Manual Workflow Research Articles

Curating models from BioModels: Developing a workflow for creating OMEX files.

The reproducibility of computational biology models can be greatly facilitated by widely adopted standards and public repositories. We examined 50 models from the BioModels Database and attempted to validate the original curation and correct some of them if necessary. For each model, we reproduced these published results using Tellurium. Once reproduced we manually created a new set of files, with the model information stored by the Systems Biology Markup Language (SBML), and simulation instructions stored by the Simulation Experiment Description Markup Language (SED-ML), and everything included in an Open Modeling EXchange (OMEX) file, which could be used with a variety of simulators to reproduce the same results. On the one hand, the reproducibility procedure of 50 models developed a manual workflow that we would use to build an automatic platform to help users more easily curate and verify models in the future. On the other hand, these exercises allowed us to find the limitations and possible enhancement of the current curation and tooling to verify and curate models.

Accurate and Automated Genotyping of the CFTR Poly-T/TG Tract with CFTR-TIPS.

Cystic fibrosis is caused by biallelic pathogenic variants in the CFTR gene, which contains a polymorphic (TG)mTn sequence (the "poly-T/TG tract") in intron 9. While T9 and T7 alleles are benign, T5 alleles with longer TG repeats, e.g., (TG)12T5 and (TG)13T5, are clinically significant. Thus, professional medical societies currently recommend reporting the TG repeat size when T5 is detected. Sanger sequencing is a cost-effective method of genotyping the (TG)mTn tract; however, its polymorphic length substantially complicates data analysis. We developed CFTR-TIPS, a freely available web-based software tool that infers the (TG)mTn genotype from Sanger sequencing data. This tool detects the (TG)mTn tract in the chromatograms, quantifies goodness of fit with expected patterns, and visualizes the results in a graphical user interface. It is broadly compatible with any Sanger chromatogram that contains the (TG)mTn tract ± 15 bp. We evaluated CFTR-TIPS using 835 clinical samples previously analyzed in a CLIA-certified, CAP-accredited laboratory. When operated fully automatically, CFTR-TIPS achieved 99.8% concordance with our clinically validated manual workflow, while generally taking less than 10 s per sample. There were two discordant samples: one due to a co-occurring heterozygous duplication that confounded the tool and the other due to incomplete (TG)mTn tract detection in the reverse chromatogram. No clinically significant misclassifications were observed. CFTR-TIPS is a free, accurate, and rapid tool for CFTR (TG)mTn tract genotyping using cost-effective Sanger sequencing. This tool is suitable both for automated use and as an aid to manual review to enhance accuracy and reduce analysis time.

P-236 Simulated cohort ranking analysis estimates the number of IVF cycles needed to reach fetal heartbeat with and without artificial intelligence (AI) embryo triage

Abstract Study question What is the impact of AI-based embryo triage on the number of IVF cycles needed to reach fetal heartbeat (FH)? Summary answer Simulated cohort ranking via the AIVF Day-5 AI model demonstrates a 27.5% reduction in the number of IVF cycles needed to achieve FH. What is known already Published reports show FH rates per first-time IVF/single euploid embryo transfer ranging from 26-43% (age <38 yrs). Likewise, the average number of autologous retrieval cycles required for FH is 2.25±0.8, depending on oocyte age and demographic. While conventional embryo morphology ranking is used as a proxy indicator for IVF success, its use is associated with observer variability and manual workflow. This leads to a non-efficient cycle number-per-FH ratio, preventing the clinic’s ability to service more patients. Study design, size, duration A single center database (N = 10,000 embryos) was used to generate 1,288 simulated cohorts reflecting known maternal age/cohort size distributions. Only cohorts containing ≥1 FH+ embryo(s) were included. The model ranked embryos within each cohort by scalar output (1-9.9). The position of the first FH+ embryo within each AI-ranked cohort was calculated and averaged across cohorts to compute number of cycles required to reach FH. Participants/materials, setting, methods Data was stratified into 5 age categories; the prevalence of top/good/fair graded embryos per cohort were included according to their known (%) prevalences per stratified age group. Data was randomized into 1,288 cohorts; positions of the first FH+ embryo in each cohort was calculated (for example, if the 2nd AI-ranked embryo was FH+, a value of 2 was assigned, respectively). Values were averaged and compared to conventional IVF. Main results and the role of chance A mean ± SD of 1.63 ±1.06 cycles was needed to reach the first FH+ embryo with AI cohort ranking (pooled mean across all age groups <38 years) with a mean cohort size N = 7.8 embryos. The relative percent reduction in cycles needed to reach FH when compared to conventional IVF without AI ranking (2.25±0.8; age <38 yrs) was 27.5%. Ranking and electronic recording time for ranking per embryo with/without AI implementation was 186.0 versus 30.9 seconds per embryo, respectively. Relative ranking time reduction was 83%. Limitations, reasons for caution Results were pooled across age groups and not stratified per age. Prospective validation is ongoing to show the influence of demographics/clinical characteristics on AI utility. Wider implications of the findings This study demonstrates the potential of shortening the time to pregnancy by using AI-based quantitative embryo assessment for cohort ranking. Trial registration number Not applicable

