Search

Purpose: Most studies evaluating artificial intelligence (AI) models that detect abnormalities in neuroimaging are either tested on unrepresentative patient cohorts or are insufficiently well-validated, leading to poor generalisability to real-world tasks. The aim was to determine the diagnostic test accuracy and summarise the evidence supporting the use of AI models performing first-line, high-volume neuroimaging tasks. Methods: Medline, Embase, Cochrane library and Web of Science were searched until September 2021 for studies that temporally or externally validated AI capable of detecting abnormalities in first-line CT or MR neuroimaging. A bivariate random-effects model was used for meta-analysis where appropriate. PROSPERO: CRD42021269563. Results: Only 16 studies were eligible for inclusion. Included studies were not compromised by unrepresentative datasets or inadequate validation methodology. Direct comparison with radiologists was available in 4/16 studies. 15/16 had a high risk of bias. Meta-analysis was only suitable for intracranial haemorrhage detection in CT imaging (10/16 studies), where AI systems had a pooled sensitivity and specificity 0.90 (95% CI 0.85 - 0.94) and 0.90 (95% CI 0.83 - 0.95) respectively. Other AI studies using CT and MRI detected target conditions other than haemorrhage (2/16), or multiple target conditions (4/16). Only 3/16 studies implemented AI in clinical pathways, either for pre-read triage or as post-read discrepancy identifiers. Conclusion: The paucity of eligible studies reflects that most abnormality detection AI studies were not adequately validated in representative clinical cohorts. The few studies describing how abnormality detection AI could impact patients and clinicians did not explore the full ramifications of clinical implementation.

This letter critically examines the recent article by Infante et al. assessing the utility of large language models (LLMs) like GPT-4, Perplexity, and Bard in identifying urgent findings in emergency radiology reports. While acknowledging the potential of LLMs in generating labels for computer vision, concerns are raised about the ethical implications of using patient data without explicit approval, highlighting the necessity of stringent data protection measures under GDPR.

Artificial neural networks trained on large, expert-labelled datasets are considered state-of-the-art for a range of medical image recognition tasks. However, categorically labelled datasets are time-consuming to generate and constrain classification to a pre-defined, fixed set of classes. For neuroradiological applications in particular, this represents a barrier to clinical adoption. To address these challenges, we present a self-supervised text-vision framework that learns to detect clinically relevant abnormalities in brain MRI scans by directly leveraging the rich information contained in accompanying free-text neuroradiology reports. Our training approach consisted of two-steps. First, a dedicated neuroradiological language model - NeuroBERT - was trained to generate fixed-dimensional vector representations of neuroradiology reports (N = 50,523) via domain-specific self-supervised learning tasks. Next, convolutional neural networks (one per MRI sequence) learnt to map individual brain scans to their corresponding text vector representations by optimising a mean square error loss. Once trained, our text-vision framework can be used to detect abnormalities in unreported brain MRI examinations by scoring scans against suitable query sentences (e.g., 'there is an acute stroke', 'there is hydrocephalus' etc.), enabling a range of classification-based applications including automated triage. Potentially, our framework could also serve as a clinical decision support tool, not only by suggesting findings to radiologists and detecting errors in provisional reports, but also by retrieving and displaying examples of pathologies from historical examinations that could be relevant to the current case based on textual descriptors., Comment: Under Review

Function-as-a-Service is a cloud computing paradigm offering an event-driven execution model to applications. It features serverless attributes by eliminating resource management responsibilities from developers and offers transparent and on-demand scalability of applications. Typical serverless applications have stringent response time and scalability requirements and therefore rely on deployed services to provide quick and fault-tolerant feedback to clients. However, the FaaS paradigm suffers from cold starts as there is a non-negligible delay associated with on-demand function initialization. This work focuses on reducing the frequency of cold starts on the platform by using Reinforcement Learning. Our approach uses Q-learning and considers metrics such as function CPU utilization, existing function instances, and response failure rate to proactively initialize functions in advance based on the expected demand. The proposed solution was implemented on Kubeless and was evaluated using a normalised real-world function demand trace with matrix multiplication as the workload. The results demonstrate a favourable performance of the RL-based agent when compared to Kubeless' default policy and function keep-alive policy by improving throughput by up to 8.81% and reducing computation load and resource wastage by up to 55% and 37%, respectively, which is a direct outcome of reduced cold starts., Comment: 13 figures, 10 pages, 3 tables

