AI and machine learning in medical imaging: key points from development to translation.

Innovation in medical imaging artificial intelligence (AI)/machine learning (ML) demands extensive data collection, algorithmic advancements, and rigorous performance assessments encompassing aspects such as generalizability, uncertainty, bias, fairness, trustworthiness, and interpretability. Achieving widespread integration of AI/ML algorithms into diverse clinical tasks will demand a steadfast commitment to overcoming issues in model design, development, and performance assessment. The complexities of AI/ML clinical translation present substantial challenges, requiring engagement with relevant stakeholders, assessment of cost-effectiveness for user and patient benefit, timely dissemination of information relevant to robust functioning throughout the AI/ML lifecycle, consideration of regulatory compliance, and feedback loops for real-world performance evidence. This commentary addresses several hurdles for the development and adoption of AI/ML technologies in medical imaging. Comprehensive attention to these underlying and often subtle factors is critical not only for tackling the challenges but also for exploring novel opportunities for the advancement of AI in radiology.

Reframing Communication about Fall Prevention Programs to Increase Older Adults' Intentions to Participate.

INTRODUCTION: Falls and fall-related injuries in community-dwelling older adults are a growing global health concern. Despite effective exercise-based fall prevention programs (FPPs), low enrollment rates persist due to negative connotations associated with falls and aging. This study aimed to investigate whether positive framing in communication leads to a higher intention to participate in an FPP among community-dwelling older adults. METHODS: We conducted a two-sequence randomized crossover study. We designed two flyers, a standard flyer containing standard terminology regarding FPPs for older adults, and a reframed flyer highlighting fitness and activity by reframing 'fall prevention' as an 'exercise program' and 'old' as 'over 65 years'. With a Mann-Whitney U test, we investigated group differences regarding the intention to participate between the flyers. A sensitivity analysis and subgroup analyses were performed. We conducted qualitative thematic analysis on open-ended answers to gain a deeper understanding of participants' intention to participate. RESULTS: In total, we included 133 participants. Findings indicated a significantly higher intention to participate in the reframed flyer (median = 4; interquartile range = 1-6) compared to the standard flyer (median = 2; interquartile range = 1-4) (p = 0.038). Participants favored more general terms such as 'over 65 years' over 'older adults'. Older adults who were female, not at high fall risk, perceived themselves as not at fall risk, and maintained a positive attitude to aging showed greater receptivity to positively-framed communications in the reframed flyer. Additionally, already being engaged in physical activities and a lack of practical information about the FPP appeared to discourage participation intentions. DISCUSSION: The results in favor of the reframed flyer provide practical insights for designing and implementing effective (mass-)media campaigns on both (inter)national and local levels, as well as for interacting with this population on an individual basis. Aging-related terminology in promotional materials hinders engagement, underscoring the need for more positive messaging and leaving out terms such as 'older'. Tailored positively framed messages and involving diverse older adults in message development are essential for promoting participation in FPPs across various population subgroups to promote participation in FPPs among community-dwelling older adults.

Artificial intelligence in medicine: mitigating risks and maximizing benefits via quality assurance, quality control, and acceptance testing.

The adoption of artificial intelligence (AI) tools in medicine poses challenges to existing clinical workflows. This commentary discusses the necessity of context-specific quality assurance (QA), emphasizing the need for robust QA measures with quality control (QC) procedures that encompass (1) acceptance testing (AT) before clinical use, (2) continuous QC monitoring, and (3) adequate user training. The discussion also covers essential components of AT and QA, illustrated with real-world examples. We also highlight what we see as the shared responsibility of manufacturers or vendors, regulators, healthcare systems, medical physicists, and clinicians to enact appropriate testing and oversight to ensure a safe and equitable transformation of medicine through AI.

Distinct functions for the paralogous RBM41 and U11/U12-65K proteins in the minor spliceosome.

