Machine Learning-Based Estimation of Experimental Artifacts and Image Quality in Fluorescence Microscopy - Supporting Data

<p>Supporting data for the reproduction of the results reported in Corbetta, E., Bocklitz, T., Machine learning based estimation of experimental artifacts and image quality in fluorescence microscopy (2024) [1].</p> <h2><strong>MM-IQA_Images_png and </strong><strong>MM-IQA_Images_tif</strong></h2> <p>Folder containing all the supporting datasets of the publication.</p> <ul> <li>images_manual_inspection: semisynthetic dataset used for the manual inspection of the quality metrics.</li> <li>images_lda_training: semisynthetic dataset used to train the Linear Discriminant Analysis (LDA) model.</li> <li>images_lda_prediction: datasets predicted by the LDA model. <ul> <li>images_experimental: every subfolder is a dataset composed of measurements of a different sample. Images are from publicly available datasets from [2] and [3].</li> <li>images_known_semisynthetic: knwon semisynthetic dataset used for prediction, assessment and interpretation of the trained model.</li> </ul> </li> </ul> <p>Tif files are the original data used for the study.</p> <h2><strong>MM-IQA_Source_data</strong></h2> <p>Folder containing all the supporting metadata of the publication.</p> <ul> <li>manual_inspection: quality metrics computed for the semisynthetic dataset used for the manual inspection. <ul> <li>manual_inspection_bg: indices for the selection of the background region in each sample.</li> <li>manual_inspection_free_parameters: parameters used for the generation of the simulated artifacts.</li> <li>manual_inspection_metrics: quality metrics computed for the dataset, used for the manual inspection.</li> <li>manual_inspection_samples: free parameters associated to each image of the dataset for manual inspection.</li> </ul> </li> <li>LDA_training: semisynthetic dataset used to train the Linear Discriminant Analysis (LDA) model. <ul> <li>lda_metrics_synthetic+semisynthetic_uniform_max: quality metrics computed for the training dataset, with maximum normalization of the images. (Not used in the manuscript)</li> <li>lda_metrics_synthetic+semisynthetic_uniform_rescale01: quality metrics computed for the training dataset, with image values rescaled between 0 and 1.</li> <li>parameters_all_degradations: parameters used for the generation of the simulated artifacts.</li> </ul> </li> <li>LDA_prediction: datasets predicted by the LDA model. <ul> <li>experimental: quality metrics computed for measurements of different samples. Images are from publicly available datasets from [2] and [3].</li> <li>known_semisynthetic: metadata for the knwon semi-synthetic dataset used for prediction, assessment and interpretation of the trained model: <ul> <li>known_semisynthetic_free_parameters: parameters used for the generation of the simulated artifacts.</li> <li>known_semisynthetic_metrics_rescale01: quality metrics computed for the training dataset, with image values rescaled between 0 and 1.</li> <li>known_semisynthetic_maxnorm_lda_results: lda prediction results, when metrics are maximum normalized to the training dataset.</li> <li>known_semisynthetic_znorm_lda_results: lda prediction results, when metrics are z-score normalized to the training dataset.</li> </ul> </li> </ul> </li> </ul> <p>Source data can be used to reproduce the results of the manuscript, using the codes shared in the public GitLab repository&nbsp;<em><a href="https://git.photonicdata.science/elena.corbetta/multi-marker-iqa" target="_blank" rel="noopener">multi-marker-IQA</a>.</em></p> <h3><em>How to use the source data</em></h3> <p>The following table describes which data can be used in the scripts provided in the public GitLab repository&nbsp;<em><a href="https://git.photonicdata.science/elena.corbetta/multi-marker-iqa" target="_blank" rel="noopener">multi-marker-IQA</a>.</em></p> <table> <tbody> <tr> <td><strong>Script</strong></td> <td><strong>Data to use</strong></td> <td><strong>Details</strong></td> </tr> <tr> <td>01_quality_metrics</td> <td>Subfolders of MM-IQA_Images_png</td> <td>Include all the images to evaluate in a single subfolder in&nbsp;<code>/test_images</code></td> </tr> <tr> <td>&nbsp;</td> <td>background_idx.xlsx</td> <td>The indices for the samples to evaluate must be included in the table</td> </tr> <tr> <td> <p>01_quality_metrics_visualization</p> <p>01_quality_metrics_visualization_notebook</p> </td> <td>manual_inspection_metrics.xlsx</td> <td>&nbsp;</td> </tr> <tr> <td>&nbsp;</td> <td>known_semisynthetic_metrics_rescale01</td> <td>&nbsp;</td> </tr> <tr> <td>&nbsp;</td> <td>Every metadata included in LDA_predcition/experimental/</td> <td>&nbsp;</td> </tr> <tr> <td> <p>02_lda_training+prediction</p> </td> <td>lda_metrics_synthetic+semisynthetic_uniform_rescale01</td> <td>As training dataset</td> </tr> <tr> <td>&nbsp;</td> <td>known_semisynthetic_metrics_rescale01</td> <td>As prediction dataset</td> </tr> <tr> <td>&nbsp;</td> <td>Every metadata included in LDA_predcition/experimental/</td> <td>As prediction dataset</td> </tr> <tr> <td> <p>02_lda_visualization</p> <p>02_lda_visualization_notebook</p> </td> <td>known_semisynthetic_maxnorm_lda_results</td> <td>&nbsp;</td> </tr> <tr> <td>&nbsp;</td> <td>known_semisynthetic_znorm_lda_results</td> <td>&nbsp;</td> </tr> <tr> <td>Notebook_test-iqa</td> <td>A small dataset with image data and the relative background index, if available.</td> <td>Use a limited number of images.</td> </tr> <tr> <td>Notebook_mm-iqa_workflow</td> <td>Any image dataset with the relative background indices</td> <td>For quality assessment and as prediction dataset</td> </tr> <tr> <td>&nbsp;</td> <td>lda_metrics_synthetic+semisynthetic_uniform_rescale01</td> <td>As training dataset</td> </tr> </tbody> </table> <p>&nbsp;</p> <h2>MM-IQA_Scripts</h2> <ul> <li><strong>multi-marker-iqa-main</strong>: original GitLab repository for MM-IQA, version available at the date of manuscript publication.</li> <li><strong>Notebooks_peer_review</strong>: additional notebooks generated during the peer-review process with the computation of metrics for natural images and correlation measures.</li> </ul>