Explore the vast range of titles available.

The analysis of histopathology images with artificial intelligence aims to enable clinical decision support systems and precision medicine. The success of such applications depends on the ability to model the diverse patterns observed in pathology images. To this end, we present Virchow, the largest foundation model for computational pathology to date. In addition to the evaluation of biomarker prediction and cell identification, we demonstrate that a large foundation model enables pan-cancer detection, achieving 0.95 specimen-level area under the (receiver operating characteristic) curve across nine common and seven rare cancers. Furthermore, we show that with less training data, the pan-cancer detector built on Virchow can achieve similar performance to tissue-specific clinical-grade models in production and outperform them on some rare variants of cancer. Virchow’s performance gains highlight the value of a foundation model and open possibilities for many high-impact applications with limited amounts of labeled training data. Trained on 1.5 million whole-slide images from 100,000 patients, a pathology foundation model is shown to improve performance of specialized models in detection of rare cancers.

Prostate cancer is a leading cause of morbidity and mortality for adult males in the US. The diagnosis of prostate carcinoma is usually made on prostate core needle biopsies obtained through a transrectal approach. These biopsies may account for a significant portion of the pathologists' workload, yet variability in the experience and expertise, as well as fatigue of the pathologist may adversely affect the reliability of cancer detection. Machine-learning algorithms are increasingly being developed as tools to aid and improve diagnostic accuracy in anatomic pathology. The Paige Prostate AI-based digital diagnostic is one such tool trained on the digital slide archive of New York's Memorial Sloan Kettering Cancer Center (MSKCC) that categorizes a prostate biopsy whole-slide image as either “Suspicious” or “Not Suspicious” for prostatic adenocarcinoma. To evaluate the performance of this program on prostate biopsies secured, processed, and independently diagnosed at an unrelated institution, we used Paige Prostate to review 1876 prostate core biopsy whole-slide images (WSIs) from our practice at Yale Medicine. Paige Prostate categorizations were compared to the pathology diagnosis originally rendered on the glass slides for each core biopsy. Discrepancies between the rendered diagnosis and categorization by Paige Prostate were each manually reviewed by pathologists with specialized genitourinary pathology expertise. Paige Prostate showed a sensitivity of 97.7% and positive predictive value of 97.9%, and a specificity of 99.3% and negative predictive value of 99.2% in identifying core biopsies with cancer in a data set derived from an independent institution. Areas for improvement were identified in Paige Prostate's handling of poor quality scans. Overall, these results demonstrate the feasibility of porting a machine-learning algorithm to an institution remote from its training set, and highlight the potential of such algorithms as a powerful workflow tool for the evaluation of prostate core biopsies in surgical pathology practices.

Prostate cancer diagnosis rests on accurate assessment of tissue by a pathologist. The application of artificial intelligence (AI) to digitized whole slide images (WSIs) can aid pathologists in cancer diagnosis, but robust, diverse evidence in a simulated clinical setting is lacking. To compare the diagnostic accuracy of pathologists reading WSIs of prostatic biopsy specimens with and without AI assistance. Eighteen pathologists, 2 of whom were genitourinary subspecialists, evaluated 610 prostate needle core biopsy WSIs prepared at 218 institutions, with the option for deferral. Two evaluations were performed sequentially for each WSI: initially without assistance, and immediately thereafter aided by Paige Prostate (PaPr), a deep learning-based system that provides a WSI-level binary classification of suspicious for cancer or benign and pinpoints the location that has the greatest probability of harboring cancer on suspicious WSIs. Pathologists' changes in sensitivity and specificity between the assisted and unassisted modalities were assessed, together with the impact of PaPr output on the assisted reads. Using PaPr, pathologists improved their sensitivity and specificity across all histologic grades and tumor sizes. Accuracy gains on both benign and cancerous WSIs could be attributed to PaPr, which correctly classified 100% of the WSIs showing corrected diagnoses in the PaPr-assisted phase. This study demonstrates the effectiveness and safety of an AI tool for pathologists in simulated diagnostic practice, bridging the gap between computational pathology research and its clinical application, and resulted in the first US Food and Drug Administration authorization of an AI system in pathology.

