@inproceedings{RosbachAmmelingKruegeletal.2025, author = {Rosbach, Emely and Ammeling, Jonas and Kr{\"u}gel, Sebastian and Kießig, Angelika and Fritz, Alexis and Ganz, Jonathan and Puget, Chlo{\´e} and Donovan, Taryn and Klang, Andrea and K{\"o}ller, Maximilian C. and Bolfa, Pompei and Tecilla, Marco and Denk, Daniela and Kiupel, Matti and Paraschou, Georgios and Kok, Mun Keong and Haake, Alexander F. H. and de Krijger, Ronald R. and Sonnen, Andreas F.-P. and Kasantikul, Tanit and Dorrestein, Gerry M. and Smedley, Rebecca C. and Stathonikos, Nikolas and Uhl, Matthias and Bertram, Christof and Riener, Andreas and Aubreville, Marc}, title = {"When Two Wrongs Don't Make a Right" - Examining Confirmation Bias and the Role of Time Pressure During Human-AI Collaboration in Computational Pathology}, pages = {528}, booktitle = {CHI'25: Proceedings of the 2025 CHI Conference on Human Factors in Computing Systems}, editor = {Yamashita, Naomi and Evers, Vanessa and Yatani, Koji and Ding, Xianghua and Lee, Bongshin and Chetty, Marshini and Toups-Dugas, Phoebe}, publisher = {ACM}, address = {New York}, isbn = {979-8-4007-1394-1}, doi = {https://doi.org/10.1145/3706598.3713319}, year = {2025}, abstract = {Artificial intelligence (AI)-based decision support systems hold promise for enhancing diagnostic accuracy and efficiency in computational pathology. However, human-AI collaboration can introduce and amplify cognitive biases, like confirmation bias caused by false confirmation when erroneous human opinions are reinforced by inaccurate AI output. This bias may increase under time pressure, a ubiquitous factor in routine pathology, as it strains practitioners' cognitive resources. We quantified confirmation bias triggered by AI-induced false confirmation and examined the role of time constraints in a web-based experiment, where trained pathology experts (n=28) estimated tumor cell percentages. Our results suggest that AI integration fuels confirmation bias, evidenced by a statistically significant positive linear-mixed-effects model coefficient linking AI recommendations mirroring flawed human judgment and alignment with system advice. Conversely, time pressure appeared to weaken this relationship. These findings highlight potential risks of AI in healthcare and aim to support the safe integration of clinical decision support systems.}, language = {en} } @unpublished{RosbachAmmelingKruegeletal.2024, author = {Rosbach, Emely and Ammeling, Jonas and Kr{\"u}gel, Sebastian and Kießig, Angelika and Fritz, Alexis and Ganz, Jonathan and Puget, Chlo{\´e} and Donovan, Taryn and Klang, Andrea and K{\"o}ller, Maximilian C. and Bolfa, Pompei and Tecilla, Marco and Denk, Daniela and Kiupel, Matti and Paraschou, Georgios and Kok, Mun Keong and Haake, Alexander F. H. and de Krijger, Ronald R. and Sonnen, Andreas F.-P. and Kasantikul, Tanit and Dorrestein, Gerry M. and Smedley, Rebecca C. and Stathonikos, Nikolas and Uhl, Matthias and Bertram, Christof and Riener, Andreas and Aubreville, Marc}, title = {"When TwoWrongs Don't Make a Right" - Examining Confirmation Bias and the Role of Time Pressure During Human-AI Collaboration in Computational Pathology}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2411.01007}, year = {2024}, language = {en} } @unpublished{AubrevilleStathonikosDonovanetal.2023, author = {Aubreville, Marc and Stathonikos, Nikolas and Donovan, Taryn and Klopfleisch, Robert and Ganz, Jonathan and Ammeling, Jonas and Wilm, Frauke and Veta, Mitko and Jabari, Samir and Eckstein, Markus and Annuscheit, Jonas and Krumnow, Christian and Bozaba, Engin and Cayir, Sercan and Gu, Hongyan and Chen, Xiang and Jahanifar, Mostafa and Shephard, Adam and Kondo, Satoshi and Kasai, Satoshi and Kotte, Sujatha and Saipradeep, Vangala and Lafarge, Maxime W. and Koelzer, Viktor H. and Wang, Ziyue and Zhang, Yongbing and Yang, Sen and Wang, Xiyue and Breininger, Katharina and Bertram, Christof}, title = {Domain generalization across tumor types, laboratories, and species - Insights from the 2022 edition of the Mitosis Domain Generalization Challenge}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2309.15589}, year = {2023}, abstract = {Recognition of mitotic figures in histologic tumor specimens is highly relevant to patient outcome assessment. This task is challenging for algorithms and human experts alike, with deterioration of algorithmic performance under shifts in image representations. Considerable