Function-as-a-Service (FaaS) introduces a lightweight, function-based cloud execution model that finds its relevance in applications like IoT-edge data processing and anomaly detection. While CSP offer a near-infinite function elasticity, these applications often experience fluctuating workloads and stricter performance constraints. A typical CSP strategy is to empirically determine and adjust desired function instances, "autoscaling", based on monitoring-based thresholds such as CPU or memory, to cope with demand and performance. However, threshold configuration either requires expert knowledge, historical data or a complete view of environment, making autoscaling a performance bottleneck lacking an adaptable solution.RL algorithms are proven to be beneficial in analysing complex cloud environments and result in an adaptable policy that maximizes the expected objectives. Most realistic cloud environments usually involve operational interference and have limited visibility, making them partially observable. A general solution to tackle observability in highly dynamic settings is to integrate Recurrent units with model-free RL algorithms and model a decision process as a POMDP. Therefore, in this paper, we investigate a model-free Recurrent RL agent for function autoscaling and compare it against the model-free Proximal Policy Optimisation (PPO) algorithm. We explore the integration of a LSTM network with the state-of-the-art PPO algorithm to find that under our experimental and evaluation settings, recurrent policies were able to capture the environment parameters and show promising results for function autoscaling. We further compare a PPO-based autoscaling agent with commercially used threshold-based function autoscaling and posit that a LSTM-based autoscaling agent is able to improve throughput by 18%, function execution by 13% and account for 8.4% more function instances., Comment: 12 pages, 13 figures, 4 tables

The ever-growing demands for memory with larger capacity and higher bandwidth have driven recent innovations on memory expansion and disaggregation technologies based on Compute eXpress Link (CXL). Especially, CXL-based memory expansion technology has recently gained notable attention for its ability not only to economically expand memory capacity and bandwidth but also to decouple memory technologies from a specific memory interface of the CPU. However, since CXL memory devices have not been widely available, they have been emulated using DDR memory in a remote NUMA node. In this paper, for the first time, we comprehensively evaluate a true CXL-ready system based on the latest 4th-generation Intel Xeon CPU with three CXL memory devices from different manufacturers. Specifically, we run a set of microbenchmarks not only to compare the performance of true CXL memory with that of emulated CXL memory but also to analyze the complex interplay between the CPU and CXL memory in depth. This reveals important differences between emulated CXL memory and true CXL memory, some of which will compel researchers to revisit the analyses and proposals from recent work. Next, we identify opportunities for memory-bandwidth-intensive applications to benefit from the use of CXL memory. Lastly, we propose a CXL-memory-aware dynamic page allocation policy, Caption to more efficiently use CXL memory as a bandwidth expander. We demonstrate that Caption can automatically converge to an empirically favorable percentage of pages allocated to CXL memory, which improves the performance of memory-bandwidth-intensive applications by up to 24% when compared to the default page allocation policy designed for traditional NUMA systems., Comment: This paper has been accepted by MICRO'23. Please refer to the https://doi.org/10.1145/3613424.3614256 for the official version of this paper

