{"id":4950,"date":"2026-01-15T21:15:36","date_gmt":"2026-01-15T20:15:36","guid":{"rendered":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/?p=4950"},"modified":"2026-05-06T23:26:38","modified_gmt":"2026-05-06T21:26:38","slug":"segmentation-ai-author","status":"publish","type":"post","link":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/","title":{"rendered":"Segmentation AI Author Example: Train H&amp;E Vulvar Cancer AI Based on Aligned IHC Serial Sections"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">This <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/mikaia-university\/\">MIKAIA<sup>\u00ae<\/sup> University<\/a> app note illustrates how to interactively train a new H&amp;E segmentation AI with serial sections, without having to annotate tumor regions. Instead, an IHC epithelial marker (pan-cytokeratin AE1\/AE3) is thresholded through H-DAB stain unmixing to generate highly accurate tumor masks. These masks serve as training annotations in the Segmentation AI Author app to train the H&amp;E AI.<\/p>\n\n\n\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Dataset<\/strong>: Serial sections of vulvar cancer, kindly provided by <strong><a href=\"https:\/\/www.pathologie.uk-erlangen.de\/\">Universit\u00e4tsklinikum Erlangen<\/a><\/strong>. <br><strong>Markers<\/strong>: H&amp;E + <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/ihc-markers\/\">7x IHC<\/a> (AE1\/AE3 PCK, CD3, CD4, CD8, CD68, CD163, FoxP3).<\/p>\n\n\n\n<div id=\"ez-toc-container\" class=\"ez-toc-v2_0_83 counter-hierarchy ez-toc-counter ez-toc-grey ez-toc-container-direction\">\n<div class=\"ez-toc-title-container\">\n<p class=\"ez-toc-title\" style=\"cursor:inherit\">Table of Contents<\/p>\n<span class=\"ez-toc-title-toggle\"><a href=\"#\" class=\"ez-toc-pull-right ez-toc-btn ez-toc-btn-xs ez-toc-btn-default ez-toc-toggle\" aria-label=\"Toggle Table of Content\"><span class=\"ez-toc-js-icon-con\"><span class=\"\"><span class=\"eztoc-hide\" style=\"display:none;\">Toggle<\/span><span class=\"ez-toc-icon-toggle-span\"><svg style=\"fill: #999;color:#999\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" class=\"list-377408\" width=\"20px\" height=\"20px\" viewBox=\"0 0 24 24\" fill=\"none\"><path d=\"M6 6H4v2h2V6zm14 0H8v2h12V6zM4 11h2v2H4v-2zm16 0H8v2h12v-2zM4 16h2v2H4v-2zm16 0H8v2h12v-2z\" fill=\"currentColor\"><\/path><\/svg><svg style=\"fill: #999;color:#999\" class=\"arrow-unsorted-368013\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"10px\" height=\"10px\" viewBox=\"0 0 24 24\" version=\"1.2\" baseProfile=\"tiny\"><path d=\"M18.2 9.3l-6.2-6.3-6.2 6.3c-.2.2-.3.4-.3.7s.1.5.3.7c.2.2.4.3.7.3h11c.3 0 .5-.1.7-.3.2-.2.3-.5.3-.7s-.1-.5-.3-.7zM5.8 14.7l6.2 6.3 6.2-6.3c.2-.2.3-.5.3-.7s-.1-.5-.3-.7c-.2-.2-.4-.3-.7-.3h-11c-.3 0-.5.1-.7.3-.2.2-.3.5-.3.7s.1.5.3.7z\"\/><\/svg><\/span><\/span><\/span><\/a><\/span><\/div>\n<nav><ul class='ez-toc-list ez-toc-list-level-1 ' ><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-1\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_1_Align_serial_sections_with_MIKAIA%C2%AE_Slide_Align\" >Step 1: Align serial sections with MIKAIA\u00ae Slide Align<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-2\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_2_Open_aligned_serial_section_stack_in_MIKAIA%C2%AE\" >Step 2: Open aligned serial section stack in MIKAIA\u00ae<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-3\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_3_Mask_DAB_brown_regions_in_IHC_tumor_marker_PCK\" >Step 3: Mask DAB+ (brown) regions in IHC tumor marker (PCK)<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-4\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_4_Train_new_H_E_AI_with_Segmentation_AI_Author_using_the_DAB_masks_as_ground_truth\" >Step 4: Train new H&amp;E AI with Segmentation AI Author, using the DAB masks as ground truth<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-5\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_5_Use_newly_trained_AI_to_analyze_further_H_E-only_scans\" >Step 5: Use newly trained AI to analyze further H&amp;E-only scans<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-6\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_6_Quantify_immune_landscape_in_tumor-micro-environment_TME_including_video\" >Step 6: Quantify immune landscape in tumor-micro-environment (TME, including video)<\/a><\/li><li class='ez-toc-page-1 ez-toc-heading-level-2'><a class=\"ez-toc-link ez-toc-heading-7\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#Step_7_Cellular_neighborhood_analysis\" >Step 7: Cellular neighborhood analysis<\/a><\/li><\/ul><\/nav><\/div>\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_1_Align_serial_sections_with_MIKAIA%C2%AE_Slide_Align\"><\/span>Step 1: Align serial sections with MIKAIA<sup>\u00ae<\/sup> Slide Align<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The <strong><a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/mikaia-slide-align-overview\/\">MIKAIA<sup>\u00ae<\/sup><\/a> <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/mikaia-slide-align-overview\/\">Slide Align<\/a><\/strong> module is designed to co-align the serial sections from each case. The output is a deformed whole-slide-image (in SVS or OME-TIF format) per stain. Aligned sections from a case are linked together through a companion file (*.multifile.csv), which can be opened directly in MIKAIA<sup>\u00ae<\/sup>.