Safety and efficiency of a fully automatic workflow for auto-segmentation in radiotherapy using three commercially available deep learning-based applications

Advancements in radiotherapy auto-segmentation necessitate reliable and efficient workflows. Therefore, a standardized fully automatic workflow was developed for three commercially available deep learning-based auto-segmentation applications and compared to a manual workflow for safety and efficiency. The workflow underwent safety evaluation with failure mode and effects analysis. Notably, eight failure modes were reduced, including seven with severity factors ≥7, indicating the effect on patients, and two with Risk Priority Number value >125, which assesses relative risk level. Efficiency, measured by mouse clicks, showed zero clicks with the automatic workflow. This automation illustrated improvement in both safety and efficiency of workflow

An engineering biology approach to automated workflow and biodesign

Abstract The paper addresses the application of engineering biology strategies and techniques to the automation of laboratory workflow—primarily in the context of biofoundries and biodesign applications based on the Design, Build, Test and Learn paradigm. The trend toward greater automation comes with its own set of challenges. On the one hand, automation is associated with higher throughput and higher replicability. On the other hand, the implementation of an automated workflow requires an instruction set that is far more extensive than that required for a manual workflow. Automated tasks must also be conducted in the order specified in the workflow, with the right logic, utilizing suitable biofoundry resources, and at scale—while simultaneously collecting measurements and associated data. The paper describes an approach to an automated workflow that is being trialed at the London Biofoundry at SynbiCITE. The solution represents workflows with directed graphs, uses orchestrators for their execution, and relies on existing standards. The approach is highly flexible and applies to not only workflow automation in single locations but also distributed workflows (e.g. for biomanufacturing). The final section presents an overview of the implementation—using the simple example of an assay based on a dilution, measurement, and data analysis workflow.

Evaluation of the Efficiency of a Digital Workflow for Cataract Planning in Patients with Astigmatism.

To evaluate the efficiency and associated costs of a digital cataract workflow system and manual cataract workflow system for patients, with astigmatism presenting for cataract surgery in Brisbane, Australia. Sixty patients with bilateral cataract requiring toric intraocular lenses (IOL) were sequentially assigned to a manual cataract workflow (n = 30) or digital workflow (n = 30) using EQ Workplace (SW v1.7.0) running on FORUM (SW v.4.2.1.66) (Carl Zeiss Meditec, Jena, Germany). Each step of preoperative data acquisition and analysis was timed. Steps in each workflow were divided into presurgical planning time and total workflow time, the latter including the time required to input toric data into CALLISTO eye (Carl Zeiss Meditec). Secondary outcomes included staff costs within each workflow. Median presurgical planning time using a digital workflow process was 6.51 ± 0.65 minutes, and using a manual workflow process, 12.32 ± 0.56 minutes (p < 0.001). Similarly, median total workflow time using a digital workflow process was 6.93 ± 0.57 minutes and using a manual workflow process, 13.49 ± 0.47 minutes (p < 0.001). Evaluating the staff remuneration during presurgical planning and the operating costs associated with running EQ Workplace, there was a cost-reduction of 35% per patient when using the digital cataract workflow process. Using a digital cataract workflow process is more efficient and provides staff cost-savings compared to a manual workflow process when planning for toric IOL implantation.

