This dataset underlies AfriDx D6.7 report on Clinical study of NAT in COVID-19.

Summary of Project

The AfriDx Project comprises nucleic acid testing (NAT) for COVID-19 using the PATHPOD system and compared with the RT-qPCR as the gold standard. KNUST, KCCR and NMIMR received PATHPOD and cartridges from the DTU (see D2.2). The cartridges for the testing were shipped in two (2) batches. Training on PATHPOD usage was done and used for testing covid-19 samples. Data obtained was compared with the gold standard RT-PCR.

The overall the testing against RT-LAMP was circa 60% sensitive, but the specificity dropped from 80.0% in the first batch to 24% in the second batch. Analysis of the factors causing the drop in specificity is on-going.

In the first batch a total of 1,947 tests were conducted, with 531 positive and 1416 negative results, producing 302 samples returning a false positive (FP) 133 samples giving a false negative (FN). The data showed that the FNs could be related to the copy number of virus in the sample, with a strong correlation with RT-PCR C_T value and true positive/false negative LAMP outcome. The second batch of nucleic acid testing recorded a total of 1,612 test for N-gene in Ghana with 1208 positive and 404 negative tests. This showed an exceptionally high occurrence of false positive test results: 1134 (76.2%) samples returned a false positive (FP) compared with the RT-PCR and 49 (3.9%) samples gave a false negative (FN). Nevertheless, the correlation with C_T remained, suggesting that the PATHPOD was functioning correctly and that the results were revealing some contamination or deterioration of reagents or sample.

Some initial analysis of the raw data from PATHPOD revealed some characteristics of sample and/or reagent contamination as well as the outcome of a poorly sealed cartridge, that would result in an erroneous signal. Further analysis is needed to fully understand any design modifications that might be beneficial. The impact of shipping and storage on the cartridges also has potential impact and it is particularly noteworthy that the second batch of testing, using cartridges from the same manufacturing run as the first batch, performed less well.

Methodology

The PATHPOD equipment was placed on a clean and flat surface and switch on with the knob located at the back of equipment to turn on the equipment. The oropharyngeal sample in 300ul of PBS was heated at 95 ⁰C for 5min to inactivate virus. The master mix room table was disinfected with suitable disinfectant against DNA/RNA contamination. Wearing appropriate gloves, the Pathpod cartridge was removed from the refrigerator and allow 15-30 min at room temperature for the cartridge to acclimatize, and place in the chip holder. The attached temporary sealer film covering the wells was removed and discarded.

After short vortex of sample, 6 µl of sample and/or controls were added directly into the center of the well. the yellow paper from the PCR film was removed and placed over the film chip. The film was sealed properly to the chip using a soft roller ready for processing in the PATHPOD system.

Using the Pathpod keyboard, sample ID was entered and COV assay was selected. The start bottom was pressed to heat the machine, thereafter the cartridge was inserted and start bottom was pressed again to run the program. When the assay was completed, result was read from both the screen and the LEDs next to the keyboard. Ensuring that the position on the machine matches the position on the chip. The following interpretation was inferred as results:

GREEN LIGHT: NEGATIVE BLINKING RED LIGHT: POSITIVE YELLOW LIGHT: RE-TEST THE SAMPLE

Datasets

The current dataset contributes to assess the accuracy of the Empatica 4 (E4) wristband for the detection of heart rate variability (HRV) and electrodermal activity (EDA) metrics in stress-inducing conditions and growing-risk driving scenarios. Heart Rate Variability (HRV) and ElectroDermal Activity (EDA) signals were recorded over six experimental conditions (i.e., Baseline, Video Clip, Scream, No Risk Driving, Low-Risk Driving, and High-Risk Driving) and by means of two measurement systems: the E4 device and a gold standard system. The raw quality of the physiological signals was enhanced by means of robust semi-automatic reconstruction algorithms. Heart Rate Variability time-domain parameters showed high accuracy in motion-free experimental conditions, while Heart Rate Variability frequency-domain parameters reported sufficient accuracy in almost every experimental condition.