Here, we identify RBM41 as a novel unique protein component of the minor spliceosome. RBM41 has no previously recognized cellular function but has been identified as a paralog of U11/U12-65K, a known unique component of the U11/U12 di-snRNP. Both proteins use their highly similar C-terminal RRMs to bind to 3'-terminal stem-loops in U12 and U6atac snRNAs with comparable affinity. Our BioID data indicate that the unique N-terminal domain of RBM41 is necessary for its association with complexes containing DHX8, an RNA helicase, which in the major spliceosome drives the release of mature mRNA from the spliceosome. Consistently, we show that RBM41 associates with excised U12-type intron lariats, is present in the U12 mono-snRNP, and is enriched in Cajal bodies, together suggesting that RBM41 functions in the post-splicing steps of the minor spliceosome assembly/disassembly cycle. This contrasts with U11/U12-65K, which uses its N-terminal region to interact with U11 snRNP during intron recognition. Finally, while RBM41 knockout cells are viable, they show alterations in U12-type 3' splice site usage. Together, our results highlight the role of the 3'-terminal stem-loop of U12 snRNA as a dynamic binding platform for the U11/U12-65K and RBM41 proteins, which function at distinct stages of the assembly/disassembly cycle.

Identification and development of implementation strategies: the important role of codesign.

Background: To date, implementation strategies reported in the literature are commonly poorly described and take the implementation context insufficiently into account. To unravel the black box of implementation strategy development, insight is needed into effective theory-based and practical-informed strategies. The current study aims to describe the stepwise development of a practical-informed and theory-based implementation strategy bundle to implement ProMuscle, a nutrition and exercise intervention for community-dwelling older adults, in multiple settings in primary care. Methods: The first four steps of Implementation Mapping were adopted to develop appropriate implementation strategies. First, previously identified barriers to implementation were categorized into the constructs of the Consolidated Framework for Implementation Research (CFIR). Second, the CFIR-ERIC matching tool linked barriers to existing implementation strategies. Behavioral change strategies were added from the literature where necessary. Third, evidence for implementation strategies was sought. Fourth, in codesign with involved healthcare professionals and implementation experts, implementation strategies were operationalized to practical implementation activities following the guidance provided by Proctor et al. These practical implementation activities were processed into an implementation toolbox, which can be tailored to a specific context and presents prioritized implementation activities in a chronological order. Results: A previous study identified and categorized a total of 654 barriers for the implementation of a combined lifestyle intervention within the CFIR framework. Subsequently, the barriers were linked to 40 strategies. Due to the fact that many strategies impacted multiple barriers, seven overarching themes emerged based on the strategies: assessing the context, network internally, network externally, costs, knowledge, champions, and patient needs and resources. Codesign sessions with professionals and implementation experts resulted in the development of supported and tangible implementation activities for the final 20 strategies. The implementation activities were processed into a web-based implementation toolbox, which allows healthcare professionals to tailor the implementation activities to their specific context and guides healthcare professionals to prioritize implementation activities chronologically during their implementation. Conclusion: A theory-based approach in combination with codesign sessions with stakeholders is a usable Implementation Strategy Mapping Method for developing a practical implementation strategy bundle to implement ProMuscle across multiple settings in primary care. The next step involves evaluating the developed implementation strategies, including the implementation toolbox, to assess their impact on the implementation and adoption of ProMuscle.

Biosynthesis of polyhydroxybutyrate by Methylorubrum extorquens DSM13060 is essential for intracellular colonization in plant endosymbiosis.

