Goal-conditioned reinforcement learning for ultrasound navigation guidance

Transesophageal echocardiography (TEE) plays a pivotal role in cardiology for diagnostic and interventional procedures. However, using it effectively requires extensive training due to the intricate nature of image acquisition and interpretation. To enhance the efficiency of novice sonographers and reduce variability in scan acquisitions, we propose a novel ultrasound (US) navigation assistance method based on contrastive learning as goal-conditioned reinforcement learning (GCRL). We augment the previous framework using a novel contrastive patient batching method (CPB) and a data-augmented contrastive loss, both of which we demonstrate are essential to ensure generalization to anatomical variations across patients. The proposed framework enables navigation to both standard diagnostic as well as intricate interventional views with a single model. Our method was developed with a large dataset of 789 patients and obtained an average error of 6.56 mm in position and 9.36 degrees in angle on a testing dataset of 140 patients, which is competitive or superior to models trained on individual views. Furthermore, we quantitatively validate our method's ability to navigate to interventional views such as the Left Atrial Appendage (LAA) view used in LAA closure. Our approach holds promise in providing valuable guidance during transesophageal ultrasound examinations, contributing to the advancement of skill acquisition for cardiac ultrasound practitioners.

Leveraging Large Language Models (LLMs) to Support Collaborative Human-AI Online Risk Data Annotation

In this position paper, we discuss the potential for leveraging LLMs as interactive research tools to facilitate collaboration between human coders and AI to effectively annotate online risk data at scale. Collaborative human-AI labeling is a promising approach to annotating large-scale and complex data for various tasks. Yet, tools and methods to support effective human-AI collaboration for data annotation are under-studied. This gap is pertinent because co-labeling tasks need to support a two-way interactive discussion that can add nuance and context, particularly in the context of online risk, which is highly subjective and contextualized. Therefore, we provide some of the early benefits and challenges of using LLMs-based tools for risk annotation and suggest future directions for the HCI research community to leverage LLMs as research tools to facilitate human-AI collaboration in contextualized online data annotation. Our research interests align very well with the purposes of the LLMs as Research Tools workshop to identify ongoing applications and challenges of using LLMs to work with data in HCI research. We anticipate learning valuable insights from organizers and participants into how LLMs can help reshape the HCI community's methods for working with data.

Toward Safe Evolution of Artificial Intelligence (AI) based Conversational Agents to Support Adolescent Mental and Sexual Health Knowledge Discovery

Following the recent release of various Artificial Intelligence (AI) based Conversation Agents (CAs), adolescents are increasingly using CAs for interactive knowledge discovery on sensitive topics, including mental and sexual health topics. Exploring such sensitive topics through online search has been an essential part of adolescent development, and CAs can support their knowledge discovery on such topics through human-like dialogues. Yet, unintended risks have been documented with adolescents' interactions with AI-based CAs, such as being exposed to inappropriate content, false information, and/or being given advice that is detrimental to their mental and physical well-being (e.g., to self-harm). In this position paper, we discuss the current landscape and opportunities for CAs to support adolescents' mental and sexual health knowledge discovery. We also discuss some of the challenges related to ensuring the safety of adolescents when interacting with CAs regarding sexual and mental health topics. We call for a discourse on how to set guardrails for the safe evolution of AI-based CAs for adolescents.

Feature boosting with efficient attention for scene parsing

The complexity of scene parsing grows with the number of object and scene classes, which is higher in unrestricted open scenes. The biggest challenge is to model the spatial relation between scene elements while succeeding in identifying objects at smaller scales. This paper presents a novel feature-boosting network that gathers spatial context from multiple levels of feature extraction and computes the attention weights for each level of representation to generate the final class labels. A novel `channel attention module' is designed to compute the attention weights, ensuring that features from the relevant extraction stages are boosted while the others are attenuated. The model also learns spatial context information at low resolution to preserve the abstract spatial relationships among scene elements and reduce computation cost. Spatial attention is subsequently concatenated into a final feature set before applying feature boosting. Low-resolution spatial attention features are trained using an auxiliary task that helps learning a coarse global scene structure. The proposed model outperforms all state-of-the-art models on both the ADE20K and the Cityscapes datasets.

Cardiac ultrasound simulation for autonomous ultrasound navigation

Ultrasound is well-established as an imaging modality for diagnostic and interventional purposes. However, the image quality varies with operator skills as acquiring and interpreting ultrasound images requires extensive training due to the imaging artefacts, the range of acquisition parameters and the variability of patient anatomies. Automating the image acquisition task could improve acquisition reproducibility and quality but training such an algorithm requires large amounts of navigation data, not saved in routine examinations. Thus, we propose a method to generate large amounts of ultrasound images from other modalities and from arbitrary positions, such that this pipeline can later be used by learning algorithms for navigation. We present a novel simulation pipeline which uses segmentations from other modalities, an optimized volumetric data representation and GPU-accelerated Monte Carlo path tracing to generate view-dependent and patient-specific ultrasound images. We extensively validate the correctness of our pipeline with a phantom experiment, where structures' sizes, contrast and speckle noise properties are assessed. Furthermore, we demonstrate its usability to train neural networks for navigation in an echocardiography view classification experiment by generating synthetic images from more than 1000 patients. Networks pre-trained with our simulations achieve significantly superior performance in settings where large real datasets are not available, especially for under-represented classes. The proposed approach allows for fast and accurate patient-specific ultrasound image generation, and its usability for training networks for navigation-related tasks is demonstrated.