Automating the Illumina DNA library preparation kit for whole genome sequencing applications on the flowbot ONE liquid handler robot.

Whole-genome sequencing (WGS) is currently making its transition from research tool into routine (clinical) diagnostic practice. The workflow for WGS includes the highly labor-intensive library preparations (LP), one of the most critical steps in the WGS procedure. Here, we describe the automation of the LP on the flowbot ONE robot to minimize the risk of human error and reduce hands-on time (HOT). For this, the robot was equipped, programmed, and optimized to perform the Illumina DNA Prep automatically. Results obtained from 16 LP that were performed both manually and automatically showed comparable library DNA yields (median of 1.5-fold difference), similar assembly quality values, and 100% concordance on the final core genome multilocus sequence typing results. In addition, reproducibility of results was confirmed by re-processing eight of the 16 LPs using the automated workflow. With the automated workflow, the HOT was reduced to 25min compared to the 125min needed when performing eight LPs using the manual workflow. The turn-around time was 170 and 200min for the automated and manual workflow, respectively. In summary, the automated workflow on the flowbot ONE generates consistent results in terms of reliability and reproducibility, while significantly reducing HOT as compared to manual LP.

Evaluation of an automated platform for the detection of HEV RNA in plasma and stool

IntroductionWe evaluated the performance of the automated Altostar HEV RNA platform for detecting HEV RNA. Methods and resultsClinical performance was determined by testing 81 plasma samples and 10 fecal samples manually quantified previously with the Realstar RT-PCR assay using the Magnapure instrument for extraction. The assays were concordant for 79/81 plasma samples (97.5%) and 10/10 (100%) fecal samples. The two plasma samples that tested negative with the Altostar assay had a very low HEV RNA concentration (1.6 and 1.4 log10 IU/ml). Quantitative results obtained with the automated platform and the manual workflow were highly correlated (ρ= 0.98, p<0.01). The intra-run and inter-run standard deviation were 0.09 IU/ml and 0.13 IU/ml respectively. The assay was linear from 2 to 6 log IU/ml. The limit of detection determined by Probit analysis with the WHO HEV RNA standard was 7.6 [95% CI: 4.4–52.5] IU/ml. ConclusionsThe Altostar platform enables highly accurate testing for the detection of HEV RNA in stool and the quantification of HEV RNA in plasma. This allowed us to shorten turnaround times and to save time for the technical staff.

Method for Quantitative HDV RNA Detection: I, Manual Workflow (Serum and Liver Tissue) and II, Fully Automated High Throughput Workflow for Diagnostic Use.

The hepatitis delta virus (HDV) is a small RNA virus (1700 base pairs), which uses the surface proteins of the hepatitis B virus (HBV) as an envelope. Accurate and reliable quantitative detection of HDV RNA is central for scientific and translational clinical research or diagnostic purposes. However, HDV poses challenges for nucleic acid amplification techniques: (1) the circular genome displays high intramolecular base pairing; (2) high content of cytosine and guanine; and (3) enormous genomic diversity among the eight known HDV genotypes (GTs). Here, we provide step-by-step instructions for (A) a manual workflow to perform a quantitative HDV reverse transcription (RT)-PCR from serum and liver tissue and (B) a quantitative HDV RT-PCR assay with whole process control to be used for serum or plasma samples run on a fully automated system. Both assays target the conserved ribozyme region and demonstrate inclusivity for all eight HDV GTs. The choice of assay depends on the experimental needs and equipment availability. While the former is ideal for scientific research laboratories, the latter provides a useful tool in the field of translational research or diagnostics.