Folder 01 contains both HRV and EDA parameters for every experimental condition, according to the Gold Standard measurement system and the Empatica 4 device, in two separate Excel files.

Folder 02 contains supplementary material on the assessment of the signals quality.

Folder 03 contains the Bland-Altman plot for each HRV and EDA parameter and for each condition (1 .png file per each parameter), and an excel file that resumes the Bland-Altman analyses numerical outcomes.

Vision represents the most important sense of primates. To understand visual processing, various different methods are employed, e.g. electrophysiology, psychophysics or eye tracking. For the latter, researchers recently began to step outside the artificial environments of laboratory setups towards the more natural conditions we usually face in the real world. This approach was enabled by sophisticated video-based mobile eye-trackers. Yet, their mobility and light weight often comes at an expense of reduced performance and accuracy, compared to the stationary eye-trackers, which nowadays are considered a gold standard in vision science.

In order to get a better understanding of the advantages and limitations of both eye-tracking techniques, we quantitatively compared one of the most advanced mobile eye-trackers available, the EyeSeeCam (ESC, sampling at 230 Hz), with a commonly used laboratory eye-tracker, the EyeLink II (ELII, sampling at 500 Hz). We aimed to investigate whether or not fully mobile eye-trackers are capable of providing adequate data to allow for a direct comparison with data recorded with stationary eye-trackers. Therefore, we recorded three different, commonly used eye movements, i.e. fixation, saccades and smooth pursuit eye movements, with both eye trackers in successive standardized paradigms in a laboratory setting with eight human subjects.

Despite major technical differences, the values of most eye movement parameters were not statistically different between both systems. Differences could only be found in overall gaze accuracy and for time critical parameters like saccade duration, for which a higher sample frequency is especially useful. Although, the stationary ELII system proved to be superior, especially at a single subject or even at a single trial basis, the ESC showed a similar performance for averaged parameters across trials and subjects. We conclude that modern high-performance mobile eye-trackers are well suited to provide reliable oculomotor data at the required spatial and temporal resolution.

Here we provide the raw data recorded with both eye trackers in eights human subjects in the three different, commonly used, oculomotor tasks.

This research developed a Kencorpus Swahili Question Answering Dataset KenSwQuAD from raw data of Swahili language, which is a low resource language predominantly spoken in Eastern African and also has speakers in other parts of the world. Question Answering datasets are important for machine comprehension of natural language processing tasks such as internet search and dialog systems. However, before such machine learning systems can perform these tasks, they need training data such as the gold standard Question Answering (QA) set developed in this research. The research engaged annotators to formulate question answer pairs from Swahili texts that had been collected by the Kencorpus project, a Kenyan languages corpus that collected data from three Kenyan languages. The total Swahili data collection had 2,585 texts, out of which we annotated 1,445 story texts with at least 5 QA pairs each, resulting into a final dataset of 7,526 QA pairs. A quality assurance set of 12.5% of the annotated texts was subjected to re-evaluation by different annotators who confirmed that the QA pairs were all correctly annotated. A proof of concept on applying the set to machine learning on the question answering task confirmed that the dataset can be used for such practical tasks. The research therefore developed KenSwQuAD, a question-answer dataset for Swahili that is useful to the natural language processing community who need training and gold standard sets for their machine learning applications. The research also contributed to the resourcing of the Swahili language which is important for communication around the globe. Updating this set and providing similar sets for other low resource languages is an important research area that is worthy of further research. Acknowledgement of annotators: Rose Felynix Nyaboke, Alice Gachachi Muchemi, Patrick Ndung'u, Eric Omundi Magutu, Henry Masinde, Naomi Muthoni Gitau, Mark Bwire Erusmo, Victor Orembe Wandera, Frankline Owino, Geoffrey Sagwe Ombui

This repository contains:

1. data.tar.gz: raw data from Auditory Brainstem Response (ABR) measurements performed at over 4,000 mice at the German Mouse Clinic. It also contains suitably pre-processed ABR raw data of over 8,000 mice from another public data repository (10.5061/DRYAD.CV803RV) of the Wellcome Sanger Institute. All data is intended to be used by and compatible with code published on GitHub (https://github.com/ExperimentalGenetics/ABR_thresholder).
   