LGBTQ+ persons face heightened vulnerability to climate change-induced disasters due to their sexual orientation, gender identity, and gender expression. Yet, there are limited studies that examine how LGBTQ+ students, particularly in a higher-education setting, understand climate change in relation to their sexuality. With a majority of studies on LGBTQ+ persons and climate change focusing on LGBTQ+ experience during and after disaster, there is a gap in understanding LGBTQ+ perceptions of climate change in the day-to-day. Particularly in Appalachia, a region characterized by a strong place-attachment to and a collective identity with the natural environment, studies of marginalized groups’ perceptions of climate change are lacking.This thesis utilizes a case study of LGBTQ+ undergraduate college students at large, public institutions in Appalachia. Based on semi-structured interviews and focus groups, this thesis investigates the ways in which sexual and place-based identity influence LGBTQ+ undergraduate understandings of the environment and climate change.LGBTQ+ undergraduate students spoke to the unique intersection of Appalachian identity and queer identity, and how this intersection influences their understandings of the environment. As students transition from home to the college space, their understandings of their own sexual and gender identities, as well as of climate change, are further cultivated. Participants also spoke to the privilege that whiteness and class have during present and future climate change impacts, and how such privileges intersect with their queer identity. As they are constantly bombarded by information on a myriad of social and environmental topics in their coursework, the participants understand that climate change is occurring. Yet, they do not understand localized climate change impacts in Appalachia, prompting them to be uncertain about the future, their identities, and climate change.This thesis has broad implications for studies of climate change across the sub-disciplines of feminist, queer, and emotional geography. As there are limited studies on climate change perceptions at the higher education level, and almost no studies in Appalachia specifically, this thesis explores the intersections of identity, emotions, and perceptions.

Robust human pose estimation is of particular interest to the computer vision community, and can be applied to a broad range of applications such as automated surveillance, human-computer interaction, and human activity recognition. In this dissertation, we present a framework for human pose estimation based on stochastic image grammars. Humans in particular are difficult to model, as their articulated geometry, camera viewpoint, and perspective, can produce a very large number of distinct shapes in images. Furthermore, humans often exhibit highly variant and amorphous part appearances, have self-occlusion, and commonly appear in cluttered environments. Our approach capitalizes on the reconfigurable and modular nature of grammatical models to cope with this variability in both geometry and appearance. We present a human body model as a stochastic context-sensitive AND-OR graph grammar, which represents the body as a hierarchical composition of primitive parts while maintaining the articulated kinematics between parts. Each body instance can be composed from a different set of parts and relations in order to explain the unique shape or appearance of that instance. We present grammar models based on coarse-to-fine phrase-structured grammars as well as dependency grammars, and describe efficient algorithms for learning and inference from both generative and discriminative perspectives. Furthermore, we propose extensions to our model to provide ambiguity reasoning in crowded scenes through the use of composite cluster sampling, and reasoning for self-occlusion and external occlusion of parts. We also present a technique to incorporate image segmentation into the part appearance models to improve localization performance on difficult to detect parts. Finally, we demonstrate the effectiveness of our approach by showing state-of-art performance on several recent public benchmark datasets.

The use of artificial intelligence to enable precision medicine and decision support systems through the analysis of pathology images has the potential to revolutionize the diagnosis and treatment of cancer. Such applications will depend on models' abilities to capture the diverse patterns observed in pathology images. To address this challenge, we present Virchow, a foundation model for computational pathology. Using self-supervised learning empowered by the DINOv2 algorithm, Virchow is a vision transformer model with 632 million parameters trained on 1.5 million hematoxylin and eosin stained whole slide images from diverse tissue and specimen types, which is orders of magnitude more data than previous works. The Virchow model enables the development of a pan-cancer detection system with 0.949 overall specimen-level AUC across 17 different cancer types, while also achieving 0.937 AUC on 7 rare cancer types. The Virchow model sets the state-of-the-art on the internal and external image tile level benchmarks and slide level biomarker prediction tasks. The gains in performance highlight the importance of training on massive pathology image datasets, suggesting scaling up the data and network architecture can improve the accuracy for many high-impact computational pathology applications where limited amounts of training data are available.