Diseases diagnosis based on artificial intelligence and ensemble classification

BackgroundIn recent years, Computer Aided Diagnosis (CAD) has become an important research area that attracted a lot of researchers. In medical diagnostic systems, several attempts have been made to build and enhance CAD applications to avoid errors that can cause dangerously misleading medical treatments. The most exciting opportunity for promoting the performance of CAD system can be accomplished by integrating Artificial Intelligence (AI) in medicine. This allows the effective automation of traditional manual workflow, which is slow, inaccurate and affected by human errors. AimsThis paper aims to provide a complete Computer Aided Disease Diagnosis (CAD2) strategy based on Machine Learning (ML) techniques that can help clinicians to make better medical decisions. MethodsThe proposed CAD2 consists of three main sequential phases, namely; (i) Outlier Rejection Phase (ORP), (ii) Feature Selection Phase (FSP), and (iii) Classification Phase (CP). ORP is implemented to reject outliers using new Outlier Rejection Technique (ORT) that contains two sequential stages called Fast Outlier Rejection (FOR) and Accurate Outlier Rejection (AOR). The most informative features are selected through FSP using Hybrid Selection Technique (HST). HST includes two main stages called Quick Selection Stage (QS2) using fisher score as a filter method and Precise Selection Stage (PS2) using a Hybrid Bio-inspired Optimization (HBO) technique as a wrapper method. Finally, actual diagnose takes place through CP, which relies on Ensemble Classification Technique (ECT). ResultsThe proposed CAD2 has been tested experimentally against recent disease diagnostic strategies using two different datasets in which the first contains several diseases, while the second includes data for Covid-19 patients only. Experimental results have proven the high efficiency of the proposed CAD2 in terms of accuracy, error, precision, and recall compared with other competitors. Additionally, CAD2 strategy provides the best Wilcoxon signed rank test and Friedman test measurements against other strategies according to both datasets. ConclusionIt is concluded that CAD2 strategy based on ORP, FSP, and CP gave an accurate diagnosis compared to other strategies because it gave the highest accuracy and the lowest error and implementation time.

Improving Diabetes Screening in the Primary Care Clinic.

In our suburban primary care clinic, the average rate of screening for diabetes among eligible patients was only 51%, similar to national screening data. We conducted a quality improvement project to increase this rate. During the 6-month preintervention phase, we collected baseline data on the percentage of eligible patients screened per week (percentage of patients with hemoglobin A1c checked in the prior 3 years out of patients eligible for screening who completed a visit during the week). We then implemented a two-phase intervention. In phase 1 (approximately 8 months), we generated an electronic health record (EHR) report to identify eligible patients and pended laboratory orders for physicians to sign. In phase 2 (approximately 3 months), we replaced the phase 1 intervention with an EHR clinical decision support tool that automatically identifies eligible patients. We compared screening rates in the preintervention vs. intervention period. For phase 1, we also assessed laboratory completion rates and the laboratory results. We surveyed physicians regarding intervention acceptability and satisfaction at 3, 6, 9, and 12 months during the intervention period. The weekly percentage of patients screened increased from an average of 51% in the preintervention phase to 65% in the intervention phase (p < 0.001). During phase 1, most patients underwent laboratory blood testing as recommended (83% within 3 months), and results were consistent with prediabetes in 23% and with diabetes in 4%. Overall, most physicians believed that the intervention appropriately identified patients due for screening and was helpful (100% of respondents agreed at 9 months vs. 71% at 3 months). We successfully implemented a systematic screening intervention involving a manual workflow and EHR tool and improved diabetes screening rates in our clinic.