2. results.tar.gz: results and visualisations - comparison of two new, independent automatic hearing threshold finding methods and comparison with the manual gold standard method.

Dataset of heart rate measurements collected from the PineTime wristband, with a gold standard reference.

Contents

The repository contains both the raw and the "merged", clean data. The merged data is much easier to work with and should be used when building machine learning models. The raw data is provided for transparency, reproducibility, and to allow for studies that could use the other data collected from the Equivital device.

• schedule.md – schedule of the study, indicating the start and end times of each exercise and break.
• data_raw/ – raw data collected from the PineTime wristband and the Equivital device. Each subdirectory corresponds to one participant. The files are in the Feather format.
• data_merged/ – merged data series that can be used for building ML models. The files are in JSON format and follow a nested structure, where each heart rate measurement is associated with a series of acceleration measurements that preceded it. Each file corresponds to one continuous measurement session – there are sometimes multiple sessions per participant due to intermittent hardware failures.

Citation

If you use this data in research works, please cite the following paper:

Sowiński, P., Rachwał, K., Danilenka, A., Bogacka, K., Kobus, M., Dąbrowska, A., Paszkiewicz, A., et al. (2023). Frugal Heart Rate Correction Method for Scalable Health and Safety Monitoring in Construction Sites. Sensors, 23(14), 6464. MDPI AG. Retrieved from http://dx.doi.org/10.3390/s23146464

BibTeX:

@article{sowinski2023frugal,
 title={Frugal Heart Rate Correction Method for Scalable Health and Safety Monitoring in Construction Sites},
 author={Sowi{\'n}ski, Piotr and Rachwa{\l}, Kajetan and Danilenka, Anastasiya and Bogacka, Karolina and Kobus, Monika and D{\k{a}}browska, Anna and Paszkiewicz, Andrzej and Bolanowski, Marek and Ganzha, Maria and Paprzycki, Marcin},
 journal={Sensors},
 volume={23},
 number={14},
 pages={6464},
 year={2023},
 publisher={MDPI},
 url = {https://www.mdpi.com/1424-8220/23/14/6464},
 doi = {10.3390/s23146464}
}

Authors

• Monika Kobus – data collection
• Anna Dąbrowska – data collection, methodological supervision
• Piotr Sowiński – data collection and processing

Acknowledgements

This work is part of the ASSIST-IoT project that has received funding from the EU’s Horizon 2020 research and innovation programme under grant agreement No 957258.

The Central Institute for Labour Protection – National Research Institute provided facilities and equipment for data collection.

License

The dataset is licensed under the Creative Commons Attribution 4.0 International License.

CoNLL 2017 and 2018 shared tasks: Multilingual Parsing from Raw Text to Universal Dependencies

This package contains the test data in the form in which they ware presented to the participating systems: raw text files and files preprocessed by UDPipe. The metadata.json files contain lists of files to process and to output; README files in the respective folders describe the syntax of metadata.json.

For full training, development and gold standard test data, see Universal Dependencies 2.0 (CoNLL 2017) Universal Dependencies 2.2 (CoNLL 2018) See the download links at http://universaldependencies.org/.

For more information on the shared tasks, see http://universaldependencies.org/conll17/ http://universaldependencies.org/conll18/

Contents:

conll17-ud-test-2017-05-09 ... CoNLL 2017 test data conll18-ud-test-2018-05-06 ... CoNLL 2018 test data conll18-ud-test-2018-05-06-for-conll17 ... CoNLL 2018 test data with metadata and filenames modified so that it is digestible by the 2017 systems.