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1920\" height=\"1032\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config.jpg\" alt=\"screenshot of MIKAIA\u00ae Slide Align\" class=\"wp-image-4951\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config.jpg 1920w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/01_Registration_Config-740x398.jpg 740w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/><figcaption class=\"wp-element-caption\">MIKAIA<sup>\u00ae<\/sup> Slide Align can be used to co-align the serial sections<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_2_Open_aligned_serial_section_stack_in_MIKAIA%C2%AE\"><\/span>Step 2: Open aligned serial section stack in MIKAIA<sup>\u00ae<\/sup><span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The screenshot below shows an H&amp;E image, with toolbar buttons for selecting other brightfield layers visible, similar to a <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/overview-of-spatial-analysis-apps-for-mif-slides\/\">multiplex IHC scan<\/a>. However, since each layer is brightfield, only one layer is displayed at a time, and they are not mixed together.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1920\" height=\"1032\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile.jpg\" alt=\"screenshot: Open aligned serial section stack in MIKAIA\u00ae\" class=\"wp-image-4952\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile.jpg 1920w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/02_Registration_Applied_ViewMultifile-740x398.jpg 740w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/><figcaption class=\"wp-element-caption\">The aligned multi-layer stack is opened in MIKAIA<sup>\u00ae<\/sup>. The H&amp;E layer is currently visible. IHC layers available in top left corner.<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">The individual aligned slides can also be opened directly and viewed side-by-side. o synchronize the viewers, enable the sync button in the bottom toolbar (indicated by the blue button). When the mouse hovers over the master viewer, the mouse pointer is mirrored in the other viewers.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" width=\"1024\" height=\"550\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-1024x550.jpg\" alt=\"screenshot of four aligned serial sections displayed side by side.\" class=\"wp-image-4953\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews-740x398.jpg 740w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/03_Registration_Applied_ViewMultipleViews.jpg 1920w\" sizes=\"(max-width: 1024px) 100vw, 1024px\" \/><figcaption class=\"wp-element-caption\">Four aligned serial sections displayed side by side. Viewers are synced<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_3_Mask_DAB_brown_regions_in_IHC_tumor_marker_PCK\"><\/span>Step 3: Mask DAB+ (brown) regions in IHC tumor marker (PCK)<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">To train the H&amp;E tumor segmentation AI, only the H&amp;E and the Pan-Cytokeratin AE1\/AE3 layers are used in the following steps. The <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/perineural-invasion\/\"><strong>Mask by Color App<\/strong><\/a> is employed to accuarately and automatically mask brown tumor regions, as shown in the screenshots below. This app offers various masking modes; in this case, it is configured to deconvolute the hematoxylin and DAB stains and then mask only the DAB stain through thresholding.<\/p>\n\n\n\n\t\t\t\n\t<style type=\"text\/css\">\n\t\t.slider-info-4965.bafg-slider-info .bafg-slider-title {\n\t\t\t\t\t\t\tfont-size:\n\t\t\t\t\t22px\t\t\t\t;\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\n\t\t.slider-info-4965.bafg-slider-info .bafg-slider-description {\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\t\t\n\t\t\n\t\t.slider-info-4965.bafg-slider-info .bafg_slider_readmore_button {\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\ttext-align: center;\n\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\n\t\t.slider-info-4965.bafg-slider-info .bafg_slider_readmore_button:hover {\n\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\t<\/style>\n\t\n\t\t\t<div class=\"bafg-twentytwenty-container slider-4965  \"\n\t\t\t\tbafg-orientation=\"horizontal\" bafg-default-offset=\"0.5\"\n\t\t\t\tbafg-before-label=\"IHC (PCK)\"\n\t\t\t\tbafg-after-label=\"H&amp;E\" bafg-overlay=\"1\"\n\t\t\t\tbafg-move-slider-on-hover=\"\"\n\t\t\t\tbafg-click-to-move=\"\">\n\n\t\t\t\t\t\t\t\t<img class=\"skip-lazy\" data-skip-lazy\t\t\t\t\tsrc=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/04_Generate_Mask_Automatically.jpg\" alt=\"\">\n\t\t\t\t<img class=\"skip-lazy\" data-skip-lazy\t\t\t\t\tsrc=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg\" alt=\"\">\n\n\t\t\t<\/div>\n\n\t\t\t\t<div class=\"bafg-slider-info-wraper\">\n\t\t<div style=\"\" class=\"slider-info-4965 bafg-slider-info\">\n\t\t\t\t\t<\/div>\n\t<\/div>\n\t\n\t\t\t<style type=\"text\/css\">\n\t\t\t\t\t\t\t\t\t\t\t<\/style>\n\t\t\t\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_4_Train_new_H_E_AI_with_Segmentation_AI_Author_using_the_DAB_masks_as_ground_truth\"><\/span>Step 4: Train new H&amp;E AI with Segmentation AI Author, using the DAB masks as ground truth<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The masks generated from the PCK IHC layer can now be applied to the H&amp;E layer. In the following screenshot, the annotations remain visible while the viewer shows the H&amp;E layer. This approach allows for the training of an H&amp;E segmentation AI using serial sections.s.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1920\" height=\"1032\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining.jpg\" alt=\"screenshot of masks derived from the PCK IHC layer\" class=\"wp-image-4956\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining.jpg 1920w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_02_ModelTraining-740x398.jpg 740w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/><figcaption class=\"wp-element-caption\">Masks derived from the PCK IHC layer are used as training annotations for a new AI created with MIKAIA<sup>\u00ae<\/sup> Segmentation AI Author<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">In the <strong>Segmentation AI Author App<\/strong>, a new model can now be created by adding three tissue classes: &#8220;Tumor&#8221;, &#8220;Healthy&#8221;, and &#8220;Background&#8221;. Instead of manualy annotating various tumor regions, we now use the PCK mask. Since the Segmentation AI Author App employs Few Shot Learning (&#8220;few shots&#8221; meaning few training annotations) and to limit the training duration, we select only a subset of the generated PCK masks that captur the tumor&#8217;s heterogeneity, discarding large portions of the mask that would not provide any additional training value anyway. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Additionally, we quickly outline a few non-tumor tissue regions as well as white background regions. Then, we click &#8220;Train&#8221;. The screenshot shows how the training annotations are analyzed tile by tile in order to derive and store a representation of each annotated tissue class. This step only takes approximately one minute, depending on the number and size of annotations.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_5_Use_newly_trained_AI_to_analyze_further_H_E-only_scans\"><\/span>Step 5: Use newly trained AI to analyze further H&amp;E-only scans<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The new Vulvar Cancer Segmentation AI, quickly created using the Segmentation AI Author, is now ready for application to additional H&amp;E-scans. The aligned IHC layer is no longer needed. Below is the segmentation for a field of view:<\/p>\n\n\n\n\t\t\t\n\t<style type=\"text\/css\">\n\t\t.slider-info-4963.bafg-slider-info .bafg-slider-title {\n\t\t\t\t\t\t\tfont-size:\n\t\t\t\t\t22px\t\t\t\t;\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\n\t\t.slider-info-4963.bafg-slider-info .bafg-slider-description {\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\t\t\n\t\t\n\t\t.slider-info-4963.bafg-slider-info .bafg_slider_readmore_button {\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\ttext-align: center;\n\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\n\t\t.slider-info-4963.bafg-slider-info .bafg_slider_readmore_button:hover {\n\n\t\t\t\n\t\t\t\n\t\t\t\t\t}\n\t<\/style>\n\t\n\t\t\t<div class=\"bafg-twentytwenty-container slider-4963  \"\n\t\t\t\tbafg-orientation=\"horizontal\" bafg-default-offset=\"0.5\"\n\t\t\t\tbafg-before-label=\"original\"\n\t\t\t\tbafg-after-label=\"AI segmentation result overlaid\" bafg-overlay=\"1\"\n\t\t\t\tbafg-move-slider-on-hover=\"\"\n\t\t\t\tbafg-click-to-move=\"\">\n\n\t\t\t\t\t\t\t\t<img class=\"skip-lazy\" data-skip-lazy\t\t\t\t\tsrc=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_03_Segmentation_Original.jpg\" alt=\"\">\n\t\t\t\t<img class=\"skip-lazy\" data-skip-lazy\t\t\t\t\tsrc=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_03_Segmentation_Original_Overlay.jpg\" alt=\"\">\n\n\t\t\t<\/div>\n\n\t\t\t\t<div