Sharing Patient Praises With Radiology Staff: Workflow Automation and Impact on Staff

ObjectiveThis study aims to develop and evaluate a semi-automated workflow using natural language processing (NLP) for sharing positive patient feedback with radiology staff, assessing its efficiency and impact on radiology staff morale. MethodsThe HIPAA-compliant, institutional review board–waived implementation study was conducted from April 2022 to June 2023 and introduced a Patient Praises program to distribute positive patient feedback to radiology staff collected from patient surveys. The study transitioned from an initial manual workflow to a hybrid process using an NLP model trained on 1,034 annotated comments and validated on 260 holdout reports. The times to generate Patient Praises e-mails were compared between manual and hybrid workflows. Impact of Patient Praises on radiology staff was measured using a four-question Likert scale survey and an open text feedback box. Kruskal-Wallis test and post hoc Dunn’s test were performed to evaluate differences in time for different workflows. ResultsFrom April 2022 to June 2023, the radiology department received 10,643 patient surveys. Of those surveys, 95.6% contained positive comments, with 9.6% (n = 978) shared as Patient Praises to staff. After implementation of the hybrid workflow in March 2023, 45.8% of Patient Praises were sent through the hybrid workflow and 54.2% were sent manually. Time efficiency analysis on 30-case subsets revealed that the hybrid workflow without edits was the most efficient, taking a median of 0.7 min per case. A high proportion of staff found the praises made them feel appreciated (94%) and valued (90%) responding with a 5/5 agreement on 5-point Likert scale responses. ConclusionA hybrid workflow incorporating NLP significantly improves time efficiency for the Patient Praises program while increasing feelings of acknowledgment and value among staff.

Development of a semi-automated MHC-associated peptide proteomics (MAPPs) method using streptavidin bead-based immunoaffinity capture and nano LC-MS/MS to support immunogenicity risk assessment in drug development.

Major histocompatibility complex (MHC)-Associated Peptide Proteomics (MAPPs) is an ex vivo method used to assess the immunogenicity risk of biotherapeutics. MAPPs can identify potential T-cell epitopes within the biotherapeutic molecule. Using adalimumab treated human monocyte derived dendritic cells (DCs) and a pan anti-HLA-DR antibody (Ab), we systematically automated and optimized biotin/streptavidin (SA)-capture antibody coupling, lysate incubation with capture antibody, as well as the washing and elution steps of a MAPPs method using functionalized magnetic beads and a KingFisher Magnetic Particle processor. Automation of these steps, combined with capturing using biotinylated-Ab/SA magnetic beads rather than covalently bound antibody, improved reproducibility as measured by minimal inter-and intra-day variability, as well as minimal analyst-to-analyst variability. The semi-automated MAPPs workflow improved sensitivity, allowing for a lower number of cells per analysis. The method was assessed using five different biotherapeutics with varying immunogenicity rates ranging from 0.1 to 48% ADA incidence in the clinic. Biotherapeutics with ≥10%immunogenicity incidence consistently presented more peptides (1.8-28 fold) and clusters (10-21 fold) compared to those with <10% immunogenicity incidence. Our semi-automated MAPPs method provided two main advantages over a manual workflow- the robustness and reproducibility affords confidence in the epitopes identified from as few as 5 to 10 donors and the method workflow can be readily adapted to incorporate different capture Abs in addition to anti-HLA-DR. The incorporation of semi-automated MAPPs with biotinylated-Ab/SA bead-based capture in immunogenicity screening strategies allows the generation of more consistent and reliable data, helping to improve immunogenicity prediction capabilities in drug development. MHC associated peptide proteomics (MAPPs), Immunogenicity risk assessment, in vitro/ex vivo, biotherapeutics, Major Histocompatibility Complex Class II (MHC II), LC-MS, Immunoaffinity Capture, streptavidin magnetic beads.

MassSpecPreppy-An end-to-end solution for automated protein concentration determination and flexible sample digestion for proteomics applications.