Methylorubrum extorquens DSM13060 is an endosymbiont that lives in the cells of shoot tip meristems. The bacterium is methylotrophic and consumes plant-derived methanol for the production of polyhydroxybutyrate (PHB). The PHB provides protection against oxidative stress for both host and endosymbiont cells through its fragments, methyl-esterified 3-hydroxybutyrate (ME-3HB) oligomers. We evaluated the role of the genes involved in the production of ME-3HB oligomers in the host colonization by the endosymbiont M. extorquens DSM13060 through targeted genetic mutations. The strains with deletions in PHB synthase (phaC), PHB depolymerase (phaZ1), and a transcription factor (phaR) showed altered PHB granule characteristics, as ΔphaC had a significantly low number of granules, ΔphaR had a significantly increased number of granules, and ΔphaZ1 had significantly large PHB granules in the bacterial cells. When the deletion strains were exposed to oxidative stress, the ΔphaC strain was sensitive to 10 mM HO· and 20 mM H2O2. The colonization of the host, Scots pine (Pinus sylvestris L.), by the deletion strains varied greatly. The deletion strain ΔphaR colonized the host mainly intercellularly, whereas the ΔphaZ1 strain was a slightly poorer colonizer than the control. The deletion strain ΔphaC lacked the colonization potential, living mainly on the surfaces of the epidermis of pine roots and shoots in contrast to the control, which intracellularly colonized all pine tissues within the study period. In earlier studies, deletions within the PHB metabolic pathway have had a minor effect on plant colonization by rhizobia. We have previously shown the association between ME-3HB oligomers, produced by PhaC and PhaZ1, and the ability to alleviate host-generated oxidative stress during plant infection by the endosymbiont M. extorquens DSM13060. Our current results show that the low capacity for PHB synthesis leads to poor tolerance of oxidative stress and loss of colonization potential by the endosymbiont. Altogether, our findings demonstrate that the metabolism of PHB in M. extorquens DSM13060 is an important trait in the non-rhizobial endosymbiosis.

High-Throughput Miniaturized Synthesis of PROTAC-Like Molecules.

The development of miniaturized high-throughput in situ screening platforms capable of handling the entire process of drug synthesis to final screening is essential for advancing drug discovery in the future. In this study, an approach based on combinatorial solid-phase synthesis, enabling the efficient synthesis of libraries of proteolysis targeting chimeras (PROTACs) in an array format is presented. This on-chip platform allows direct biological screening without the need for transfer steps.  UV-induced release of target molecules into individual droplets facilitates further on-chip experimentation. Utilizing a mitogen-activated protein kinase kinases (MEK1/2) degrader as a template, a series of 132 novel PROTAC-like molecules is synthesized using solid-phase Ugi reaction. These compounds are further characterized using various methods, including matrix-assisted laser desorption ionization mass spectrometry (MALDI-MS) imaging, while consuming only a few milligrams of starting materials in total. Furthermore, the feasibility of culturing cancer cells on the modified spots and quantifying the effect of MEK suppression is demonstrated. The miniaturized synthesis platform lays a foundation for high-throughput in situ biological screening of potent PROTACs for potential anticancer activity and offers the potential for accelerating the drug discovery process by integrating miniaturized synthesis and biological steps on the same array.

Virtual neural network-guided optimization of non-invasive brain stimulation in Alzheimer's disease.

Transcranial direct current stimulation (tDCS) is a non-invasive brain stimulation technique with potential for counteracting disrupted brain network activity in Alzheimer's disease (AD) to improve cognition. However, the results of tDCS studies in AD have been variable due to different methodological choices such as electrode placement. To address this, a virtual brain network model of AD was used to explore tDCS optimization. We compared a large, representative set of virtual tDCS intervention setups, to identify the theoretically optimized tDCS electrode positions for restoring functional network features disrupted in AD. We simulated 20 tDCS setups using a computational dynamic network model of 78 neural masses coupled according to human structural topology. AD network damage was simulated using an activity-dependent degeneration algorithm. Current flow modeling was used to estimate tDCS-targeted cortical regions for different electrode positions, and excitability of the pyramidal neurons of the corresponding neural masses was modulated to simulate tDCS. Outcome measures were relative power spectral density (alpha bands, 8-10 Hz and 10-13 Hz), total spectral power, posterior alpha peak frequency, and connectivity measures phase lag index (PLI) and amplitude envelope correlation (AEC). Virtual tDCS performance varied, with optimized strategies improving all outcome measures, while others caused further deterioration. The best performing setup involved right parietal anodal stimulation, with a contralateral supraorbital cathode. A clear correlation between the network role of stimulated regions and tDCS success was not observed. This modeling-informed approach can guide and perhaps accelerate tDCS therapy development and enhance our understanding of tDCS effects. Follow-up studies will compare the general predictions to personalized virtual models and validate them with tDCS-magnetoencephalography (MEG) in a clinical AD patient cohort.