covariate shifts occur when assessment is performed on different tumor types, images are acquired using different digitization devices, or specimens are produced in different laboratories. This observation motivated the inception of the 2022 challenge on MItosis Domain Generalization (MIDOG 2022). The challenge provided annotated histologic tumor images from six different domains and evaluated the algorithmic approaches for mitotic figure detection provided by nine challenge participants on ten independent domains. Ground truth for mitotic figure detection was established in two ways: a three-expert consensus and an independent, immunohistochemistry-assisted set of labels. This work represents an overview of the challenge tasks, the algorithmic strategies employed by the participants, and potential factors contributing to their success. With an F1 score of 0.764 for the top-performing team, we summarize that domain generalization across various tumor domains is possible with today's deep learning-based recognition pipelines. When assessed against the immunohistochemistry-assisted reference standard, all methods resulted in reduced recall scores, but with only minor changes in the order of participants in the ranking.}, language = {en} } @unpublished{HaghoferParlakBarteletal.2023, author = {Haghofer, Andreas and Parlak, Eda and Bartel, Alexander and Donovan, Taryn and Assenmacher, Charles-Antoine and Bolfa, Pompei and Dark, Michael and Fuchs-Baumgartinger, Andrea and Klang, Andrea and J{\"a}ger, Kathrin and Klopfleisch, Robert and Merz, Sophie and Richter, Barbara and Schulman, F. Yvonne and Ganz, Jonathan and Scharinger, Josef and Aubreville, Marc and Winkler, Stephan M. and Kiupel, Matti and Bertram, Christof}, title = {Nuclear Morphometry using a Deep Learning-based Algorithm has Prognostic Relevance for Canine Cutaneous Mast Cell Tumors}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2309.15031}, year = {2023}, abstract = {Variation in nuclear size and shape is an important criterion of malignancy for many tumor types; however, categorical estimates by pathologists have poor reproducibility. Measurements of nuclear characteristics (morphometry) can improve reproducibility, but manual methods are time consuming. In this study, we evaluated fully automated morphometry using a deep learning-based algorithm in 96 canine cutaneous mast cell tumors with information on patient survival. Algorithmic morphometry was compared with karyomegaly estimates by 11 pathologists, manual nuclear morphometry of 12 cells by 9 pathologists, and the mitotic count as a benchmark. The prognostic value of automated morphometry was high with an area under the ROC curve regarding the tumor-specific survival of 0.943 (95\% CI: 0.889 - 0.996) for the standard deviation (SD) of nuclear area, which was higher than manual morphometry of all pathologists combined (0.868, 95\% CI: 0.737 - 0.991) and the mitotic count (0.885, 95\% CI: 0.765 - 1.00). At the proposed thresholds, the hazard ratio for algorithmic morphometry (SD of nuclear area ≥9.0μm2) was 18.3 (95\% CI: 5.0 - 67.1), for manual morphometry (SD of nuclear area ≥10.9μm2) 9.0 (95\% CI: 6.0 - 13.4), for karyomegaly estimates 7.6 (95\% CI: 5.7 - 10.1), and for the mitotic count 30.5 (95\% CI: 7.8 - 118.0). Inter-rater reproducibility for karyomegaly estimates was fair (κ = 0.226) with highly variable sensitivity/specificity values for the individual pathologists. Reproducibility for manual morphometry (SD of nuclear area) was good (ICC = 0.654). This study supports the use of algorithmic morphometry as a prognostic test to overcome the limitations of estimates and manual measurements.}, language = {en} } @unpublished{GanzMarzahlAmmelingetal.2024, author = {Ganz, Jonathan and Marzahl, Christian and Ammeling, Jonas and Richter, Barbara and Puget, Chlo{\´e} and Denk, Daniela and Demeter, Elena A. and Tabaran, Flaviu A. and Wasinger, Gabriel and Lipnik, Karoline and Tecilla, Marco and Valentine, Matthew J. and Dark, Michael and Abele, Niklas and Bolfa, Pompei and Erber, Ramona and Klopfleisch, Robert and Merz, Sophie and Donovan, Taryn and Jabari, Samir and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {On the Value of PHH3 for Mitotic Figure Detection on H\&E-stained Images}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2406.19899}, year = {2024}, abstract = {The count of mitotic figures (MFs) observed in hematoxylin and eosin (H\&E)-stained