In proteomics, fast, efficient, and highly reproducible sample preparation is of utmost importance, particularly in view of fast scanning mass spectrometers enabling analyses of large sample series. To address this need, we have developed the web application MassSpecPreppy that operates on the open science OT-2 liquid handling robot from Opentrons. This platform can prepare up to 96 samples at once, performing tasks like BCA protein concentration determination, sample digestion with normalization, reduction/alkylation and peptide elution into vials or loading specified peptide amounts onto Evotips in an automated and flexible manner. The performance of the developed workflows using MassSpecPreppy was compared with standard manual sample preparation workflows. The BCA assay experiments revealed an average recovery of 101.3% (SD:±7.82%) for the MassSpecPreppy workflow, while the manual workflow had a recovery of 96.3% (SD:±9.73%). The species mix used in the evaluation experiments showed that 94.5% of protein groups for OT-2 digestion and 95% for manual digestion passed the significance thresholds with comparable peptide level coefficient of variations. These results demonstrate that MassSpecPreppy is a versatile and scalable platform for automated sample preparation, producing injection-ready samples for proteomics research.

A-389 Implementation of a solicited laboratory workflow improves quality indicators for point of care blood gas testing by Respiratory Therapists

Abstract Background Quality indicators (QIs) including turnaround-time (TAT), critical result notification, corrected patient results, and unacceptable specimen are key measures of a laboratory quality management program that interrogates the pre-analytic, analytic, and post-analytic phases of the testing process. It is essential that high quality standards are maintained for critical care testing such as blood gas analysis. Achieving these high standards is difficult enough with specially trained laboratory personnel and can especially challenging with respiratory therapists (RTs) who are tasked with the difficulty of balancing patient testing while providing care. Tracking these QIs can also be cumbersome, especially if using manual methods. This retrospective study evaluates the effect of implementing a solicited laboratory information system (LIS) workflow using EPIC-Beaker on blood gas testing QIs performed at the point-of-care (POC) by RTs. Methods Blood gas QIs including TAT, percent of critical results notified, and number of corrected patient reports were compared between two workflows used by RTs. Metrics obtained between 1/1/2021–1/31/22 were from 13 POC blood gas analyzers from manufacturer #1, with manual TAT review of a sampling of specimens (5 days’ worth of samples per month) performed in manufacturer #1 middleware; manual review critical result notifications; and manually completed cards to track corrected reports. Metrics obtained between 2/1/22–12/31/22 were from 13 newly implemented POC blood gas analyzers from manufacturer #2, using a solicited EPIC-Beaker workflow with barcode-enabled labels for specimen tracking of collect and received times, automated review of TAT (received to result, collect to result, and collect to receive) with reports, and utilization of the communication log activity to document critical result notifications and unacceptable specimen. LIS reporting tools were used to summarize monthly TAT, critical result notification, unacceptable specimen, and corrected report QIs. Results Between the time periods employing the manual vs Beaker workflows, percent of results meeting the TAT were 99.5% (results within 30 minutes of receipt) vs 95.5% (results within 30 minutes of collection); percent of critical results notified was 100% vs 100%; and number of corrected reports was 132 vs 24 (demonstrating an 82% reduction), respectively. Estimated time spent reviewing/compiling QIs by RTs was ∼5 hours/week vs ∼3 hours/month with the manual vs EPIC-Beaker workflow. Conclusion Using an EPIC-Beaker solicited LIS workflow significantly improved the number of corrected patient reports compared to the previous workflow. Of the critical results audited each month, the percent with notification documentation remained the same between the workflows, however, using reporting tools allowed faster auditing of all critical results obtained each month compared to the manual workflow that only audited a sampling of the critical results per month due to the prohibitive time needed for manual review. TATs monitored also did not significantly change, however, the Beaker workflow allowed for simultaneous auditing of multiple measures of TAT, compared to the former workflow that only included electronic documentation of specimen ‘receipt’ by the RT and not the actual collect time. Employing automated LIS workflows may improve POC blood gas testing QIs while streamlining the process of compiling QIs.

Comparison of Nanotrap® Microbiome A Particles, membrane filtration, and skim milk workflows for SARS-CoV-2 concentration in wastewater.