BackgroundTreatment of nerve injuries proves to be a worldwide clinical challenge. Acellular nerve allografts are suggested to be a promising alternative for bridging a nerve gap to the current gold standard, an autologous nerve graft.ObjectiveTo systematically review the efficacy of the acellular nerve allograft, its difference from the gold standard (the nerve autograft) and to discuss its possible indications.Material and methodsPubMed, Embase and Web of Science were systematically searched until the 4th of January 2022. Original peer reviewed paper that presented 1) distinctive data; 2) a clear comparison between not immunologically processed acellular allografts and autologous nerve transfers; 3) was performed in laboratory animals of all species and sex. Meta analyses and subgroup analyses (for graft length and species) were conducted for muscle weight, sciatic function index, ankle angle, nerve conduction velocity, axon count diameter, tetanic contraction and amplitude using a Random effects model. Subgroup analyses were conducted on graft length and species.ResultsFifty articles were included in this review and all were included in the meta-analyses. An acellular allograft resulted in a significantly lower muscle weight, sciatic function index, ankle angle, nerve conduction velocity, axon count and smaller diameter, tetanic contraction compared to an autologous nerve graft. No difference was found in amplitude between acellular allografts and autologous nerve transfers. Post hoc subgroup analyses of graft length showed a significant reduced muscle weight in long grafts versus small and medium length grafts. All included studies showed a large variance in methodological design.ConclusionOur review shows that the included studies, investigating the use of acellular allografts, showed a large variance in methodological design and are as a consequence difficult to compare. Nevertheless, our results indicate that treating a nerve gap with an allograft results in an inferior nerve recovery compared to an autograft in seven out of eight outcomes assessed in experimental animals. In addition, based on our preliminary post hoc subgroup analyses we suggest that when an allograft is being used an allograft in short and medium (0-1cm, > 1-2cm) nerve gaps is preferred over an allograft in long (> 2cm) nerve gaps.

This record contains raw data related to the article "Predicting Long-Term Mortality in TAVI Patients Using Machine Learning Techniques"

Background: Whereas transcatheter aortic valve implantation (TAVI) has become the gold standard for aortic valve stenosis treatment in high-risk patients, it has recently been extended to include intermediate risk patients. However, the mortality rate at 5 years is still elevated. The aim of the present study was to develop a novel machine learning (ML) approach able to identify the best predictors of 5-year mortality after TAVI among several clinical and echocardiographic variables, which may improve the long-term prognosis. Methods: We retrospectively enrolled 471 patients undergoing TAVI. More than 80 pre-TAVI variables were collected and analyzed through different feature selection processes, which allowed for the identification of several variables with the highest predictive value of mortality. Different ML models were compared. Results: Multilayer perceptron resulted in the best performance in predicting mortality at 5 years after TAVI, with an area under the curve, positive predictive value, and sensitivity of 0.79, 0.73, and 0.71, respectively. Conclusions: We presented an ML approach for the assessment of risk factors for long-term mortality after TAVI to improve clinical prognosis. Fourteen potential predictors were identified with the organic mitral regurgitation (myxomatous or calcific degeneration of the leaflets and/or annulus) which showed the highest impact on 5 years mortality.

README

The Bitbrain Open Access Sleep (BOAS) dataset.

Overview

This project aimes at bridging the gap between gold-standard clinical sleep monitoring and emerging wearable EEG technologies. The dataset comprises 128 nights in which healthy participants were simultaneously monitored with two technologies: a Brain Quick Plus Evolution PSG system, by Micromed, and a wearable EEG headband, by Bitbrain. The Micromed PSG system provides a comprehensive and clinically validated set of sleep parameters, while the Bitbrain wearable EEG headband offers a user-friendly, self-administered alternative, limited to forehead EEG electrodes.

A relevant aspect of the dataset is the simultaneous acquisition of data from both systems, allowing for direct comparison and validation of the wearable EEG device against the established PSG standard. This dual-recording approach provides a rich resource for evaluating the performance and potential of wearable EEG technology in sleep studies.