Weather in two climatic regions shapes the diversity and drives the structure of fungal endophytic community of bilberry (Vaccinium myrtillus L.) fruit.

BACKGROUND: Bilberry (Vaccinium myrtillus L.) is one of the most important economic and natural resources in Northern Europe. Despite its importance, the endophytic fungal community of the fruits has rarely been investigated. Biogeographic patterns and determinants of the fungal diversity in the bilberry fruit are poorly understood, albeit fungal endophytes can have a close relationship with the host plants. Here, we investigated the effect of climatic regions, and their weather conditions within growth season and soil properties on fungal endophytic communities of bilberry fruits collected from northern and southern regions of Finland using high-throughput sequencing technology targeting the internal transcribed spacer 2 ribosomal DNA region for fungi. RESULTS: Species richness and beta diversity (variation in community structure) were higher in the southern compared to the studied northern region. The weather condition of the growth season drove both fungal richness and community structure. Furthermore, abundance of the genera Venturia, Cladosporium, and Podosphaera was influenced by the weather, being different between the south and north regions. CONCLUSIONS: We conclude that diversity and assembly structure of the fungal endophytes in bilberry fruits follow similar patterns as for foliar fungal endophytes, being shaped by various environmental factors, such as the climate and surrounding vegetation.

Subthalamic and pallidal neurons are modulated during externally cued movements in Parkinson's disease.

External sensory cues can reduce freezing of gait in people with Parkinson's disease (PD), yet the role of the basal ganglia in these movements is unclear. We used microelectrode recordings to examine modulations in single unit (SU) and oscillatory local field potentials (LFP) during auditory-cued rhythmic pedaling movements of the feet. We tested five blocks of increasing cue frequencies (1 Hz, 1.5 Hz, 2 Hz, 2.5 Hz, and 3 Hz) in 24 people with PD undergoing deep brain stimulation surgery of the subthalamic nucleus (STN) or globus pallidus internus (GPi). Single unit firing and beta band LFPs (13-30 Hz) in response to movement onsets or cue onsets were examined. We found that the timing accuracy of foot pedaling decreased with faster cue frequencies. Increasing cue frequencies also attenuated firing rates in both STN and GPi neurons. Peak beta power in the GPi and STN showed different responses to the task. GPi beta power showed persistent suppression with fast cues and phasic modulation with slow cues. STN beta power showed enhanced beta synchronization following movement. STN beta power also correlated with rate of pedaling. Overall, we showed task-related responses in the GPi and STN during auditory-cued movements with differential roles in sensory and motor control. The results suggest a role for both input and output basal ganglia nuclei in auditory rhythmic pacing of gait-like movements in PD.

Barriers and facilitators for implementation of a combined lifestyle intervention in community-dwelling older adults: a scoping review.