slides is an important prognostic marker as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. Deep learning algorithms can standardize this task, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithm's performance. Unlike H\&E, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H\&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H\&E stain alone, the use of this ground truth could potentially introduce noise into the H\&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. We found that the annotators' object-level agreement increased when using PHH3-assisted labeling. Subsequently, MF detectors were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. Additionally, a novel dual-stain MF detector was developed to investigate the interpretation-shift of PHH3-assisted labels used in H\&E, which clearly outperformed single-stain detectors. However, the PHH3-assisted labels did not have a positive effect on solely H\&E-based models. The high performance of our dual-input detector reveals an information mismatch between the H\&E and PHH3-stained images as the cause of this effect.}, language = {en} } @unpublished{GanzAmmelingRosbachetal.2024, author = {Ganz, Jonathan and Ammeling, Jonas and Rosbach, Emely and Lausser, Ludwig and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Is Self-Supervision Enough? Benchmarking Foundation Models Against End-to-End Training for Mitotic Figure Classification}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2412.06365}, year = {2024}, abstract = {Foundation models (FMs), i.e., models trained on a vast amount of typically unlabeled data, have become popular and available recently for the domain of histopathology. The key idea is to extract semantically rich vectors from any input patch, allowing for the use of simple subsequent classification networks potentially reducing the required amounts of labeled data, and increasing domain robustness. In this work, we investigate to which degree this also holds for mitotic figure classification. Utilizing two popular public mitotic figure datasets, we compared linear probing of five publicly available FMs against models trained on ImageNet and a simple ResNet50 end-to-end-trained baseline. We found that the end-to-end-trained baseline outperformed all FM-based classifiers, regardless of the amount of data provided. Additionally, we did not observe the FM-based classifiers to be more robust against domain shifts, rendering both of the above assumptions incorrect.}, language = {en} } @article{BertramAubrevilleGurtneretal.2020, author = {Bertram, Christof and Aubreville, Marc and Gurtner, Corinne and Bartel, Alexander and Corner, Sarah M. and Dettwiler, Martina and Kershaw, Olivia and Noland, Erica L. and Schmidt, Anja and Sledge, Dodd G. and Smedley, Rebecca C. and Thaiwong, Tuddow and Kiupel, Matti and Maier, Andreas and Klopfleisch, Robert}, title = {Computerized Calculation of Mitotic Count Distribution in Canine Cutaneous Mast Cell Tumor Sections: Mitotic Count Is Area Dependent}, volume = {57}, journal = {Veterinary Pathology}, number = {2}, publisher = {Sage}, address = {London}, issn = {1544-2217}, doi = {https://doi.org/10.1177/0300985819890686}, pages = {214 -- 226}, year = {2020}, language = {en} } @inproceedings{MarzahlAubrevilleBertrametal.2020, author = {Marzahl, Christian and Aubreville, Marc and Bertram, Christof and Gerlach, Stefan and Maier, Jennifer and Voigt, J{\"o}rn and Hill, Jenny and Klopfleisch, Robert and Maier, Andreas}, title = {Is crowd-algorithm collaboration an advanced alternative to crowd-sourcing on cytology slides?}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2020, Algorithmen - Systeme - Anwendungen. Proceedings des Workshops vom 15. bis 17. M{\"a}rz 2020 in Berlin}, editor = {Tolxdorff, Thomas and Deserno, Thomas Martin and Handels, Heinz and Maier, Andreas and Maier-Hein, Klaus H. and Palm, Christoph}, publisher = {Springer Vieweg}, address = {Wiesbaden}, isbn = {978-3-658-29266-9}, doi = {https://doi.org/10.1007/978-3-658-29267-6_5}, pages = {26 -- 31}, year = {2020}, language = {en} } @inproceedings{AubrevilleBertramJabarietal.2020, author = {Aubreville, Marc and Bertram, Christof and Jabari, Samir and Marzahl, Christian and Klopfleisch, Robert and Maier, Andreas}, title = {Inter-species, inter-tissue domain adaptation for mitotic figure assessment}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2020, Algorithmen - Systeme - Anwendungen. Proceedings des Workshops vom 15. bis 17. M{\"a}rz 2020 in Berlin}, subtitle = {learning new tricks from old dogs}, editor = {Tolxdorff, Thomas and Deserno, Thomas Martin and Handels, Heinz and Maier, Andreas and Maier-Hein, Klaus H. and Palm, Christoph}, publisher = {Springer Vieweg}, address = {Wiesbaden}, isbn = {978-3-658-29266-9}, doi = {https://doi.org/10.1007/978-3-658-29267-6_1}, pages = {1 -- 7}, year = {2020}, language = {en} } @inproceedings{GanzBertramKlopfleischetal.2022, author = {Ganz, Jonathan and Bertram, Christof and Klopfleisch, Robert and Jabari, Samir and Breininger, Katharina and Aubreville, Marc}, title = {Classification of visibility in multi-stain microscopy images}, booktitle = {Medical Imaging with Deep Learning: MIDL 2022 Short Papers}, url = {https://openreview.net/forum?id=-GsA-mUVmm}, year = {2022}, language = {en} } @unpublished{PugetGanzOstermaieretal.2024, author = {Puget, Chlo{\´e} and Ganz, Jonathan and Ostermaier, Julian and Konrad, Thomas and Parlak, Eda and Bertram, Christof and Kiupel, Matti and Breininger, Katharina and Aubreville, Marc and Klopfleisch, Robert}, title = {Deep Learning model predicts the c-Kit-11 mutational status of canine cutaneous mast cell tumors by HE stained histological slides}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2401.06169}, year = {2024}, abstract = {Numerous prognostic factors are currently assessed histopathologically in biopsies of canine mast cell tumors to evaluate clinical behavior. In addition, PCR analysis of the c-Kit exon 11 mutational status is often performed to evaluate the potential success of a tyrosine kinase inhibitor therapy. This project aimed at training deep learning models (DLMs) to identify the c-Kit-11 mutational status of MCTs solely based on morphology without additional molecular analysis. HE slides of 195 mutated and 173 non-mutated tumors were stained consecutively in two different laboratories and scanned with three different slide scanners. This resulted in six different datasets (stain-scanner variations) of whole slide images. DLMs were trained with single and mixed datasets and their performances was assessed under scanner and staining domain shifts. The DLMs correctly classified HE slides according to their c-Kit 11 mutation status in, on average, 87\% of cases for the best-suited stain-scanner variant. A relevant performance drop could be observed when the stain-scanner combination of the training and test dataset differed. Multi-variant datasets improved the average accuracy but did not reach the maximum accuracy of algorithms trained and tested on the same stain-scanner variant. In summary, DLM-assisted morphological examination of MCTs can predict c-Kit-exon 11 mutational status of MCTs with high accuracy. However, the recognition performance is impeded by a change of scanner or staining protocol. Larger data sets with higher numbers of scans originating from different laboratories and scanners may lead to more robust DLMs to identify c-Kit mutations in HE slides.}, language = {en} } @inproceedings{AubrevilleBertramKlopfleischetal.2019, author = {Aubreville, Marc and Bertram, Christof and Klopfleisch, Robert and Maier, Andreas}, title = {Field of Interest Proposal for Augmented Mitotic Cell Count}, volume = {2}, booktitle = {Proceedings of the 12th International Joint Conference on Biomedical Engineering Systems and Technologies - BIOIMAGING}, subtitle = {Comparison of Two Convolutional Networks}, publisher = {SciTePress}, address = {Set{\´u}bal}, isbn = {978-989-758-353-7}, issn = {2184-4305}, doi = {https://doi.org/10.5220/0007365700300037}, pages = {30 -- 37}, year = {2019}, language = {en} } @article{BertramAubrevilleDonovanetal.2021, author = {Bertram, Christof and Aubreville, Marc and Donovan, Taryn and Bartel, Alexander and Wilm, Frauke and Marzahl, Christian and Assenmacher, Charles-Antoine and Becker, Kathrin and Bennett, Mark and Corner, Sarah M. and Cossic, Brieuc and Denk, Daniela and Dettwiler, Martina and Garcia Gonzalez, Beatriz and Gurtner, Corinne and Haverkamp, Ann-Kathrin and Heier, Annabelle and Lehmbecker, Annika and Merz, Sophie and Noland, Erica L. and Plog, Stephanie and Schmidt, Anja and Sebastian, Franziska and Sledge, Dodd G. and Smedley, Rebecca C. and Tecilla, Marco and Thaiwong, Tuddow and Fuchs-Baumgartinger, Andrea and Meuten, Donald J. and Breininger, Katharina and Kiupel, Matti and Maier, Andreas and Klopfleisch, Robert}, title = {Computer-assisted mitotic count using a deep learning-based algorithm improves interobserver reproducibility and accuracy}, volume = {59}, journal = {Veterinary Pathology}, number = {2}, publisher = {Sage}, address = {London}, issn = {1544-2217}, doi = {https://doi.org/10.1177/03009858211067478}, pages = {211 -- 226}, year = {2021}, abstract = {The mitotic count (MC) is an important histological parameter for prognostication of malignant neoplasms. However, it has inter- and intraobserver discrepancies due to difficulties in selecting the region of interest (MC-ROI) and in identifying or classifying mitotic figures (MFs). Recent progress in the field of artificial intelligence has allowed the development of high-performance algorithms that may improve standardization of the MC. As algorithmic predictions are not flawless, computer-assisted review by pathologists may ensure reliability. In the present study, we compared partial (MC-ROI preselection) and full (additional visualization of MF candidates and display of algorithmic confidence values) computer-assisted MC analysis to the routine (unaided) MC analysis by 23 pathologists for whole-slide images of 50 canine cutaneous mast cell tumors (ccMCTs). Algorithmic predictions aimed to assist pathologists in detecting mitotic hotspot locations, reducing omission of MFs, and improving classification against imposters. The interobserver consistency for the MC significantly increased with computer assistance (interobserver correlation coefficient, ICC = 0.92) compared to the unaided approach (ICC = 0.70). Classification into prognostic stratifications had a higher accuracy with computer assistance. The algorithmically preselected hotspot MC-ROIs had a consistently higher MCs than the manually selected MC-ROIs. Compared to a ground truth (developed with immunohistochemistry for phosphohistone H3), pathologist performance in detecting individual MF was augmented when using computer assistance (F1-score of 0.68 increased to 0.79) with a reduction in false negatives by 38\%. The results of this study demonstrate that computer assistance may lead to more reproducible and accurate MCs in ccMCTs.}, language = {en} } @article{MarzahlAubrevilleBertrametal.2020, author = {Marzahl, Christian and Aubreville, Marc and Bertram, Christof and Stayt, Jason and Jasensky, Anne-Katherine and Bartenschlager, Florian and Fragoso-Garcia, Marco and Barton, Ann K. and Elsemann, Svenja and Jabari, Samir and Krauth, Jens and Madhu, Prathmesh and Voigt, J{\"o}rn and Hill, Jenny and Klopfleisch, Robert and Maier, Andreas}, title = {Deep Learning-based quantification of pulmonary hemosiderophages in cytology slides}, volume = {10}, pages = {9795}, journal = {Scientific Reports}, publisher = {Springer Nature}, address = {London}, issn = {2045-2322}, doi = {https://doi.org/10.1038/s41598-020-65958-2}, year = {2020}, abstract = {Exercise-induced pulmonary hemorrhage (EIPH) is a common condition in sport horses with negative impact on performance. Cytology of bronchoalveolar lavage fluid by use of a scoring system is considered the most sensitive diagnostic method. Macrophages are classified depending on the degree of cytoplasmic hemosiderin content. The current gold standard is manual grading, which is however monotonous and time-consuming. We evaluated state-of-the-art deep learning-based methods for single cell macrophage classification and compared them against the performance of nine cytology experts and evaluated inter- and intra-observer variability. Additionally, we evaluated object detection methods on a novel data set of 17 completely annotated cytology whole slide images (WSI) containing 78,047 hemosiderophages. Our deep learning-based approach reached a concordance of 0.85, partially exceeding human expert concordance (0.68 to 0.86, mean of 0.73, SD of 0.04). Intra-observer variability was high (0.68 to 0.88) and inter-observer concordance was moderate (Fleiss' kappa = 0.67). Our object detection approach has a mean average precision of 0.66 over the five classes from the whole slide gigapixel image and a computation time of below two minutes. To mitigate the high inter- and intra-rater variability, we propose our automated object detection pipeline, enabling accurate, reproducible and quick EIPH scoring in WSI.