To ensure robust and reliable sleep staging, we employed a rigorous labeling process. Three expert sleep scorers independently annotated the PSG recordings following the criteria developed by the American Academy of Sleep Medicine (AASM) (Berry et al., 2015), and a consensus label was derived from these annotations by a fourth expert. This consensus labeling approach addresses the inherent variability in human sleep staging, which has an estimated inter-scorer agreement of approximately 85% (Danker-Hopfe et al., 2009; Rosenberg and Van Hout, 2013). The consensus labels were then applied to the corresponding wearable EEG recordings, leveraging the simultaneous data acquisition. Moreover, we utilized a deep learning model to analyze the dataset (Esparza-Iaizzo et al., 2024). By implementing a cross-validation procedure, we trained and validated the model separately on the PSG and wearable EEG datasets. The model achieved an 87.08% match between the human-consensus labels and the network-provided labels for the PSG data, and an 86.64% match for the wearable EEG data.

Our dataset, therefore, includes:

1. Raw and labeled PSG recordings from 128 nights.

2. Raw and labeled wearable EEG recordings from the same nights.

3. Human-consensus sleep stage labels for both PSG and wearable EEG data.

4. AI-generated sleep stage labels for both datasets.

Format

The dataset is formatted according to the Brain Imaging Data Structure.

The folder of each participant contains the data recorded with the PSG (sub-xx_task-Sleep_acq-psg_eeg.edf) and with the wearable EEG headband (sub-xx_task-Sleep_acq-headband_eeg.edf). Note that not all the PSG sensors were used with all the participants. The full list of available sources of activity for each recording can be obtained on the 'channels.tsv' file.

Meaning of all the channel groups:

- PSG_EEG: EEG signals recorded with the PSG system. Channels available are F3, F4, C3, C4, O1, O2 (PSG_F3;PSG_F4;PSG_C3;PSG_C4;PSG_O1;PSG_O2).
- PSG_EOG: EOG signals recorded with the PSG system. Some of the participants have just one EOG derivation (PSG_EOG), whereas others have 2 lateral derivations, referenced to the mastoid (PSG_EOGR;PSG_EOGL).
- PSG_EMG: EMG signals recorded with the PSG system. One chin EMG derivation is available (PSG_EMG).
- PSG_BELTS: Breathing activity recorded the PSG system using breathing belts. Abdominal and thoracic breathing belts were used (PSG_ABD;PSG_THOR).
- PSG_THER: Respiratory airflow recorded with the PSG system using a thermistor (PSG_THER).
- PSG_CAN: Respiratory airflow recorded with the PSG system using a nasal cannula (PSG_CAN).
- PSG_PPG: Photopletismographic activity recorded with the PSG system. Channels available are pulse (PSG_PULSE), heart beat (PSG_BEAT) and oxygen saturation (PSG_SPO2).

- HB_EEG: EEG signals recorded with the wearable EEG headband. The headband measuring locaations are equivalent to AF7 and AF8 (HB_1;HB_2).
- HB_IMU: Movement activity recorded from an accelerometer and gyroscope. Both sensors record in 3 dimensions (x, y, z) each, (HB_IMU_1;HB_IMU_2;HB_IMU_3;HB_IMU_4;HB_IMU_5;HB_IMU_6)
- HB_PULSE: Pulse activity recorded with the wearable EEG headband using a PPG sensor (HB_PULSE).

The sleep stages of each night are coded as events at the corresponding recording folder. The sleep stages obtained as the consensus of the three experts, as well as the labels obtained by the AI using the EEG activity recorded with the PSG can be found in the 'sub-xx_task-Sleep_acq-psg_events.tsv' files. The sleep stages obtained by the AI using the EEG activity recorded with the wearable headband can be found in the 'sub-xx_task-Sleep_acq-headband_events.tsv' files.

References

Berry, R. B., Brooks, R., Gamaldo, C. E., Harding, S. M., Lloyd, R. M., Marcus, C. L., et al. (2015). The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications, Version 2.2. Darien, Illinois.

Danker-Hopfe, H., Anderer, P., Zeitlhofer, J., Boeck, M., Dorn, H., Gruber, G., et al. (2009). Interrater reliability for sleep scoring according to the Rechtschaffen & Kales and the new AASM standard. J. Sleep Res. 18, 74–84. doi: 10.1111/j.1365-2869.2008.00700.x.