Background: Lifestyle interventions, combining nutrition and exercise, are effective in improving the physical functioning of community-dwelling older adults and preventing healthcare risks due to loss in muscle mass. However, the potential of these types of interventions is not being fully exploited due to insufficient implementation. Having insight into the determinants that could hinder or facilitate the implementation of a combined lifestyle intervention could improve the development of matching implementation strategies and enhance the implementation of such lifestyle interventions. The aim of this study was to identify barriers and facilitators for the successful implementation of a combined lifestyle intervention for community-dwelling older adults. Method: A scoping review was conducted. A literature search was conducted in four electronic databases, and references were checked for additional inclusion. Studies were screened if they met the inclusion criteria. Barriers and facilitators were extracted from the included studies. To validate the results of the literature search, healthcare professionals and community-dwelling older adults were interviewed. Barriers and facilitators were categorized by two researchers according to the constructs of the Consolidated Framework for Implementation Research (CFIR). Results: The search identified 12,364 studies, and 23 were found eligible for inclusion in the review. Barriers and facilitators for 26 of the 39 constructs of the CFIR were extracted. The interviews with healthcare professionals and older adults yielded six extra barriers and facilitators for implementation, resulting in determinants for 32 of the 39 CFIR constructs. According to literature and healthcare professionals, cosmopolitanism (network with external organizations), patient needs and resources, readiness for implementation, costs, knowledge and beliefs about the intervention, network and communication, and engaging were found to be the most important determinants for implementation of a combined lifestyle intervention. Conclusion: A broad range of barriers and facilitators across all domains of the CFIR framework emerged in this study. The results of this review reflect on determinants that should be taken into account when planning for the implementation of a combined lifestyle intervention. A further step in the implementation process is the development of implementation strategies aiming at the identified determinants to enhance the implementation of a combined lifestyle intervention in community care.

Bayesian uncertainty estimation for detection of long-tailed and unseen conditions in medical images.

Purpose: Deep supervised learning provides an effective approach for developing robust models for various computer-aided diagnosis tasks. However, there is often an underlying assumption that the frequencies of the samples between the different classes of the training dataset are either similar or balanced. In real-world medical data, the samples of positive classes often occur too infrequently to satisfy this assumption. Thus, there is an unmet need for deep-learning systems that can automatically identify and adapt to the real-world conditions of imbalanced data. Approach: We propose a deep Bayesian ensemble learning framework to address the representation learning problem of long-tailed and out-of-distribution (OOD) samples when training from medical images. By estimating the relative uncertainties of the input data, our framework can adapt to imbalanced data for learning generalizable classifiers. We trained and tested our framework on four public medical imaging datasets with various imbalance ratios and imaging modalities across three different learning tasks: semantic medical image segmentation, OOD detection, and in-domain generalization. We compared the performance of our framework with those of state-of-the-art comparator methods. Results: Our proposed framework outperformed the comparator models significantly across all performance metrics (pairwise t-test: p<0.01) in the semantic segmentation of high-resolution CT and MR images as well as in the detection of OOD samples (p<0.01), thereby showing significant improvement in handling the associated long-tailed data distribution. The results of the in-domain generalization also indicated that our framework can enhance the prediction of retinal glaucoma, contributing to clinical decision-making processes. Conclusions: Training of the proposed deep Bayesian ensemble learning framework with dynamic Monte-Carlo dropout and a combination of losses yielded the best generalization to unseen samples from imbalanced medical imaging datasets across different learning tasks.

CRISPR screens in sister chromatid cohesion defective cells reveal PAXIP1-PAGR1 as regulator of chromatin association of cohesin.

The cohesin complex regulates higher order chromosome architecture through maintaining sister chromatid cohesion and folding chromatin by DNA loop extrusion. Impaired cohesin function underlies a heterogeneous group of genetic syndromes and is associated with cancer. Here, we mapped the genetic dependencies of human cell lines defective of cohesion regulators DDX11 and ESCO2. The obtained synthetic lethality networks are strongly enriched for genes involved in DNA replication and mitosis and support the existence of parallel sister chromatid cohesion establishment pathways. Among the hits, we identify the chromatin binding, BRCT-domain containing protein PAXIP1 as a novel cohesin regulator. Depletion of PAXIP1 severely aggravates cohesion defects in ESCO2 mutant cells, leading to mitotic cell death. PAXIP1 promotes global chromatin association of cohesin, independent of DNA replication, a function that cannot be explained by indirect effects of PAXIP1 on transcription or DNA repair. Cohesin regulation by PAXIP1 requires its binding partner PAGR1 and a conserved FDF motif in PAGR1. PAXIP1 co-localizes with cohesin on multiple genomic loci, including active gene promoters and enhancers. Possibly, this newly identified role of PAXIP1-PAGR1 in regulating cohesin occupancy on chromatin is also relevant for previously described functions of PAXIP1 in transcription, immune cell maturation and DNA repair.