In this paper, we present a new feature representation for first-person videos. In first-person video understanding (e.g., activity recognition), it is very important to capture both entire scene dynamics (i.e., egomotion) and salient local motion observed in videos. We describe a representation framework based on time series pooling, which is designed to abstract short-term/long-term changes in feature descriptor elements. The idea is to keep track of how descriptor values are changing over time and summarize them to represent motion in the activity video. The framework is general, handling any types of per-frame feature descriptors including conventional motion descriptors like histogram of optical flows (HOF) as well as appearance descriptors from more recent convolutional neural networks (CNN). We experimentally confirm that our approach clearly outperforms previous feature representations including bag-of-visual-words and improved Fisher vector (IFV) when using identical underlying feature descriptors. We also confirm that our feature representation has superior performance to existing state-of-the-art features like local spatio-temporal features and Improved Trajectory Features (originally developed for 3rd-person videos) when handling first-person videos. Multiple first-person activity datasets were tested under various settings to confirm these findings.

Privacy protection from surreptitious video recordings is an important societal challenge. We desire a computer vision system (e.g., a robot) that can recognize human activities and assist our daily life, yet ensure that it is not recording video that may invade our privacy. This paper presents a fundamental approach to address such contradicting objectives: human activity recognition while only using extreme low-resolution (e.g., 16x12) anonymized videos. We introduce the paradigm of inverse super resolution (ISR), the concept of learning the optimal set of image transformations to generate multiple low-resolution (LR) training videos from a single video. Our ISR learns different types of sub-pixel transformations optimized for the activity classification, allowing the classifier to best take advantage of existing high-resolution videos (e.g., YouTube videos) by creating multiple LR training videos tailored for the problem. We experimentally confirm that the paradigm of inverse super resolution is able to benefit activity recognition from extreme low-resolution videos.

With the advent of drones, aerial video analysis becomes increasingly important; yet, it has received scant attention in the literature. This paper addresses a new problem of parsing low-resolution aerial videos of large spatial areas, in terms of 1) grouping, 2) recognizing events and 3) assigning roles to people engaged in events. We propose a novel framework aimed at conducting joint inference of the above tasks, as reasoning about each in isolation typically fails in our setting. Given noisy tracklets of people and detections of large objects and scene surfaces (e.g., building, grass), we use a spatiotemporal AND-OR graph to drive our joint inference, using Markov Chain Monte Carlo and dynamic programming. We also introduce a new formalism of spatiotemporal templates characterizing latent sub-events. For evaluation, we have collected and released a new aerial videos dataset using a hex-rotor flying over picnic areas rich with group events. Our results demonstrate that we successfully address above inference tasks under challenging conditions.

Adoptive chimeric antigen receptor T-cell (CAR-T) therapy is transformative and approved for hematologic malignancies. It is also being developed for the treatment of solid tumors, autoimmune disorders, heart disease, and aging. Despite unprecedented clinical outcomes, CAR-T and other engineered cell therapies face a variety of manufacturing and safety challenges. Traditional methods, such as lentivirus transduction and electroporation, result in random integration or cause significant cellular damage, which can limit the safety and efficacy of engineered cell therapies. We present hydroporation as a gentle and effective alternative for intracellular delivery. Hydroporation resulted in 1.7- to 2-fold higher CAR-T yields compared to electroporation with superior cell viability and recovery. Hydroporated cells exhibited rapid proliferation, robust target cell lysis, and increased pro-inflammatory and regulatory cytokine secretion in addition to improved CAR-T yield by day 5 post-transfection. We demonstrate that scaled-up hydroporation can process 5 × 10 8 cells in less than 10 s, showcasing the platform as a viable solution for high-yield CAR-T manufacturing with the potential for improved therapeutic outcomes.