A5079688100- Cell Image Analysis Techniques
- AI in cancer detection
- Single-cell and spatial transcriptomics
- Cancer Cells and Metastasis
- Cancer Genomics and Diagnostics
- Radiomics and Machine Learning in Medical Imaging
- Advanced Biosensing Techniques and Applications
- Cancer Research and Treatments
- Medical Imaging Techniques and Applications
- Endometrial and Cervical Cancer Treatments
- Lung Cancer Diagnosis and Treatment
- Digital Imaging for Blood Diseases
- Nanoplatforms for cancer theranostics
- 3D Printing in Biomedical Research
- Cancer Immunotherapy and Biomarkers
- Remote-Sensing Image Classification
- Photodynamic Therapy Research Studies
- Neutrophil, Myeloperoxidase and Oxidative Mechanisms
- Diagnosis and treatment of tuberculosis
- Long-Term Effects of COVID-19
- Lung Cancer Treatments and Mutations
- Cervical Cancer and HPV Research
- Ovarian cancer diagnosis and treatment
- vaccines and immunoinformatics approaches
- COVID-19 Clinical Research Studies
Sidney Kimmel Comprehensive Cancer Center
2025
Johns Hopkins University
2023-2025
Bloomberg (United States)
2025
Bruker (United States)
2024
Johns Hopkins Medicine
2023-2024
Universidad de Navarra
2019-2023
Clinica Universidad de Navarra
2022
Lawrence Berkeley National Laboratory
2022
Understanding the spatial interactions between elements of tumor microenvironment -i.e. cells. fibroblasts, immune cells- and how these relate to diagnosis or prognosis a is one goals computational pathology. We present NaroNet, deep learning framework that models multi-scale from multiplex-stained cancer tissue images provides patient-level interpretable predictions using seamless end-to-end pipeline. Trained only with their corresponding clinical labels, NaroNet unsupervisedly learns which...
Objectives Multiplex immunohistochemistry and immunofluorescence (mIHC/IF) are emerging technologies that can be used to help define complex immunophenotypes in tissue, quantify immune cell subsets, assess the spatial arrangement of marker expression. mIHC/IF assays require concerted efforts optimize validate multiplex staining protocols prior their application on slides. The best practice guidelines for validation across platforms were previously published by this task force. current effort...
Abstract Endometrial tumors show substantial heterogeneity in their immune microenvironment. This could be used to improve the accuracy of current outcome prediction tools. We assessed microenvironment 235 patients diagnosed with low‐grade, early‐stage endometrial cancer. Multiplex quantitative immunofluorescence was carried out measure CD8, CD68, FOXP3, PD‐1, and PD‐L1 markers, as well cytokeratin (CK), on tissue microarrays. Clustering results revealed five robust response patterns, each...
Abstract Predicting recurrence in low-grade, early-stage endometrial cancer (EC) is both challenging and clinically relevant. We present a weakly-supervised deep learning framework, NaroNet, that can learn, without manual expert annotation, the complex tumor-immune interrelations at three levels: local phenotypes, cellular neighborhoods, tissue areas. It uses multiplexed immunofluorescence for simultaneous visualization quantification of CD68 + macrophages, CD8 T cells, FOXP3 regulatory...
The tissue immune microenvironment is associated with key aspects of tumor biology. interaction between the system and cancer cells has predictive prognostic potential across different types. Spatially resolved tissue-based technologies allowed researchers to simultaneously quantify populations in samples. However, bare quantification fails harness spatial nature technologies. Tumor-immune interactions are specific patterns that can be measured. In recent years, several computational tools...
Abstract Motivation Recent advances in multiplex immunostaining and multispectral cytometry have opened the door to simultaneously visualizing an unprecedented number of biomarkers both liquid solid samples. Properly unmixing fluorescent emissions is a challenging task, which normally requires characterization individual fluorochromes from control As increases, cost time use reagents becomes prohibitively high. Here, we present fully unsupervised blind spectral method for separation highly...