class=\"bafg-slider-info-wraper\">\n\t\t<div style=\"\" class=\"slider-info-4963 bafg-slider-info\">\n\t\t\t\t\t<\/div>\n\t<\/div>\n\t\n\t\t\t<style type=\"text\/css\">\n\t\t\t\t\t\t\t\t\t\t\t<\/style>\n\t\t\t\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_6_Quantify_immune_landscape_in_tumor-micro-environment_TME_including_video\"><\/span>Step 6: Quantify immune landscape in tumor-micro-environment (TME, including video)<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Now that the tumor has been masked, it is also possible to quantify the immune landscape. The <strong>Universal IHC Cell AI App<\/strong> is used to detect cells in each IHC layer, measure the H &amp; DAB stain expression per cell, and classify them as negative or positive (or grade them based on the DAB expression). In a standard single layer IHC scan, the app categorizes cells as &#8220;positive&#8221; or &#8220;negative&#8221;. In a multi-layer stack, however, the app automatically uses the stain names for classification.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1920\" height=\"1032\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection.jpg\" alt=\"screenshot of Universal IHC Cell AI\" class=\"wp-image-4979\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection.jpg 1920w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_04_CD3_CellDetection-740x398.jpg 740w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/><figcaption class=\"wp-element-caption\">Universal IHC Cell AI app detects cells per IHC layer in tumor (red) and tumor margin (green) <\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">The &#8220;Add concentric margins&#8221; option (toolbar | Actions) is used to mask the tumor margin by dilating the tumor masks. In the screenshot below, the tumor is highlighted in red, while the margin appears in green. We have renamed the &#8220;outer margin&#8221; to &#8220;Stroma&#8221;. <\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The &#8220;Add concentric margins&#8221; option creates an &#8220;outermost contour&#8221; class that includes the outlines containing both the margin and the tumor. We select this contour and click the &#8220;Analyze ROI &#8221; button to limit the cell quantification to only this region of interest. We configure the app to only show positive cells, and select the &#8220;Tumor&#8221; and &#8220;Stroma&#8221; classes in the &#8220;<a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/divide-by-roi\/\">Divide by ROI<\/a>&#8221; section to group detected cells accordingly. After analyzing all IHC layers, the following additional annotation classes are generated: &#8220;PCK+ (tumor)&#8221;, PCK+ (stroma), &#8220;CD3+ (tumor)&#8221;, &#8220;CD3+ (stroma), and more.<\/p>\n\n\n\n<figure class=\"wp-block-video\"><video height=\"1080\" style=\"aspect-ratio: 1920 \/ 1080;\" width=\"1920\" controls poster=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_Zelldetection_Alle_Overlay_Tumor_HE.jpg\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/MIKAIA-IHC-Cell-AI-In-Aligned-Serial-Sections.mp4\"><\/video><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\"><span class=\"ez-toc-section\" id=\"Step_7_Cellular_neighborhood_analysis\"><\/span>Step 7: Cellular neighborhood analysis<span class=\"ez-toc-section-end\"><\/span><\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">The <a href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/mikaia-cellular-neighborhood-app\/\"><strong>Cellular Neighborhood App<\/strong><\/a> quantifies the cell composition of the cells&#8217; neighborhood for each cell type. Specifically, it iterates over all cells within the selected cell classes. For each cell, it counts the cell types in the cell&#8217;s vicinity (k nearest neighbor and\/or within a specified radius). These neighborhood composition vectors are then averaged per cell type and categorized by distance.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1920\" height=\"1032\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results.jpg\" alt=\"screenshot of MIKAIA\u00ae Cellular Neighborhood Ap\" class=\"wp-image-4981\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results.jpg 1920w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-300x161.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-1024x550.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-768x413.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-1536x826.jpg 1536w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-370x199.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-270x145.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-570x306.