Staphylococcus aureus surgical site infection rates in 5 European countries.

OBJECTIVE: To determine the overall and procedure-specific incidence of surgical site infections (SSI) caused by Staphylococcus aureus (S. aureus) as well as risk factors for such across all surgical disciplines in Europe. METHODS: This is a retrospective cohort of patients with surgical procedures performed at 14 European centres in 2016, with a nested case-control analysis. S. aureus SSI were identified by a semi-automated crossmatching bacteriological and electronic health record data. Within each surgical procedure, cases and controls were matched using optimal propensity score matching. RESULTS: A total of 764 of 178 902 patients had S. aureus SSI (0.4%), with 86.0% of these caused by methicillin susceptible and 14% by resistant pathogens. Mean S. aureus SSI incidence was similar for all surgical specialties, while varying by procedure. CONCLUSIONS: This large procedure-independent study of S. aureus SSI proves a low overall infection rate of 0.4% in this cohort. It provides proof of principle for a semi-automated approach to utilize big data in epidemiological studies of healthcare-associated infections. Trials registration The study was registered at clinicaltrials.gov under NCT03353532 (11/2017).

Health perspectives and handling of PFAS contamination in Denmark.

Per- and polyfluorinated alkyl substances (PFAS) are a group of man-made extremely persistent chemicals that have been spread worldwide since the 1950s due to their properties as effective water- and grease-repellent and heat- and temperature-resistant. They are associated with an increasing number of health effects including immune and hormonal disturbances and some types of cancer. We present a review of PFAS pollution in Denmark, the current human exposure, suspected health effects, and patient management in environmental medicine.

ShrinkCRISPR: a flexible method for differential fitness analysis of CRISPR-Cas9 screen data.

BACKGROUND: CRISPR screens provide large-scale assessment of cellular gene functions. Pooled libraries typically consist of several single guide RNAs (sgRNAs) per gene, for a large number of genes, which are transduced in such a way that every cell receives at most one sgRNA, resulting in the disruption of a single gene in that cell. This approach is often used to investigate effects on cellular fitness, by measuring sgRNA abundance at different time points. Comparing gene knockout effects between different cell populations is challenging due to variable cell-type specific parameters and between replicates variation. Failure to take those into account can lead to inflated or false discoveries. RESULTS: We propose a new, flexible approach called ShrinkCRISPR that can take into account multiple sources of variation. Impact on cellular fitness between conditions is inferred by using a mixed-effects model, which allows to test for gene-knockout effects while taking into account sgRNA-specific variation. Estimates are obtained using an empirical Bayesian approach. ShrinkCRISPR can be applied to a variety of experimental designs, including multiple factors. In simulation studies, we compared ShrinkCRISPR results with those of drugZ and MAGeCK, common methods used to detect differential effect on cell fitness. ShrinkCRISPR yielded as many true discoveries as drugZ using a paired screen design, and outperformed both drugZ and MAGeCK for an independent screen design. Although conservative, ShrinkCRISPR was the only approach that kept false discoveries under control at the desired level, for both designs. Using data from several publicly available screens, we showed that ShrinkCRISPR can take data for several time points into account simultaneously, helping to detect early and late differential effects. CONCLUSIONS: ShrinkCRISPR is a robust and flexible approach, able to incorporate different sources of variations and to test for differential effect on cell fitness at the gene level. These improve power to find effects on cell fitness, while keeping multiple testing under the correct control level and helping to improve reproducibility. ShrinkCrispr can be applied to different study designs and incorporate multiple time points, making it a complete and reliable tool to analyze CRISPR screen data.