}, language = {en} } @article{DonovanMooreBertrametal.2021, author = {Donovan, Taryn and Moore, Frances M. and Bertram, Christof and Luong, Richard and Bolfa, Pompei and Klopfleisch, Robert and Tvedten, Harold and Salas, Elisa N. and Whitley, Derick and Aubreville, Marc and Meuten, Donald J.}, title = {Mitotic Figures - Normal, Atypical, and Imposters: A Guide to Identification}, volume = {58}, journal = {Veterinary pathology}, number = {2}, publisher = {Sage}, address = {London}, issn = {1544-2217}, doi = {https://doi.org/10.1177/0300985820980049}, pages = {243 -- 257}, year = {2021}, language = {en} } @inproceedings{AubrevilleBertramKlopfleischetal.2019, author = {Aubreville, Marc and Bertram, Christof and Klopfleisch, Robert and Maier, Andreas}, title = {Augmented mitotic cell count using field of interest proposal}, booktitle = {Bildverarbeitung f{\"u}r die Medizin 2019, Algorithmen - Systeme - Anwendungen, Proceedings des Workshops vom 17. bis 19. M{\"a}rz 2019 in L{\"u}beck}, editor = {Handels, Heinz and Deserno, Thomas Martin and Maier, Andreas and Maier-Hein, Klaus H. and Palm, Christoph and Tolxdorff, Thomas}, publisher = {Springer Vieweg}, address = {Wiesbaden}, isbn = {978-3-658-25325-7}, doi = {https://doi.org/10.1007/978-3-658-25326-4_71}, pages = {321 -- 326}, year = {2019}, language = {en} } @unpublished{GanzMarzahlAmmelingetal.2024, author = {Ganz, Jonathan and Marzahl, Christian and Ammeling, Jonas and Rosbach, Emely and Richter, Barbara and Puget, Chlo{\´e} and Denk, Daniela and Demeter, Elena A. and Tabaran, Flaviu A. and Wasinger, Gabriel and Lipnik, Karoline and Tecilla, Marco and Valentine, Matthew J. and Dark, Michael and Abele, Niklas and Bolfa, Pompei and Erber, Ramona and Klopfleisch, Robert and Merz, Sophie and Donovan, Taryn and Jabari, Samir and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Information Mismatch in PHH3-Assisted Mitosis Annotation Leads to Interpretation Shifts in H\&E Slide Analysis}, titleParent = {Research Square}, publisher = {Research Square}, address = {Durham}, doi = {https://doi.org/10.21203/rs.3.rs-4900505/v1}, year = {2024}, abstract = {The count of mitotic figures (MFs) observed in hematoxylin and eosin (H\&E)-stained slides is an important prognostic marker, as it is a measure for tumor cell proliferation. However, the identification of MFs has a known low inter-rater agreement. In a computer-aided setting, deep learning algorithms can help to mitigate this, but they require large amounts of annotated data for training and validation. Furthermore, label noise introduced during the annotation process may impede the algorithms' performance. Unlike H\&E, where identification of MFs is based mainly on morphological features, the mitosis-specific antibody phospho-histone H3 (PHH3) specifically highlights MFs. Counting MFs on slides stained against PHH3 leads to higher agreement among raters and has therefore recently been used as a ground truth for the annotation of MFs in H\&E. However, as PHH3 facilitates the recognition of cells indistinguishable from H\&E staining alone, the use of this ground truth could potentially introduce an interpretation shift and even label noise into the H\&E-related dataset, impacting model performance. This study analyzes the impact of PHH3-assisted MF annotation on inter-rater reliability and object level agreement through an extensive multi-rater experiment. Subsequently, MF detectors, including a novel dual-stain detector, were evaluated on the resulting datasets to investigate the influence of PHH3-assisted labeling on the models' performance. We found that the annotators' object-level agreement significantly increased when using PHH3-assisted labeling (F1: 0.53 to 0.74). However, this enhancement in label consistency did not translate to improved performance for H\&E-based detectors, neither during the training phase nor the evaluation phase. Conversely, the dual-stain detector was able to benefit from the higher consistency. This reveals an information mismatch between the H\&E and PHH3-stained images as the cause of this effect, which renders PHH3-assisted annotations not well-aligned for use with H\&E-based detectors. Based on our findings, we propose an improved PHH3-assisted labeling procedure.