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) RNA monitoring in wastewater has become an important tool for Coronavirus Disease 2019 (COVID-19) surveillance. Grab (quantitative) and passive samples (qualitative) are two distinct wastewater sampling methods. Although many viral concentration methods such as the usage of membrane filtration and skim milk are reported, these methods generally require large volumes of wastewater, expensive lab equipment, and laborious processes. The objectives of this study were to compare two workflows (Nanotrap® Microbiome A Particles coupled with MagMax kit and membrane filtration workflows coupled with RNeasy kit) for SARS-CoV-2 recovery in grab samples and two workflows (Nanotrap® Microbiome A Particles and skim milk workflows coupled with MagMax kit) for SARS-CoV-2 recovery in Moore swab samples. The Nanotrap particle workflow was initially evaluated with and without the addition of the enhancement reagent 1 (ER1) in 10 mL wastewater. RT-qPCR targeting the nucleocapsid protein was used for detecting SARS-CoV-2 RNA. Adding ER1 to wastewater prior to viral concentration significantly improved viral concentration results (P < 0.0001) in 10 mL grab and swab samples processed by automated or manual Nanotrap workflows. SARS-CoV-2 concentrations in 10 mL grab and Moore swab samples with ER1 processed by the automated workflow as a whole showed significantly higher (P < 0.001) results than 150 mL grab samples using the membrane filtration workflow and 250 mL swab samples using the skim milk workflow, respectively. Spiking known genome copies (GC) of inactivated SARS-CoV-2 into 10 mL wastewater indicated that the limit of detection of the automated Nanotrap workflow was ~11.5 GC/mL using the RT-qPCR and 115 GC/mL using the digital PCR methods. These results suggest that Nanotrap workflows could substitute the traditional membrane filtration and skim milk workflows for viral concentration without compromising the assay sensitivity. The manual workflow can be used in resource-limited areas, and the automated workflow is appropriate for large-scale COVID-19 wastewater-based surveillance.

Evaluation of a fully automated anatomical left atrial mapping pipeline enabled by artficial intelligence

Abstract Funding Acknowledgements Type of funding sources: Public Institution(s). Main funding source(s): Dr. Fabbricatore and Dr Valeriano have been supported by a research grant provided by the Cardiopath PhD program (University of Naples Federico II) Background Pulmonary vein isolation (PVI) is increasingly performed globally for the treatment of atrial fibrillation [1]. Using a standardized and reproducible workflow, we previously reported an average procedure time of 76.1 ± 26.2 min [2]. Over the last few years, due to workflow optimization and new technological developments, we were able to reduce the skin-to-skin time towards 47.3 ± 7.1 min (unpublished data). Our workflow implements rotational angiography (RA) of the left atrium (LA), which is segmented manually, then converted to a 3D mesh and finally imported into the electroanatomical mapping system (EAM). A second 3D mesh is created by cutting the left bottom anterior part of the LA, including the left atrial appendage, in order to clearly visualize the ridge (Figure 1). Recently, we developed an integrated artificial intelligence (AI) pipeline to automate this process (Figure 2). OBJECTIVE The aim of the study was to evaluate the efficiency of the new AI-based fully automated anatomical mapping of the LA and the efficacy of the automated segmentation. Methods The last 50 PVI procedures that were manually segmented in our lab were processed once more using the AI pipeline. The efficiency of the AI pipeline was assessed by comparing the elapsed time, for both our current manual workflow and the AI pipeline, between the reconstruction of the computed tomography (CT) volume from the 2D RA and the moment both 3D meshes are imported into the EAM. The efficacy of the pipeline was analyzed by calculating the Dice coefficient between the manual and automated segmentation as well as by visually grading segmentation quality on a 4-point Likert scale (1: Bad,2: Sufficient, 3: Good, 4: Excellent) by an experienced cardiac electrophysiologist. Significance testing of timing and visual grading differences between both workflows was performed using a one-sided Wilcoxon signed-rank test. Variables were summarized using median (Q1-Q3). Results The time to segment the LA, create the 3D meshes and import them into the EAM was significantly reduced from 560.0 (513.0–636.5) seconds to 52.0 (49.0–56.0) seconds for the manual and automated workflow respectively (p&amp;lt;0.0001). This reduction is equivalent to about 18% of the current skin-to-skin time for PVI in our lab. Furthermore, we reported a high Dice coefficient between the automated and manual segmentations valued at 0.928 (0.917–0.939). Finally, the subjective quality of the automated segmentation was superior compared to the manual segmentation, with respective values of 4 (3–4) and 3 (3–3.5) (p=0.006). Conclusion We developed an efficient and effective AI pipeline enabling fully automated anatomical mapping of the LA, potentially reducing our current PVI skin-to-skin time by 18%. The robustness of the AI segmentation model has been validated objectively (Dice coefficient) and subjectively. The efficacy of the AI cut model, the second model in the pipeline, remains to be verified.