Esparza-Iaizzo, M., Sierra-Torralba, M., Klinzing, J. G., Minguez, J., Montesano, L., and López-Larraz, E. (2024). Automatic sleep scoring for real-time monitoring and stimulation in individuals with and without sleep apnea. bioRxiv, 2024.06.12.597764. doi: 10.1101/2024.06.12.597764.

Rosenberg, R. S., and Van Hout, S. (2013). The American Academy of Sleep Medicine inter-scorer reliability program: sleep stage scoring. J. Clin. sleep Med. 9, 81–87. doi: 10.5664/jcsm.2350.

Contact

If you have any questions or comments, please contact:

Eduardo López-Larraz: eduardo.lopez@bitbrain.com Jens G. Klinzing: jens.klinzing@bitbrain.com

Corpus for the ICDAR2019 Competition on Post-OCR Text Correction (October 2019)Christophe Rigaud, Antoine Doucet, Mickael Coustaty, Jean-Philippe Moreuxhttp://l3i.univ-larochelle.fr/ICDAR2019PostOCR-------------------------------------------------------------------------------These are the supplementary materials for the ICDAR 2019 paper ICDAR 2019 Competition on Post-OCR Text CorrectionPlease use the following citation:@inproceedings{rigaud2019pocr,title=""ICDAR 2019 Competition on Post-OCR Text Correction"",author={Rigaud, Christophe and Doucet, Antoine and Coustaty, Mickael and Moreux, Jean-Philippe},year={2019},booktitle={Proceedings of the 15th International Conference on Document Analysis and Recognition (2019)}}

Description: The corpus accounts for 22M OCRed characters along with the corresponding Gold Standard (GS). The documents come from different digital collections available, among others, at the National Library of France (BnF) and the British Library (BL). The corresponding GS comes both from BnF's internal projects and external initiatives such as Europeana Newspapers, IMPACT, Project Gutenberg, Perseus and Wikisource. Repartition of the dataset- ICDAR2019_Post_OCR_correction_training_18M.zip: 80% of the full dataset, provided to train participants' methods.- ICDAR2019_Post_OCR_correction_evaluation_4M: 20% of the full dataset used for the evaluation (with Gold Standard made publicly after the competition).- ICDAR2019_Post_OCR_correction_full_22M: full dataset made publicly available after the competition. Special case for Finnish language Material from the National Library of Finland (Finnish dataset FI > FI1) are not allowed to be re-shared on other website. Please follow these guidelines to get and format the data from the original website.1. Go to https://digi.kansalliskirjasto.fi/opendata/submit?set_language=en;2. Download OCR Ground Truth Pages (Finnish Fraktur) [v1](4.8GB) from Digitalia (2015-17) package;3. Convert the Excel file ""~/metadata/nlf_ocr_gt_tescomb5_2017.xlsx"" as Comma Separated Format (.csv) by using save as function in a spreadsheet software (e.g. Excel, Calc) and copy it into ""FI/FI1/HOWTO_get_data/input/"";4. Go to ""FI/FI1/HOWTO_get_data/"" and run ""script_1.py"" to generate the full ""FI1"" dataset in ""output/full/"";4. Run ""script_2.py"" to split the ""output/full/"" dataset into ""output/training/"" and ""output/evaluation/"" sub sets.At the end of the process, you should have a ""training"", ""evaluation"" and ""full"" folder with 1579528, 380817 and 1960345 characters respectively.