}, language = {en} } @article{AubrevilleBertramMarzahletal.2020, author = {Aubreville, Marc and Bertram, Christof and Marzahl, Christian and Gurtner, Corinne and Dettwiler, Martina and Schmidt, Anja and Bartenschlager, Florian and Merz, Sophie and Fragoso-Garcia, Marco and Kershaw, Olivia and Klopfleisch, Robert and Maier, Andreas}, title = {Deep learning algorithms out-perform veterinary pathologists in detecting the mitotically most active tumor region}, volume = {10}, pages = {16447}, journal = {Scientific reports}, publisher = {Springer Nature}, address = {London}, issn = {2045-2322}, doi = {https://doi.org/10.1038/s41598-020-73246-2}, year = {2020}, abstract = {Manual count of mitotic figures, which is determined in the tumor region with the highest mitotic activity, is a key parameter of most tumor grading schemes. It can be, however, strongly dependent on the area selection due to uneven mitotic figure distribution in the tumor section. We aimed to assess the question, how significantly the area selection could impact the mitotic count, which has a known high inter-rater disagreement. On a data set of 32 whole slide images of H\&E-stained canine cutaneous mast cell tumor, fully annotated for mitotic figures, we asked eight veterinary pathologists (five board-certified, three in training) to select a field of interest for the mitotic count. To assess the potential difference on the mitotic count, we compared the mitotic count of the selected regions to the overall distribution on the slide. Additionally, we evaluated three deep learning-based methods for the assessment of highest mitotic density: In one approach, the model would directly try to predict the mitotic count for the presented image patches as a regression task. The second method aims at deriving a segmentation mask for mitotic figures, which is then used to obtain a mitotic density. Finally, we evaluated a two-stage object-detection pipeline based on state-of-the-art architectures to identify individual mitotic figures. We found that the predictions by all models were, on average, better than those of the experts. The two-stage object detector performed best and outperformed most of the human pathologists on the majority of tumor cases. The correlation between the predicted and the ground truth mitotic count was also best for this approach (0.963-0.979). Further, we found considerable differences in position selection between pathologists, which could partially explain the high variance that has been reported for the manual mitotic count. To achieve better inter-rater agreement, we propose to use a computer-based area selection for support of the pathologist in the manual mitotic count.}, language = {en} } @unpublished{AmmelingHeckerGanzetal.2023, author = {Ammeling, Jonas and Hecker, Moritz and Ganz, Jonathan and Donovan, Taryn and Klopfleisch, Robert and Bertram, Christof and Breininger, Katharina and Aubreville, Marc}, title = {Automated Volume Corrected Mitotic Index Calculation Through Annotation-Free Deep Learning using Immunohistochemistry as Reference Standard}, publisher = {arXiv}, address = {Ithaca}, doi = {https://doi.org/10.48550/arXiv.2311.08949}, year = {2023}, abstract = {The volume-corrected mitotic index (M/V-Index) was shown to provide prognostic value in invasive breast carcinomas. However, despite its prognostic significance, it is not established as the standard method for assessing aggressive biological behaviour, due to the high additional workload associated with determining the epithelial proportion. In this work, we show that using a deep learning pipeline solely trained with an annotation-free, immunohistochemistry-based approach, provides accurate estimations of epithelial segmentation in canine breast carcinomas. We compare our automatic framework with the manually annotated M/V-Index in a study with three board-certified pathologists. Our results indicate that the deep learning-based pipeline shows expert-level performance, while providing time efficiency and reproducibility.}, language = {en} } @article{BertramAubrevilleGurtneretal.2020, author = {Bertram, Christof and Aubreville, Marc and Gurtner, Corinne and Bartel, Alexander and Corner, Sarah M. and Dettwiler, Martina and Kershaw, Olivia and Noland, Erica L. and Schmidt, Anja and Sledge, Dodd G. and Smedley, Rebecca C. and Thaiwong, Tuddow and Kiupel, Matti and Maier, Andreas and Klopfleisch, Robert}, title = {Mitotic count in canine cutaneous mast cell tumours}, volume = {2020}, journal = {Journal of Comparative Pathology}, subtitle = {not accurate but reproducible}, number = {174}, publisher = {Elsevier}, address = {London}, issn = {1532-3129}, doi = {https://doi.org/10.1016/j.jcpa.2019.10.015}, pages = {143}, year = {2020}, language = {en} }