Full-system simulation of computer systems is critical to capture the complex interplay between various hardware and software components in future systems. Modeling the network subsystem is indispensable to the fidelity of the full-system simulation due to the increasing importance of scale-out systems. The network software stack has undergone major changes over the last decade, and kernel-bypass networking stacks and data-plane networks are rapidly replacing the conventional kernel network stack. Nevertheless, the current state-of-the-art architectural simulators still use kernel networking which precludes realistic network application scenarios. In this work, we enable kernel bypass networking stack on gem5, the state-of-the-art full-system architectural simulator. We extend gem5's NIC hardware model and device driver to enable the support for userspace device drivers to run the DPDK framework. We also implement a network load generator hardware model in gem5 to generate various traffic patterns and perform per-packet timestamp and latency measurements without introducing packet loss. Our experimental results show that DPDK's simulated network bandwidth scales with the number of cores and NIC ports. As two use cases, we analyze the sensitivity of (1) network performance to several microarchitectural parameters, and (2) direct cache access (DCA) technology to DPDK burst size., Comment: 4 Pages, 4 Figures and 1 Table

An important inference from Neural Tangent Kernel (NTK) theory is the existence of spectral bias (SB), that is, low frequency components of the target function of a fully connected Artificial Neural Network (ANN) being learnt significantly faster than the higher frequencies during training. This is established for Mean Square Error (MSE) loss functions with very low learning rate parameters. Physics Informed Neural Networks (PINNs) are designed to learn the solutions of differential equations (DE) of arbitrary orders; in PINNs the loss functions are obtained as the residues of the conservative form of the DEs and represent the degree of dissatisfaction of the equations. So there has been an open question whether (a) PINNs also exhibit SB and (b) if so, how does this bias vary across the orders of the DEs. In this work, a series of numerical experiments are conducted on simple sinusoidal functions of varying frequencies, compositions and equation orders to investigate these issues. It is firmly established that under normalized conditions, PINNs do exhibit strong spectral bias, and this increases with the order of the differential equation.

The growing demand for head magnetic resonance imaging (MRI) examinations, along with a global shortage of radiologists, has led to an increase in the time taken to report head MRI scans around the world. For many neurological conditions, this delay can result in increased morbidity and mortality. An automated triaging tool could reduce reporting times for abnormal examinations by identifying abnormalities at the time of imaging and prioritizing the reporting of these scans. In this work, we present a convolutional neural network for detecting clinically-relevant abnormalities in $\text{T}_2$-weighted head MRI scans. Using a validated neuroradiology report classifier, we generated a labelled dataset of 43,754 scans from two large UK hospitals for model training, and demonstrate accurate classification (area under the receiver operating curve (AUC) = 0.943) on a test set of 800 scans labelled by a team of neuroradiologists. Importantly, when trained on scans from only a single hospital the model generalized to scans from the other hospital ($\Delta$AUC $\leq$ 0.02). A simulation study demonstrated that our model would reduce the mean reporting time for abnormal examinations from 28 days to 14 days and from 9 days to 5 days at the two hospitals, demonstrating feasibility for use in a clinical triage environment., Comment: Accepted as an oral presentation at Medical Imaging with Deep Learning (MIDL) 2021

We describe a light-weight, weather and lighting invariant, Semantic Bird's Eye View (S-BEV) signature for vision-based vehicle re-localization. A topological map of S-BEV signatures is created during the first traversal of the route, which are used for coarse localization in subsequent route traversal. A fine-grained localizer is then trained to output the global 3-DoF pose of the vehicle using its S-BEV and its coarse localization. We conduct experiments on vKITTI2 virtual dataset and show the potential of the S-BEV to be robust to weather and lighting. We also demonstrate results with 2 vehicles on a 22 km long highway route in the Ford AV dataset., Comment: 7 pages, 8 figures