Towards Establishing Dense Correspondence on Multiview Coronary Angiography: From Point-to-Point to Curve-to-Curve Query Matching

Coronary angiography is the gold standard imaging technique for studying and diagnosing coronary artery disease. However, the resulting 2D X-ray projections lose 3D information and exhibit visual ambiguities. In this work, we aim to establish dense correspondence in multi-view angiography, serving as a fundamental basis for various clinical applications and downstream tasks. To overcome the challenge of unavailable annotated data, we designed a data simulation pipeline using 3D Coronary Computed Tomography Angiography (CCTA). We formulated the problem of dense correspondence estimation as a query matching task over all points of interest in the given views. We established point-to-point query matching and advanced it to curve-to-curve correspondence, significantly reducing errors by minimizing ambiguity and improving topological awareness. The method was evaluated on a set of 1260 image pairs from different views across 8 clinically relevant angulation groups, demonstrating compelling results and indicating the feasibility of establishing dense correspondence in multi-view angiography.

EEG-Based Reaction Time Prediction with Fuzzy Common Spatial Patterns and Phase Cohesion using Deep Autoencoder Based Data Fusion

Drowsiness state of a driver is a topic of extensive discussion due to its significant role in causing traffic accidents. This research presents a novel approach that combines Fuzzy Common Spatial Patterns (CSP) optimised Phase Cohesive Sequence (PCS) representations and fuzzy CSP-optimized signal amplitude representations. The research aims to examine alterations in Electroencephalogram (EEG) synchronisation between a state of alertness and drowsiness, forecast drivers' reaction times by analysing EEG data, and subsequently identify the presence of drowsiness. The study's findings indicate that this approach successfully distinguishes between alert and drowsy mental states. By employing a Deep Autoencoder-based data fusion technique and a regression model such as Support Vector Regression (SVR) or Least Absolute Shrinkage and Selection Operator (LASSO), the proposed method outperforms using individual feature sets in combination with a regressor model. This superiority is measured by evaluating the Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and Correlation Coefficient (CC). In other words, the fusion of autoencoder-based amplitude EEG power features and PCS features, when used in regression, outperforms using either of these features alone in a regressor model. Specifically, the proposed data fusion method achieves a 14.36% reduction in RMSE, a 25.12% reduction in MAPE, and a 10.12% increase in CC compared to the baseline model using only individual amplitude EEG power features and regression.

Automated CT Lung Cancer Screening Workflow using 3D Camera

Despite recent developments in CT planning that enabled automation in patient positioning, time-consuming scout scans are still needed to compute dose profile and ensure the patient is properly positioned. In this paper, we present a novel method which eliminates the need for scout scans in CT lung cancer screening by estimating patient scan range, isocenter, and Water Equivalent Diameter (WED) from 3D camera images. We achieve this task by training an implicit generative model on over 60,000 CT scans and introduce a novel approach for updating the prediction using real-time scan data. We demonstrate the effectiveness of our method on a testing set of 110 pairs of depth data and CT scan, resulting in an average error of 5mm in estimating the isocenter, 13mm in determining the scan range, 10mm and 16mm in estimating the AP and lateral WED respectively. The relative WED error of our method is 4%, which is well within the International Electrotechnical Commission (IEC) acceptance criteria of 10%.

Temporal DINO: A Self-supervised Video Strategy to Enhance Action Prediction

The emerging field of action prediction plays a vital role in various computer vision applications such as autonomous driving, activity analysis and human-computer interaction. Despite significant advancements, accurately predicting future actions remains a challenging problem due to high dimensionality, complex dynamics and uncertainties inherent in video data. Traditional supervised approaches require large amounts of labelled data, which is expensive and time-consuming to obtain. This paper introduces a novel self-supervised video strategy for enhancing action prediction inspired by DINO (self-distillation with no labels). The Temporal-DINO approach employs two models; a 'student' processing past frames; and a 'teacher' processing both past and future frames, enabling a broader temporal context. During training, the teacher guides the student to learn future context by only observing past frames. The strategy is evaluated on ROAD dataset for the action prediction downstream task using 3D-ResNet, Transformer, and LSTM architectures. The experimental results showcase significant improvements in prediction performance across these architectures, with our method achieving an average enhancement of 9.9% Precision Points (PP), highlighting its effectiveness in enhancing the backbones' capabilities of capturing long-term dependencies. Furthermore, our approach demonstrates efficiency regarding the pretraining dataset size and the number of epochs required. This method overcomes limitations present in other approaches, including considering various backbone architectures, addressing multiple prediction horizons, reducing reliance on hand-crafted augmentations, and streamlining the pretraining process into a single stage. These findings highlight the potential of our approach in diverse video-based tasks such as activity recognition, motion planning, and scene understanding.

Large Scale Mask Optimization Via Convolutional Fourier Neural Operator and Litho-Guided Self Training

Machine learning techniques have been extensively studied for mask optimization problems, aiming at better mask printability, shorter turnaround time, better mask manufacturability, and so on. However, most of these researches are focusing on the initial solution generation of small design regions. To further realize the potential of machine learning techniques on mask optimization tasks, we present a Convolutional Fourier Neural Operator (CFNO) that can efficiently learn layout tile dependencies and hence promise stitch-less large-scale mask optimization with the limited intervention of legacy tools. We discover the possibility of litho-guided self-training (LGST) through a trained machine learning model when solving non-convex optimization problems, which allows iterative model and dataset update and brings significant model performance improvement. Experimental results show that, for the first time, our machine learning-based framework outperforms state-of-the-art academic numerical mask optimizers with an order of magnitude speedup.