Indiana University-Purdue University Indianapolis (IUPUI) / Background: Cancer is the leading cause of death worldwide and as a result is one of the most studied topics in public health. Breast cancer and prostate cancer are the most common cancers among women and men respectively. Gene expression and image features are independently prognostic of patient survival. However, it is sometimes difficult to discern how the molecular profile, e.g., gene expression, of given cells relate to their spatial layout, i.e., topology, in the tumor microenvironment (TME). However, with the advent of spatial transcriptomics (ST) and integrative bioinformatics analysis techniques, we are now able to better understand the TME of common cancers. Method: In this paper, we aim to determine the genes that are correlated with image topology features (ITFs) in common cancers which we denote topology associated genes (TAGs). To achieve this objective, we generate the correlation coefficient between genes and image features after identifying the optimal number of clusters for each of them. Applying this correlation matrix to heatmap using R package pheatmap to visualize the correlation between the two sets. The objective of this study is to identify common themes for the genes correlated with ITFs and we can pursue this using functional enrichment analysis. Moreover, we also find the similarity between gene clusters and some image features clusters using the ranking of correlation coefficient in order to identify, compare and contrast the TAGs across breast and prostate cancer ST slides. Result: The analysis shows that there are groups of gene ontology terms that are common within breast cancer, prostate cancer, and across both cancers. Notably, extracellular matrix (ECM) related terms appeared regularly in all ST slides. Conclusion: We identified TAGs in every ST slide regardless of cancer type. These TAGs were enriched for ontology terms that add context to the ITFs generated from ST cancer slides.

No description available.

No description available.

<p dir="ltr">Generative Image-to-Image Translation (I2IT) involves transforming an input image from one domain to another. Typically, this transformation retains the content in the input image while adjusting the domain-dependent style elements. Generative I2IT finds utility in a wide range of applications, yet its effectiveness hinges on adaptations to the unique characteristics of the data at hand. This dissertation pushes the boundaries of I2IT by applying it to stain-related problems in computational pathology. Particularly, the main contributions span two major applications of stain translation: H&E-to-H&E and H&E-to-IHC, each with its unique requirements and challenges. More specifically, the first contribution addresses the generalization challenge posed by the high variability in H&E stain appearances to any task-specific machine learning models. To this end, the Generative Stain Augmentation Network (G-SAN) is introduced to augment the training images in any downstream task with random and diverse H&E stain appearances. Experimental results demonstrate G-SAN’s ability to enhance model generalization across stain variations in downstream tasks. The second key contribution in this dissertation focuses on H&E-to-IHC stain translation. The major challenge in learning accurate H&E-to-IHC stain translation is the frequent and sometimes severe inconsistencies in the groundtruth H&E-IHC image pairs. To make training more robust to these inconsistencies, a novel contrastive learning based loss, named the Adaptive Supervised PatchNCE (ASP) loss is presented. Experimental results suggest that the proposed ASP-based framework outperforms the state-of-the-art in H&E-to-IHC stain translation by significant margins. Additionally, a new dataset for H&E-to-IHC translation – the Multi-IHC Stain Translation (MIST) dataset, is released to the public, featuring paired images from H&E to four different IHC stains. For future directions of generative I2IT in stain translation problems, a proof-of-concept study of applying the latest diffusion model based I2IT methods to the problem of virtual H&E staining is presented.</p>

Tesis por compendio / [ES] Esta tesis presenta soluciones de vanguardia basadas en algoritmos de computer vision (CV) y machine learning (ML) para ayudar a los expertos en el diagnóstico clínico. Se centra en dos áreas relevantes en el campo de la imagen médica: la patología digital y la oftalmología. Este trabajo propone diferentes paradigmas de machine learning y deep learning para abordar diversos escenarios de supervisión en el estudio del cáncer de próstata, el cáncer de vejiga y el glaucoma. En particular, se consideran métodos supervisados convencionales para segmentar y clasificar estructuras específicas de la próstata en imágenes histológicas digitalizadas. Para el reconocimiento de patrones específicos de la vejiga, se llevan a cabo enfoques totalmente no supervisados basados en técnicas de deep-clustering. Con respecto a la detección del glaucoma, se aplican algoritmos de memoria a corto plazo (LSTMs) que permiten llevar a cabo un aprendizaje recurrente a partir de volúmenes de tomografía por coherencia óptica en el dominio espectral (SD-OCT). Finalmente, se propone el uso de redes neuronales prototípicas (PNN) en un marco de few-shot learning para determinar el nivel de gravedad del glaucoma a partir de imágenes OCT circumpapilares. Los métodos de inteligencia artificial (IA) que se detallan en esta tesis proporcionan una valiosa herramienta de ayuda al diagnóstico por imagen, ya sea para el diagnóstico histológico del cáncer de próstata y vejiga o para la evaluación del glaucoma a partir de datos de OCT. / [CA] Aquesta tesi presenta solucions d'avantguarda basades en algorismes de *computer *vision (CV) i *machine *learning (ML) per a ajudar als experts en el diagnòstic clínic. Se centra en dues àrees rellevants en el camp de la imatge mèdica: la patologia digital i l'oftalmologia. Aquest treball proposa diferents paradigmes de *machine *learning i *deep *learning per a abordar diversos escenaris de supervisió en l'estudi del càncer de pròstata, el càncer de bufeta i el glaucoma. En particular, es consideren mètodes supervisats convencionals per a segmentar i classificar estructures específiques de la pròstata en imatges histològiques digitalitzades. Per al reconeixement de patrons específics de la bufeta, es duen a terme enfocaments totalment no supervisats basats en tècniques de *deep-*clustering. Respecte a la detecció del glaucoma, s'apliquen algorismes de memòria a curt termini (*LSTMs) que permeten dur a terme un aprenentatge recurrent a partir de volums de tomografia per coherència òptica en el domini espectral (SD-*OCT). Finalment, es proposa l'ús de xarxes neuronals *prototípicas (*PNN) en un marc de *few-*shot *learning per a determinar el nivell de gravetat del glaucoma a partir d'imatges *OCT *circumpapilares. Els mètodes d'intel·ligència artificial (*IA) que es detallen en aquesta tesi proporcionen una valuosa eina d'ajuda al diagnòstic per imatge, ja siga per al diagnòstic histològic del càncer de pròstata i bufeta o per a l'avaluació del glaucoma a partir de dades d'OCT. / [EN] This thesis presents cutting-edge solutions based on computer vision (CV) and machine learning (ML) algorithms to assist experts in clinical diagnosis. It focuses on two relevant areas at the forefront of medical imaging: digital pathology and ophthalmology. This work proposes different machine learning and deep learning paradigms to address various supervisory scenarios in the study of prostate cancer, bladder cancer and glaucoma. In particular, conventional supervised methods are considered for segmenting and classifying prostate-specific structures in digitised histological images. For bladder-specific pattern recognition, fully unsupervised approaches based on deep-clustering techniques are carried out. Regarding glaucoma detection, long-short term memory algorithms (LSTMs) are applied to perform recurrent learning from spectral-domain optical coherence tomography (SD-OCT) volumes. Finally, the use of prototypical neural networks (PNNs) in a few-shot learning framework is proposed to determine the severity level of glaucoma from circumpapillary OCT images. The artificial intelligence (AI) methods detailed in this thesis provide a valuable tool to aid diagnostic imaging, whether for the histological diagnosis of prostate and bladder cancer or glaucoma assessment from OCT data. / García Pardo, JG. (2022). Machine learning strategies for diagnostic imaging support on histopathology and optical coherence tomography [Tesis doctoral]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/182400 / Compendio