Fine localization is a crucial task for autonomous vehicles. Although many algorithms have been explored in the literature for this specific task, the goal of getting accurate results from commodity sensors remains a challenge. As autonomous vehicles make the transition from expensive prototypes to production items, the need for inexpensive, yet reliable solutions is increasing rapidly. This article considers scenarios where images are captured with inexpensive cameras and localization takes place using pre-loaded fine maps of local roads as side information. The techniques proposed herein extend schemes based on normalized mutual information by leveraging the likelihood of shades rather than exact sensor readings for localization in noisy environments. This algorithmic enhancement, rooted in statistical signal processing, offers substantial gains in performance. Numerical simulations are used to highlight the benefits of the proposed techniques in representative application scenarios. Analysis of a Ford image set is performed to validate the core findings of this work., Comment: 5 pages, 9 figures, to be published in 2020 IEEE Conference on Applied Signal Processing (ASPCON)

This paper presents a localization technique using aerial imagery maps and LIDAR based ground reflectivity for autonomous vehicles in urban environments. Traditional localization techniques using LIDAR reflectivity rely on high definition reflectivity maps generated from a mapping vehicle. The cost and effort required to maintain such prior maps are generally very high because it requires a fleet of expensive mapping vehicles. In this work we propose a localization technique where the vehicle localizes using aerial/satellite imagery, eradicating the need to develop and maintain complex high-definition maps. The proposed technique has been tested on a real world dataset collected from a test track in Ann Arbor, Michigan. This research concludes that aerial imagery based maps provides real-time localization performance similar to state-of-the-art LIDAR based maps for autonomous vehicles in urban environments at reduced costs., Comment: 6 pages, 7 figures, Submitted to International Conference on Intelligent Robots and Systems (IROS-2020), For the video, see https://www.youtube.com/watch?v=vcY74Z9bOLk

This paper presents a challenging multi-agent seasonal dataset collected by a fleet of Ford autonomous vehicles at different days and times during 2017-18. The vehicles traversed an average route of 66 km in Michigan that included a mix of driving scenarios such as the Detroit Airport, freeways, city-centers, university campus and suburban neighbourhoods, etc. Each vehicle used in this data collection is a Ford Fusion outfitted with an Applanix POS-LV GNSS system, four HDL-32E Velodyne 3D-lidar scanners, 6 Point Grey 1.3 MP Cameras arranged on the rooftop for 360-degree coverage and 1 Pointgrey 5 MP camera mounted behind the windshield for the forward field of view. We present the seasonal variation in weather, lighting, construction and traffic conditions experienced in dynamic urban environments. This dataset can help design robust algorithms for autonomous vehicles and multi-agent systems. Each log in the dataset is time-stamped and contains raw data from all the sensors, calibration values, pose trajectory, ground truth pose, and 3D maps. All data is available in Rosbag format that can be visualized, modified and applied using the open-source Robot Operating System (ROS). We also provide the output of state-of-the-art reflectivity-based localization for bench-marking purposes. The dataset can be freely downloaded at our website., Comment: 7 pages, 7 figures, Submitted to International Journal of Robotics Research (IJRR), Visit website at https://avdata.ford.com

Fine localization in autonomous driving platforms is a task of broad interest, receiving much attention in recent years. Some localization algorithms use the Euclidean distance as a similarity measure between the local image acquired by a camera and a global map, which acts as side information. The global map is typically expressed in terms of the coordinate system of the road plane. Yet, a road image captured by a camera is subject to distortion in that nearby features on the road have much larger footprints on the focal plane of the camera compared with those of equally-sized features that lie farther ahead of the vehicle. Using commodity computational tools, it is straightforward to execute a transformation and, thereby, bring the distorted image into the frame of reference of the global map. However, this nonlinear transformation results in unequal noise amplification. The noise profile induced by this transformation should be accounted for when trying to match an acquired image to a global map, with more reliable regions being given more weight in the process. This physical reality presents an algorithmic opportunity to improve existing localization algorithms, especially in harsh conditions. This article reviews the physics of road feature acquisition through a camera, and it proposes an improved matching method rooted in statistical analysis. Findings are supported by numerical simulations., Comment: 5 pages, 7 figures, to be published in 2019 IEEE Global Conference on Signal and Information Processing (GlobalSIP)