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_06_Cellular_Neighborhood_Results-740x398.jpg 740w\" sizes=\"(max-width: 1920px) 100vw, 1920px\" \/><figcaption class=\"wp-element-caption\">The MIKAIA<sup>\u00ae<\/sup> Cellular Neighborhood App analyzed the local cell compositions and identified 5 cellular neighborhood types CN-1 &#8211; CN-5. The Cluster Explorer diagram informs what cell types the CNs are composed of.<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">As a result, for any given cell type, a diagram shows the average cell type composition within 0-10 \u00b5m radius, within 10-20 \u00b5m, etc. These distance histograms can be viewed side by side per analyzed cell type (screenshot below). Additionally, all data including per-cell-type statistics and even per-cell statistics can be exported to a CSV spreadsheet for a more elaborate downstream analysis outside of MIKAIA<sup>\u00ae<\/sup>.<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1275\" height=\"986\" src=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram.jpg\" alt=\"screenshot of results in MIKAIA\u00ae Cellular Neighborhood App\" class=\"wp-image-4982\" srcset=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram.jpg 1275w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-300x232.jpg 300w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-1024x792.jpg 1024w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-768x594.jpg 768w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-370x286.jpg 370w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-270x209.jpg 270w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-570x441.jpg 570w, https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/Exp3_07_Cellular_Composition_Diagram-740x572.jpg 740w\" sizes=\"(max-width: 1275px) 100vw, 1275px\" \/><figcaption class=\"wp-element-caption\">Result of MIKAIA<sup>\u00ae<\/sup> Cellular Neighborhood App shows per cell type, how on average the cellular neighborhood is composed, broken down by increasing distance in 15 \u00b5m bands<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\" style=\"font-size:25px\"><strong>Contributors &amp; partners<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Institute of Pathology, Erlangen University Hospital, Comprehensive Cancer Center Erlangen\u2013European Metropolitan Area of Nuremberg (CCC ER-EMN), Bavarian Cancer Research Center (BZKF), Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg, Krankenhausstrasse 8\u201310, 91054 Erlangen, Germany<\/li>\n\n\n\n<li>Department of Gynecology and Obstetrics, Erlangen University Hospital, Comprehensive Cancer Center Erlangen\u2013European Metropolitan Area of Nuremberg (CCC ER-EMN), Bavarian Cancer Research Center (BZKF), Erlangen, Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg, Universitaetsstrasse 21\u201323, 91054 Erlangen, Germany<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\" style=\"font-size:25px\"><strong>Funding<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This MIKAIA<sup>\u00ae<\/sup> extension has been kindly made possible thanks to venture capital provided by the <strong><a href=\"https:\/\/www.fraunhofer-zukunftsstiftung.de\/\">Fraunhofer Future Foundation <\/a><\/strong>(Fraunhofer Zukunfsstiftung). Project: &#8220;Histology AI Author&#8221;, consortium: Fraunhofer IIS &amp; Fraunhofer MEVIS.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>This MIKAIA\u00ae University app note illustrates how to interactively train a new H&amp;E segmentation AI with serial sections, without having to annotate tumor regions. Instead, an IHC epithelial marker (pan-cytokeratin AE1\/AE3) is thresholded through H-DAB stain unmixing to generate highly accurate tumor masks. These masks serve as training annotations in the Segmentation AI Author app [&hellip;]<\/p>\n","protected":false},"author":4,"featured_media":4955,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[3,35,28],"tags":[37,87,7,29,66,111],"coauthors":[57],"class_list":["post-4950","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-digital-pathology","category-life-science","category-mikaia-university","tag-ai","tag-ihc","tag-mikaia","tag-mikaia-app-note","tag-multimodal-data","tag-workflow"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.6 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Segmentation AI Author: Serial Sections<\/title>\n<meta name=\"description\" content=\"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Segmentation AI Author: Serial Sections\" \/>\n<meta property=\"og:description\" content=\"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/\" \/>\n<meta property=\"og:site_name\" content=\"SMART SENSING insights\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/FraunhoferIIS\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-15T20:15:36+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-05-06T21:26:38+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1920\" \/>\n\t<meta