Next-Generation Morphometry for pathomics-data mining in histopathology.

Pathology diagnostics relies on the assessment of morphology by trained experts, which remains subjective and qualitative. Here we developed a framework for large-scale histomorphometry (FLASH) performing deep learning-based semantic segmentation and subsequent large-scale extraction of interpretable, quantitative, morphometric features in non-tumour kidney histology. We use two internal and three external, multi-centre cohorts to analyse over 1000 kidney biopsies and nephrectomies. By associating morphometric features with clinical parameters, we confirm previous concepts and reveal unexpected relations. We show that the extracted features are independent predictors of long-term clinical outcomes in IgA-nephropathy. We introduce single-structure morphometric analysis by applying techniques from single-cell transcriptomics, identifying distinct glomerular populations and morphometric phenotypes along a trajectory of disease progression. Our study provides a concept for Next-generation Morphometry (NGM), enabling comprehensive quantitative pathology data mining, i.e., pathomics.

Stain-Independent Deep Learning-Based Analysis of Digital Kidney Histopathology.

Convolutional neural network (CNN)-based image analysis applications in digital pathology (eg, tissue segmentation) require a large amount of annotated data and are mostly trained and applicable on a single stain. Here, a novel concept based on stain augmentation is proposed to develop stain-independent CNNs requiring only one annotated stain. In this benchmark study on stain independence in digital pathology, this approach is comprehensively compared with state-of-the-art techniques including image registration and stain translation, and several modifications thereof. A previously developed CNN for segmentation of periodic acid-Schiff-stained kidney histology was used and applied to various immunohistochemical stainings. Stain augmentation showed very high performance in all evaluated stains and outperformed all other techniques in all structures and stains. Without the need for additional annotations, it enabled segmentation on immunohistochemical stainings with performance nearly comparable to that of the annotated periodic acid-Schiff stain and could further uphold performance on several held-out stains not seen during training. Herein, examples of how this framework can be applied for compartment-specific quantification of immunohistochemical stains for inflammation and fibrosis in animal models and patient biopsy specimens are presented. The results show that stain augmentation is a highly effective approach to enable stain-independent applications of deep-learning segmentation algorithms. This opens new possibilities for broad implementation in digital pathology.

Immune cell infiltration pattern in non-small cell lung cancer PDX models is a model immanent feature and correlates with a distinct molecular and phenotypic make-up.

BACKGROUND: The field of cancer immunology is rapidly moving towards innovative therapeutic strategies, resulting in the need for robust and predictive preclinical platforms reflecting the immunological response to cancer. Well characterized preclinical models are essential for the development of predictive biomarkers in the oncology as well as the immune-oncology space. In the current study, gold standard preclinical models are being refined and combined with novel image analysis tools to meet those requirements. METHODS: A panel of 14 non-small cell lung cancer patient-derived xenograft models (NSCLC PDX) was propagated in humanized NOD/Shi-scid/IL-2Rnull mice. The models were comprehensively characterized for relevant phenotypic and molecular features, including flow cytometry, immunohistochemistry, histology, whole exome sequencing and cytokine secretion. RESULTS: Models reflecting hot (>5% tumor-infiltrating lymphocytes/TILs) as opposed to cold tumors (<5% TILs) significantly differed regarding their cytokine profiles, molecular genetic aberrations, stroma content, and programmed cell death ligand-1 status. Treatment experiments including anti cytotoxic T-lymphocyte-associated protein 4, anti-programmed cell death 1 or the combination thereof across all 14 models in the single mouse trial format showed distinctive tumor growth response and spatial immune cell patterns as monitored by computerized analysis of digitized whole-slide images. Image analysis provided for the first time qualitative evaluation of the extent to which PDX models retain the histological features from their original human donors. CONCLUSIONS: Deep phenotyping of PDX models in a humanized setting by combinations of computational pathology, immunohistochemistry, flow cytometry and proteomics enables the exhaustive analysis of innovative preclinical models and paves the way towards the development of translational biomarkers for immuno-oncology drugs.

Multimodal 3D medical image registration guided by shape encoder-decoder networks.

PURPOSE: Nonlinear multimodal image registration, for example, the fusion of computed tomography (CT) and magnetic resonance imaging (MRI), fundamentally depends on a definition of image similarity. Previous methods that derived modality-invariant representations focused on either global statistical grayscale relations or local structural similarity, both of which are prone to local optima. In contrast to most learning-based methods that rely on strong supervision of aligned multimodal image pairs, we aim to overcome this limitation for further practical use cases. METHODS: We propose a new concept that exploits anatomical shape information and requires only segmentation labels for both modalities individually. First, a shape-constrained encoder-decoder segmentation network without skip connections is jointly trained on labeled CT and MRI inputs. Second, an iterative energy-based minimization scheme is introduced that relies on the capability of the network to generate intermediate nonlinear shape representations. This further eases the multimodal alignment in the case of large deformations. RESULTS: Our novel approach robustly and accurately aligns 3D scans from the multimodal whole-heart segmentation dataset, outperforming classical unsupervised frameworks. Since both parts of our method rely on (stochastic) gradient optimization, it can be easily integrated in deep learning frameworks and executed on GPUs. CONCLUSIONS: We present an integrated approach for weakly supervised multimodal image registration. Achieving promising results due to the exploration of intermediate shape features as registration guidance encourages further research in this direction.