New machine learning models designed to capture the histopathology of tissues should account not only for phenotype and morphology cells, but also learn complex spatial relationships between them. To achieve this, we represent tissue as an interconnected graph, where previously segmented cells become nodes graph. Then are learned embedded into a low-dimensional vector, using Graph Neural Network. We name this Representation Learning based strategy NARO (NAtural biological Objects),...
<h3>Background</h3> In recent years, our understanding of tumor development has advanced substantially by revealing significant transcriptomic and proteomic heterogeneity within the tumor-immune microenvironment (TIME). However, most analytical, diagnostic approaches only provide one type biological readout either lack spatial information or multiplexing capabilities to effectively address TIME heterogeneity. Spatial multi-omics combines high-plex transcriptomics proteomics investigate in...
<h3>Background</h3> The demand for high-plex biomarker detection <i>in situ</i> has surged in recent years. Consequently, various cyclic immunofluorescence staining techniques have been developed to increase plexity. These require removal of fluorescence signals after each cycle, generally by chemical stripping antibodies, photobleaching fluorophores, or deployment barcoded antibodies. Each these signal methods limitations which contribute unique challenges the development, validation, and...
<h3>Background</h3> The identification of protein biomarkers in situ holds great promise for advancing disease diagnoses and treatment options. To be considered inclusion, markers must meet two main criteria: specificity, indicating they should unique to the tissue or cell type being studied, sensitivity, meaning that their levels sufficient accurately detect changes expression. This study assesses sensitivity CellScape™ Precise Spatial Multiplexing platform using High Dynamic Range (HDR)...
Abstract Understanding the spatial distribution of key cell populations is critical in advancing biomedical research and development novel therapeutics. Highly multiplexed biomarker analysis achieved with single-cell context on CellScape, a microscopy platform that enables quantitative phenotyping entire tissue sections mounted standard histology slides. Leveraging high-resolution, high dynamic range imaging, automated reagent delivery, CellScape represents an attractive for biology...
Multiplex immunofluorescence is a novel, high-content imaging technique that allows simultaneous in situ labeling of multiple tissue antigens. This growing relevance the study tumor microenvironment, and discovery biomarkers disease progression or response to immune-based therapies. Given number markers potential complexity spatial interactions involved, analysis these images requires use machine learning tools rely for their training on availability large image datasets, extremely laborious...
Multiplex tissue immunostaining is a novel technology of growing relevance as it can capture in situ the complex interactions existing between multiple elements tumor microenvironment. The existence and availability large, annotated image datasets key for objective development, training benchmarking bioimage analysis algorithms, especially those based on machine learning. Manual annotations multiplex images are, however, extremely laborious, often impracticable. In this paper, we present...
<h3>Background</h3> Lung adenocarcinomas (LUAD) with co-mutations in KRAS (K-only) and the SKT11/LKB1 (KS) or TP53 (KP) genes define patient subgroups distinct responses to anti-PD1/PD-L1 immunotherapy. In fact, multicentric studies showed that objective response rates PD-1 blockade differed significantly among KS (7.4%), KP (35.7%), K-only (28.6%) subgroups.<sup>1</sup> The association of such specific genetic profiling morphological patterns assessed on routine H&E tissue slides may...
Many efforts have been made to discover tumor-specific microenvironment elements (TMEs) from immunostained tissue sections. However, the identification of yet unknown but relevant TMEs multiplex tissues remains a challenge, due number markers involved (tens) and complexity their spatial interactions. We present NaroNet, which uses machine learning identify annotate known as well novel self-supervised embeddings cells, organized at different levels (local cell phenotypes cellular...
Multiplex tissue immunostaining is a technology of growing relevance as it can capture in situ the complex interactions existing between elements tumor microenvironment. The existence and availability large, annotated image datasets key for objective development benchmarking bioimage analysis algorithms. Manual annotation multiplex images, however, laborious, often impracticable. In this paper, we present Synplex, simulation system able to generate immunostained images based on user-defined...