property=\"og:image:height\" content=\"1032\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"Michaela Benz\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Michaela Benz\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"7 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/\"},\"author\":{\"name\":\"Michaela Benz\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#\\\/schema\\\/person\\\/1adf6207108ff297b4c299a7c17f9cd2\"},\"headline\":\"Segmentation AI Author Example: Train H&amp;E Vulvar Cancer AI Based on Aligned IHC Serial Sections\",\"datePublished\":\"2026-01-15T20:15:36+00:00\",\"dateModified\":\"2026-05-06T21:26:38+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/\"},\"wordCount\":1218,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/05_View_Mask_HE.jpg\",\"keywords\":[\"AI\",\"IHC\",\"MIKAIA\u00ae\",\"MIKAIA\u00ae App Note\",\"Multimodal Data\",\"Workflow\"],\"articleSection\":[\"Digital Pathology\",\"Life Science\",\"MIKAIA University\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/\",\"name\":\"Segmentation AI Author: Serial Sections\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/05_View_Mask_HE.jpg\",\"datePublished\":\"2026-01-15T20:15:36+00:00\",\"dateModified\":\"2026-05-06T21:26:38+00:00\",\"description\":\"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#primaryimage\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/05_View_Mask_HE.jpg\",\"contentUrl\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/05_View_Mask_HE.jpg\",\"width\":1920,\"height\":1032},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/segmentation-ai-author\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Segmentation AI Author Example: Train H&amp;E Vulvar Cancer AI Based on Aligned IHC Serial Sections\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#website\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/\",\"name\":\"SMART SENSING insights\",\"description\":\"learn more about our focus research areas sensor technology, electronics, and artificial intelligence\",\"publisher\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#organization\",\"name\":\"Fraunhofer IIS\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Fraunhofer-IIS-1.png\",\"contentUrl\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Fraunhofer-IIS-1.png\",\"width\":826,\"height\":299,\"caption\":\"Fraunhofer IIS\"},\"image\":{\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/www.facebook.com\\\/FraunhoferIIS\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/fraunhofer-iis\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/#\\\/schema\\\/person\\\/1adf6207108ff297b4c299a7c17f9cd2\",\"name\":\"Michaela Benz\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2023\\\/10\\\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpgd34362660dd3bb203b7cf9c216210eb1\",\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2023\\\/10\\\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpg\",\"contentUrl\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/wp-content\\\/uploads\\\/2023\\\/10\\\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpg\",\"caption\":\"Michaela Benz\"},\"description\":\"Michaela is the Chief Scientist of Fraunhofer IIS\u2019 Medical Image Analysis group. Her research focuses on computational pathology and biomedical image analysis. She holds a doctorate in the field of physics from FAU Erlangen-N\u00fcrnberg and has been with Fraunhofer since 2007.\",\"sameAs\":[\"https:\\\/\\\/orcid.org\\\/0000-0003-0625-4418\",\"https:\\\/\\\/www.linkedin.com\\\/in\\\/michaela-benz-096aab72\\\/\"],\"url\":\"https:\\\/\\\/websites.fraunhofer.de\\\/smart-sensing-insights\\\/author\\\/michaela-benziis-fraunhofer-de\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Segmentation AI Author: Serial Sections","description":"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/","og_locale":"en_US","og_type":"article","og_title":"Segmentation AI Author: Serial Sections","og_description":"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.","og_url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/","og_site_name":"SMART SENSING insights","article_publisher":"https:\/\/www.facebook.com\/FraunhoferIIS","article_published_time":"2026-01-15T20:15:36+00:00","article_modified_time":"2026-05-06T21:26:38+00:00","og_image":[{"width":1920,"height":1032,"url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg","type":"image\/jpeg"}],"author":"Michaela Benz","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Michaela Benz","Est. reading time":"7 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#article","isPartOf":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/"},"author":{"name":"Michaela Benz","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#\/schema\/person\/1adf6207108ff297b4c299a7c17f9cd2"},"headline":"Segmentation