-day agar and broth cultivation method that is impractical for short shelf-life biologics. Several commercial molecular platforms have been marketed for faster raw material and product release testing; however, little performance data are available in the literature. In this study, we performed a proof-of-principle analysis to evaluate the performance of five commercial molecular assays, including the MycoSEQ Mycoplasma detection kit (Life Technologies), the MycoTOOL Mycoplasma real-time detection kit (Roche), the VenorGEM qOneStep kit (Minerva Biolabs), the ATCC universal Mycoplasma detection kit, and the Biofire Mycoplasma assay (bioMérieux Industry) using 10 cultured Mollicutes spp., with each at four log-fold dilutions (1,000 CFU/mL to 1 CFU/mL) in biological duplicates with three replicates per condition (n = 6) to assess limit of detection (LOD) and repeatability. Additional testing was performed in the presence of tumor infiltrating lymphocytes (TILs). Based on LOD alone, the Biofire Mycoplasma assay was most sensitive followed by the MycoSEQ and MycoTOOL which were comparable. We showed that not all assays were capable of meeting the ≤10 CFU/mL LOD to replace culture-based methods according to European and Japanese pharmacopeia standards. No assay interference was observed when testing in the presence of TILs.

Benchmarking of two bioinformatic workflows for the analysis of whole-genome sequenced Staphylococcus aureus collected from patients with suspected sepsis

BackgroundThe rapidly growing area of sequencing technologies, and more specifically bacterial whole-genome sequencing, could offer applications in clinical microbiology, including species identification of bacteria, prediction of genetic antibiotic susceptibility and virulence genes simultaneously. To accomplish the aforementioned points, the commercial cloud-based platform, 1928 platform (1928 Diagnostics, Gothenburg, Sweden) was benchmarked against an in-house developed bioinformatic pipeline as well as to reference methods in the clinical laboratory.MethodsWhole-genome sequencing data retrieved from 264 Staphylococcus aureus isolates using the Illumina HiSeq X next-generation sequencing technology was used. The S. aureus isolates were collected during a prospective observational study of community-onset severe sepsis and septic shock in adults at Skaraborg Hospital, in the western region of Sweden. The collected isolates were characterized according to accredited laboratory methods i.e., species identification by MALDI-TOF MS analysis and phenotypic antibiotic susceptibility testing (AST) by following the EUCAST guidelines. Concordance between laboratory methods and bioinformatic tools, as well as concordance between the bioinformatic tools was assessed by calculating the percent of agreement.ResultsThere was an overall high agreement between predicted genotypic AST and phenotypic AST results, 98.0% (989/1006, 95% CI 97.3–99.0). Nevertheless, the 1928 platform delivered predicted genotypic AST results with lower very major error rates but somewhat higher major error rates compared to the in-house pipeline. There were differences in processing times i.e., minutes versus hours, where the 1928 platform delivered the results faster. Furthermore, the bioinformatic workflows showed overall 99.4% (1267/1275, 95% CI 98.7–99.7) agreement in genetic prediction of the virulence gene characteristics and overall 97.9% (231/236, 95% CI 95.0–99.2%) agreement in predicting the sequence types (ST) of the S. aureus isolates.ConclusionsAltogether, the benchmarking disclosed that both bioinformatic workflows are able to deliver results with high accuracy aiding diagnostics of severe infections caused by S. aureus. It also illustrates the need of international agreement on quality control and metrics to facilitate standardization of analytical approaches for whole-genome sequencing based predictions.

Organ-on-a-Chip and Microfluidic Platforms for Oncology in the UK.

Organ-on-chip systems are capable of replicating complex tissue structures and physiological phenomena. The fine control of biochemical and biomechanical cues within these microphysiological systems provides opportunities for cancer researchers to build complex models of the tumour microenvironment. Interest in applying organ chips to investigate mechanisms such as metastatsis and to test therapeutics has grown rapidly, and this review draws together the published research using these microfluidic platforms to study cancer. We focus on both in-house systems and commercial platforms being used in the UK for fundamental discovery science and therapeutics testing. We cover the wide variety of cancers being investigated, ranging from common carcinomas to rare sarcomas, as well as secondary cancers. We also cover the broad sweep of different matrix microenvironments, physiological mechanical stimuli and immunological effects being replicated in these models. We examine microfluidic models specifically, rather than organoids or complex tissue or cell co-cultures, which have been reviewed elsewhere. However, there is increasing interest in incorporating organoids, spheroids and other tissue cultures into microfluidic organ chips and this overlap is included. Our review includes a commentary on cancer organ-chip models being developed and used in the UK, including work conducted by members of the UK Organ-on-a-Chip Technologies Network. We conclude with a reflection on the likely future of this rapidly expanding field of oncological research.