Nanoliter Scale Parallel Liquid-Liquid Extraction for High-Throughput Purification on a Droplet Microarray.

In the current drug discovery process, the synthesis of compound libraries is separated from biological screenings both conceptually and technologically. One of the reasons is that parallel on-chip high-throughput purification of synthesized compounds is still a major challenge. Here, on-chip miniaturized high-throughput liquid-liquid extraction in volumes down to 150 nL with efficiency comparable to or better than large-scale extraction utilizing separation funnels is demonstrated. The method is based on automated and programmable merging of arrays of aqueous nanoliter droplets with organic droplets. Multi-step extraction performed simultaneously or with changing conditions as well as handling of femtomoles of compounds are demonstrated. In addition, the extraction efficiency is analyzed with a fast optical readout as well as matrix-assisted laser desorption ionization-mass spectrometry on-chip detection. The new massively parallel and miniaturized purification method adds another important tool to the chemBIOS concept combining chemical combinatorial synthesis with biological screenings on the same miniaturized droplet microarray platform, which will be essential to accelerate drug discovery.

AAPM task group report 273: Recommendations on best practices for AI and machine learning for computer-aided diagnosis in medical imaging.

Rapid advances in artificial intelligence (AI) and machine learning, and specifically in deep learning (DL) techniques, have enabled broad application of these methods in health care. The promise of the DL approach has spurred further interest in computer-aided diagnosis (CAD) development and applications using both "traditional" machine learning methods and newer DL-based methods. We use the term CAD-AI to refer to this expanded clinical decision support environment that uses traditional and DL-based AI methods. Numerous studies have been published to date on the development of machine learning tools for computer-aided, or AI-assisted, clinical tasks. However, most of these machine learning models are not ready for clinical deployment. It is of paramount importance to ensure that a clinical decision support tool undergoes proper training and rigorous validation of its generalizability and robustness before adoption for patient care in the clinic. To address these important issues, the American Association of Physicists in Medicine (AAPM) Computer-Aided Image Analysis Subcommittee (CADSC) is charged, in part, to develop recommendations on practices and standards for the development and performance assessment of computer-aided decision support systems. The committee has previously published two opinion papers on the evaluation of CAD systems and issues associated with user training and quality assurance of these systems in the clinic. With machine learning techniques continuing to evolve and CAD applications expanding to new stages of the patient care process, the current task group report considers the broader issues common to the development of most, if not all, CAD-AI applications and their translation from the bench to the clinic. The goal is to bring attention to the proper training and validation of machine learning algorithms that may improve their generalizability and reliability and accelerate the adoption of CAD-AI systems for clinical decision support.

Fluorogenic RNA aptamers to probe transcription initiation and co-transcriptional RNA folding by multi-subunit RNA polymerases.

Transcription is the first and most highly regulated step in gene expression. Experimental techniques for monitoring transcription are, thus, important for studying gene expression and gene regulation as well as for translational research and drug development. Fluorescence methods are often superior to other techniques for real-time monitoring of biochemical processes. Green fluorescent proteins have long served as valuable tools for studying the process of translation. Here we present two methods that utilize fluorescent light-up RNA aptamers (FLAPs), the RNA mimics of green fluorescent proteins, to monitoring transcription and co-transcriptional RNA folding. FLAPs adopt defined three-dimensional folds that bind low molecular weight compounds called fluorogens with concomitant increase in fluorescence by many folds. FLAPs provide a strong fluorescence signal with low background that allows monitoring of transcription in real time in vitro and in vivo. However, it takes several seconds for RNA polymerase to synthesize FLAPs and the subsequent folding of the fluorogen-binding platform takes additional seconds or minutes. Here we show that Broccoli-FLAP is well suited for monitoring the rate of transcription initiation in a multi-round setup that mitigates the slow rate of the FLAP maturation. Furthermore, we demonstrate that a relatively slow and inefficient folding of iSpinach-FLAP can be taken advantage of for monitoring the action of RNA folding chaperones.