AI Author Example: Train H&amp;E Vulvar Cancer AI Based on Aligned IHC Serial Sections","datePublished":"2026-01-15T20:15:36+00:00","dateModified":"2026-05-06T21:26:38+00:00","mainEntityOfPage":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/"},"wordCount":1218,"commentCount":0,"publisher":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#organization"},"image":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#primaryimage"},"thumbnailUrl":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg","keywords":["AI","IHC","MIKAIA\u00ae","MIKAIA\u00ae App Note","Multimodal Data","Workflow"],"articleSection":["Digital Pathology","Life Science","MIKAIA University"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/","name":"Segmentation AI Author: Serial Sections","isPartOf":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#website"},"primaryImageOfPage":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#primaryimage"},"image":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#primaryimage"},"thumbnailUrl":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg","datePublished":"2026-01-15T20:15:36+00:00","dateModified":"2026-05-06T21:26:38+00:00","description":"Use MIKAIA Slide Align to align serial sections. Then mask DAB+ regions in pancytokeratin and use for training with MIKAIA Segmentation AI Author.","breadcrumb":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#primaryimage","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg","contentUrl":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2026\/01\/05_View_Mask_HE.jpg","width":1920,"height":1032},{"@type":"BreadcrumbList","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/segmentation-ai-author\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/"},{"@type":"ListItem","position":2,"name":"Segmentation AI Author Example: Train H&amp;E Vulvar Cancer AI Based on Aligned IHC Serial Sections"}]},{"@type":"WebSite","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#website","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/","name":"SMART SENSING insights","description":"learn more about our focus research areas sensor technology, electronics, and artificial intelligence","publisher":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#organization","name":"Fraunhofer IIS","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#\/schema\/logo\/image\/","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2023\/06\/Fraunhofer-IIS-1.png","contentUrl":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2023\/06\/Fraunhofer-IIS-1.png","width":826,"height":299,"caption":"Fraunhofer IIS"},"image":{"@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/FraunhoferIIS","https:\/\/www.linkedin.com\/company\/fraunhofer-iis"]},{"@type":"Person","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/#\/schema\/person\/1adf6207108ff297b4c299a7c17f9cd2","name":"Michaela Benz","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2023\/10\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpgd34362660dd3bb203b7cf9c216210eb1","url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2023\/10\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpg","contentUrl":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-content\/uploads\/2023\/10\/cropped-Benz_Michaela_DSC_0162_Pulkert-scaled-1-96x96.jpg","caption":"Michaela Benz"},"description":"Michaela is the Chief Scientist of Fraunhofer IIS\u2019 Medical Image Analysis group. Her research focuses on computational pathology and biomedical image analysis. She holds a doctorate in the field of physics from FAU Erlangen-N\u00fcrnberg and has been with Fraunhofer since 2007.","sameAs":["https:\/\/orcid.org\/0000-0003-0625-4418","https:\/\/www.linkedin.com\/in\/michaela-benz-096aab72\/"],"url":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/author\/michaela-benziis-fraunhofer-de\/"}]}},"_links":{"self":[{"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/posts\/4950","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/users\/4"}],"replies":[{"embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/comments?post=4950"}],"version-history":[{"count":23,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/posts\/4950\/revisions"}],"predecessor-version":[{"id":5355,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/posts\/4950\/revisions\/5355"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/media\/4955"}],"wp:attachment":[{"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/media?parent=4950"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/categories?post=4950"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/tags?post=4950"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/websites.fraunhofer.de\/smart-sensing-insights\/wp-json\/wp\/v2\/coauthors?post=4950"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}