Experimental Structural Model: From Manual Paper Garment to Fabrication as an Architectural Practice-Based Approach for Fashion Design Education

The study presents the integration of architectural design approaches in the fashion design process suggesting a new educational method based on structural model fabrication. The paper addresses the output of an experimental collaborative practice-based workshop titled ‘Fashion Clash’ that mixes both architects and fashion designers. The workshop focused on testing self-structural garments following a manual workflow which is divided into three main phases, (1) modeling and form-finding, (2) assembly, and (3) fabrication. Paper-based materials are used for transforming full-scale garments into textiles. The results presented seven garments displayed at a fashion show that show the effect of the folding techniques in reaching stability and highlighting the interdisciplinary integration of architects and fashion designers. The study concludes that implementing a parametric design logic based on architectural perspective in fashion would generate innovative ways of testing self-supporting geometry. Digitally computing the forces and structure before fabrication are left for further research.

Mapping Imprecision: How to Geocode Data from Inaccurate Historic Maps

This paper aims to present and discuss the method of geocoding historical place names from historic maps that cannot be georeferenced in the GIS environment. This concerns especially maps drawn in the early modern period, i.e., before the common use of precise topographic surveys. Such maps are valuable sources of place names and geocoding them is an asset to historical and geographical analyses. Geocoding is a process of matching spatial data (such as place names) with reference datasets (databases, gazetteers) and therefore giving them geographic coordinates. Such referencing can be done using multiple tools (online, desktop), reference datasets (modern, historical) and methods (manual, semi-automatic, automatic), but no suitable approach to handling inaccurate historic maps has yet been proposed. In this paper, selected geocoding strategies were described, as well as the author’s method of matching place names from inaccurate cartographic sources. The study was based on Charles Perthées maps of Polish palatinates (1:225,000, 1783–1804)—maps that are not mathematically precise enough to be georeferenced. The proposed semi-automatic and curated approach results in 85% accuracy. It reflects the manual workflow of historical geographers who identify place names with their modern counterparts by analysing their location and proper name.

SCIP: software for efficient clinical interpretation of copy number variants detected by whole-genome sequencing

Copy number variants (CNVs) represent major etiologic factors in rare genetic diseases. Current clinical CNV interpretation workflows require extensive back-and-forth with multiple tools and databases. This increases complexity and time burden, potentially resulting in missed genetic diagnoses. We present the Suite for CNV Interpretation and Prioritization (SCIP), a software package for the clinical interpretation of CNVs detected by whole-genome sequencing (WGS). The SCIP Visualization Module near-instantaneously displays all information necessary for CNV interpretation (variant quality, population frequency, inheritance pattern, and clinical relevance) on a single page—supported by modules providing variant filtration and prioritization. SCIP was comprehensively evaluated using WGS data from 1027 families with congenital cardiac disease and/or autism spectrum disorder, containing 187 pathogenic or likely pathogenic (P/LP) CNVs identified in previous curations. SCIP was efficient in filtration and prioritization: a median of just two CNVs per case were selected for review, yet it captured all P/LP findings (92.5% of which ranked 1st). SCIP was also able to identify one pathogenic CNV previously missed. SCIP was benchmarked against AnnotSV and a spreadsheet-based manual workflow and performed superiorly than both. In conclusion, SCIP is a novel software package for efficient clinical CNV interpretation, substantially faster and more accurate than previous tools (available at https://github.com/qd29/SCIP, a video tutorial series is available at https://bit.ly/SCIPVideos).