Markus D. Herrmann, MD, PhD
Markus D. Herrmann, MD, PhD
Markus D. Herrmann, MD, PhDAssistant Professor of Pathology, Harvard Medical School
Director of Computational Pathology, Massachusetts General Hospital
Assistant Computational Pathologist, Massachusetts General Hospital
About the Lab
Histopathology relies on the microscopic examination of stained tissue section specimens. While this 150-year-old methodology remains a cornerstone of tissue-based studies and the gold standard for tissue-based diagnostics, it has several limitations. Most notably, observations are qualitative and subjective and consequently display high inter- and intraobserver variability. Our goal is to move the field from qualitative evaluations and subjective interpretations towards measurements and quantifiable predictions to promote evidencebased clinical decision-making. We drive the transformation of pathology into a quantitative science using a systems approach that combines advanced automated microscopy imaging, statistics and machine learning to study the pathogenesis of complex diseases, discover novel phenotypic biomarkers, and develop image-based clinical tests for rendering histopathologic diagnoses and predicting response to therapy.
In collaboration with other groups within the department as well as the wider hospital and university network, we develop novel image-based methods to study phenotypic manifestations and molecular mechanisms of diseases in human primary tissue from the centimeter to the nanometer length scale using various in vitro and in vivo imaging modalities, including multiplexed immunofluorescence microscopy, electron microscopy, computed tomography and optical coherence tomography. Applying representation learning and statistical inference, we computationally extract combined spatial morphologic and molecular proteomic features from large numbers of digital images and model relationships between image-derived variates and clinical outcomes to test hypotheses about underlying disease mechanisms and make predictions about disease progression. One of our focus areas is the study of biochemical and biophysical interactions between cells and their tissue microenvironment as well as the formation and maintenance of cellular compartments in tissue context. Specifically, we aim to improve our understanding of how these essential biological processes are disrupted during tumorigenesis – especially during invasive growth and metastasis of malignant neoplasms – and build on pathophysiological insights to discover clinically actionable biomarkers and targets for therapeutic intervention.
Our group actively engages in the translation of scientific discoveries and technological advances into improved clinical services. We concentrate on the development of multivariate image-based biomarkers and computational diagnostic, prognostic and predictive tests that can support and guide clinical decisions. As part of these efforts, we develop computational methods, tools and services to evaluate the technical performance of machine learning models, to assess their generalizability to relevant clinical conditions and questions, and to facilitate the interpretation of model predictions in clinical context. In addition, we partner with clinicians and regulatory experts on designing and conducting clinical reader studies for clinical validation of computational tests as well as on devising and implementing strategies to continuously monitor the performance and safety of machine learning-based software during clinical application.
Fedorov A, Beichel R, Kalpathy-Cramer J, Clunie D, Onken M, Riesmeier J, Herz C, Bauer C, Beers A, Fillion-Robin JC, Lasso A, Pinter C, Pieper S, Nolden M, Maier-Hein K, Herrmann MD, Saltz J, Prior F, Fennessy F, Buatti J, Kikinis R. Quantitative Imaging Informatics for Cancer Research. JCO Clinical Cancer Informatics. 2020 May 11;4:444—453
Herrmann MD, Clunie DA, Fedorov A, Doyle SW, Pieper S, Klepeis V, Le LP, Mutter GL, Milstone DS, Schultz TJ, Kikinis R, Kotecha GK, Hwang DH, Andriole KP, Iafrate AJ, Brink JA, Boland GW, Dreyer KJ, Michalski M, Golden JA, Louis DN, and Lennerz JK. Implementing the DICOM standard for digital pathology. J Pathol Inform. 2018 Nov 2;9:37
Gut G, Herrmann MD, Pelkmans L. Multiplexed protein maps link subcellular organization to
cellular states. Science. 2018 Aug 3;361(6401):eaar7042
Caicedo JC, Cooper S, Heigwer F, Warchal S, Qiu P, Molnar C, Vasilevich AS, Barry JD, Bansal HS, Kraus O, Wawer M, Paavolainen L, Herrmann MD, Rohban M, Hung J, Hennig H, Concannon J, Smith I, Clemons PA, Singh S, Rees P, Horvath P, Linington RG, Carpenter AE. Data analysis strategies for image-based cell profiling. Nature Methods. 2017. Aug 31;14(9):849—863 Stoeger T, Battich M, Herrmann MD, Yakimovich Y, Pelkmans L. Computer vision for imagebased transcriptomics. Methods. 2014 Sep 1;85:44–53