Licenses: free to use for non-commercial uses, according to sources in details- BG1: IMPACT - National Library of Bulgaria: CC BY NC ND- CZ1: IMPACT - National Library of the Czech Republic: CC BY NC SA- DE1: Front pages of Swiss newspaper NZZ: Creative Commons Attribution 4.0 International (https://zenodo.org/record/3333627)- DE2: IMPACT - German National Library: CC BY NC ND- DE3: GT4Hist-dta19 dataset: CC-BY-SA 4.0 (https://zenodo.org/record/1344132)- DE4: GT4Hist - EarlyModernLatin: CC-BY-SA 4.0 (https://zenodo.org/record/1344132)- DE5: GT4Hist - Kallimachos: CC-BY-SA 4.0 (https://zenodo.org/record/1344132)- DE6: GT4Hist - RefCorpus-ENHG-Incunabula: CC-BY-SA 4.0 (https://zenodo.org/record/1344132)- DE7: GT4Hist - RIDGES-Fraktur: CC-BY-SA 4.0 (https://zenodo.org/record/1344132)- EN1: IMPACT - British Library: CC BY NC SA 3.0- ES1: IMPACT - National Library of Spain: CC BY NC SA- FI1: National Library of Finland: no re-sharing allowed, follow the above section to get the data. (https://digi.kansalliskirjasto.fi/opendata)- FR1: HIMANIS Project: CC0 (https://www.himanis.org)- FR2: IMPACT - National Library of France: CC BY NC SA 3.0- FR3: RECEIPT dataset: CC0 (http://findit.univ-lr.fr)- NL1: IMPACT - National library of the Netherlands: CC BY- PL1: IMPACT - National Library of Poland: CC BY- SL1: IMPACT - Slovak National Library: CC BY NCText post-processing such as cleaning and alignment have been applied on the resources mentioned above, so that the Gold Standard and the OCRs provided are not necessarily identical to the originals.

Structure- **Content** [./lang_type/sub_folder/#.txt] - ""[OCR_toInput] "" => Raw OCRed text to be de-noised. - ""[OCR_aligned] "" => Aligned OCRed text. - ""[ GS_aligned] "" => Aligned Gold Standard text.The aligned OCRed/GS texts are provided for training and test purposes. The alignment was made at the character level using ""@"" symbols. ""#"" symbols correspond to the absence of GS either related to alignment uncertainties or related to unreadable characters in the source document. For a better view of the alignment, make sure to disable the ""word wrap"" option in your text editor.The Error Rate and the quality of the alignment vary according to the nature and the state of degradation of the source documents. Periodicals (mostly historical newspapers) for example, due to their complex layout and their original fonts have been reported to be especially challenging. In addition, it should be mentioned that the quality of Gold Standard also varies as the dataset aggregates resources from different projects that have their own annotation procedure, and obviously contains some errors.

ICDAR2019 competitionInformation related to the tasks, formats and the evaluation metrics are details on :https://sites.google.com/view/icdar2019-postcorrectionocr/evaluation

References - IMPACT, European Commission's 7th Framework Program, grant agreement 215064 - Uwe Springmann, Christian Reul, Stefanie Dipper, Johannes Baiter (2018). Ground Truth for training OCR engines on historical documents in German Fraktur and Early Modern Latin. - https://digi.nationallibrary.fi , Wiipuri, 31.12.1904, Digital Collections of National Library of Finland- EU Horizon 2020 research and innovation programme grant agreement No 770299

Contact- christophe.rigaud(at)univ-lr.fr- antoine.doucet(at)univ-lr.fr- mickael.coustaty(at)univ-lr.fr- jean-philippe.moreux(at)bnf.frL3i - University of la Rochelle, http://l3i.univ-larochelle.frBnF - French National Library, http://www.bnf.fr

Advanced therapy medicinal products (ATMP) are required to maintain their quality and safety throughout the production cycle, and they must be free of microbial contaminations. Among them, mycoplasma contaminations are difficult to detect and undesirable in ATMP, especially for immunosuppressed patients. Mycoplasma detection tests suggested by European Pharmacopoeia are the "culture method" and "indicator cell culture method" which, despite their effectiveness, are time consuming and laborious. Alternative methods are accepted, provided they are adequate and their results are comparable with those of the standard methods. To validate a novel in-house method, we performed and optimized, a real time PCR protocol, using a commercial kit and an automatic extraction system, in which we tested different volumes of matrix, maximizing the detection sensitivity. The results were compared with those obtained with the gold standard methods. From a volume of 10 ml, we were able to recognize all the mycoplasmas specified by the European Pharmacopoeia, defined as genomic copies per colony forming unit ratio (GC/CFU). Our strategy allows to achieve faster and reproducible results when compared with conventional methods and meets the sensitivity and robustness criteria required for an alternative approach to mycoplasmas detection for in-process and product-release testing of ATMP.