Standardization of SARS-CoV-2 Cycle Threshold Values: Multisite Investigation Evaluating Viral Quantitation across Multiple Commercial COVID-19 Detection Platforms.

The demand for testing during the coronavirus disease 2019 (COVID-19) pandemic has resulted in the production of several different commercial platforms and laboratory-developed assays for the detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). This has created several challenges, including, but not limited to, the standardization of diagnostic testing, utilization of cycle threshold (CT) values for quantitation and clinical interpretation, and data harmonization. Using reference standards consisting of a linear range of SARS-CoV-2 concentrations quantitated by viral culture-based methods and droplet digital PCR, we investigated the commutability and standardization of SARS-CoV-2 quantitation across different laboratories in the United States. We assessed SARS-CoV-2 CT values generated on multiple reverse transcription-PCR (RT-PCR) platforms and analyzed PCR efficiencies, linearity, gene targets, and CT value agreement. Our results demonstrate the inappropriateness of using SARS-CoV-2 CT values without established standards for viral quantitation. Further, we emphasize the importance of using reference standards and controls validated to independent assays, to compare results across different testing platforms and move toward better harmonization of COVID-19 quantitative test results. IMPORTANCE From the onset of the COVID-19 pandemic, the demand for SARS-CoV-2 testing has resulted in an explosion of analytical tests with very different approaches and designs. The variability in testing modalities, compounded by the lack of available commercial reference materials for standardization early in the pandemic, has led to several challenges regarding data harmonization for viral quantitation. In this study, we assessed multiple commercially available RT-PCR platforms across different laboratories within the United States using standardized reference materials characterized by viral culture methods and droplet digital PCR. We observed variability in the results generated by different instruments and laboratories, further emphasizing the importance of utilizing validated reference standards for quantitation, to better harmonize SARS-CoV-2 test results.

Agreement Between Experts and an Untrained Crowd for Identifying Dermoscopic Features Using a Gamified App: Reader Feasibility Study.

Dermoscopy is commonly used for the evaluation of pigmented lesions, but agreement between experts for identification of dermoscopic structures is known to be relatively poor. Expert labeling of medical data is a bottleneck in the development of machine learning (ML) tools, and crowdsourcing has been demonstrated as a cost- and time-efficient method for the annotation of medical images. The aim of this study is to demonstrate that crowdsourcing can be used to label basic dermoscopic structures from images of pigmented lesions with similar reliability to a group of experts. First, we obtained labels of 248 images of melanocytic lesions with 31 dermoscopic "subfeatures" labeled by 20 dermoscopy experts. These were then collapsed into 6 dermoscopic "superfeatures" based on structural similarity, due to low interrater reliability (IRR): dots, globules, lines, network structures, regression structures, and vessels. These images were then used as the gold standard for the crowd study. The commercial platform DiagnosUs was used to obtain annotations from a nonexpert crowd for the presence or absence of the 6 superfeatures in each of the 248 images. We replicated this methodology with a group of 7 dermatologists to allow direct comparison with the nonexpert crowd. The Cohen κ value was used to measure agreement across raters. In total, we obtained 139,731 ratings of the 6 dermoscopic superfeatures from the crowd. There was relatively lower agreement for the identification of dots and globules (the median κ values were 0.526 and 0.395, respectively), whereas network structures and vessels showed the highest agreement (the median κ values were 0.581 and 0.798, respectively). This pattern was also seen among the expert raters, who had median κ values of 0.483 and 0.517 for dots and globules, respectively, and 0.758 and 0.790 for network structures and vessels. The median κ values between nonexperts and thresholded average-expert readers were 0.709 for dots, 0.719 for globules, 0.714 for lines, 0.838 for network structures, 0.818 for regression structures, and 0.728 for vessels. This study confirmed that IRR for different dermoscopic features varied among a group of experts; a similar pattern was observed in a nonexpert crowd. There was good or excellent agreement for each of the 6 superfeatures between the crowd and the experts, highlighting the similar reliability of the crowd for labeling dermoscopic images. This confirms the feasibility and dependability of using crowdsourcing as a scalable solution to annotate large sets of dermoscopic images, with several potential clinical and educational applications, including the development of novel, explainable ML tools.

An optimized radiosynthesis of [18 F]DK222, a PET radiotracer for imaging PD-L1.

A radiochemical synthesis of [18 F]DK222, a peptide binder of programmed death ligand 1 protein, suitable for human PET studies is described, and results from validation productions are presented. The high specific activity radiotracer product is prepared as a sterile, apyrogenic solution that conforms to current Good Manufacturing Practice (cGMP) requirements. In addition, the production is extended to use a commercial synthesizer platform (General Electric FASTlab 2).