Publications

Assessing Intrapartum Risk of Hypoxic-Ischemic Encephalopathy using Fetal Heart Rate with Long Short-term Memory Networks
"Derek Kweku DEGBEDZUI
Michael Kuzniewicz
Cornet Marie-Coralie
Yvonne Wu
Heather Forquer
Lawrence Gerstley
Emily Hamilton
Doina Precup
Philip Warrick
Robert Kearney"
This study investigated the prediction of the risk of hypoxic ischemic encephalopathy using intrapartum cardiotocography records with a long… (voir plus) short-term memory re-current neural network. Across the 12 hours of labour, HIE sensitivity rose from 0.25 to 0.56 as delivery approached while specificity remained approximately constant with a mean of 0.71 and standard deviation of 0.04. The results show that classification improves as delivery approaches but that performance needs improvement. Future work will address the limitations of this preliminary study by investigating input signal transformations and the use of other network architectures to improve the model performance.
2022-12-30
Computing in Cardiology Conference (CinC) (publié)
doi.org
Offline Policy Optimization in RL with Variance Regularizaton
Riashat Islam
Samarth Sinha
Homanga Bharadhwaj
Samin Yeasar Arnob
Zhuoran Yang
Animesh Garg
Zhaoran Wang
Lihong Li
Doina Precup
2022-12-28
arXiv (prépublication)
doi.org
arxiv.org
Orientation and Context Entangled Network for Retinal Vessel Segmentation
Xinxu Wei
Kaifu Yang
Danilo Bzdok
Yongjie Li
Most of the existing deep learning based methods for vessel segmentation neglect two important aspects of retinal vessels, one is the orient… (voir plus)ation information of vessels, and the other is the contextual information of the whole fundus region. In this paper, we propose a robust Orientation and Context Entangled Network (denoted as OCE-Net), which has the capability of extracting complex orientation and context information of the blood vessels. To achieve complex orientation aware, a Dynamic Complex Orientation Aware Convolution (DCOA Conv) is proposed to extract complex vessels with multiple orientations for improving the vessel continuity. To simultaneously capture the global context information and emphasize the important local information, a Global and Local Fusion Module (GLFM) is developed to simultaneously model the long-range dependency of vessels and focus sufficient attention on local thin vessels. A novel Orientation and Context Entangled Non-local (OCE-NL) module is proposed to entangle the orientation and context information together. In addition, an Unbalanced Attention Refining Module (UARM) is proposed to deal with the unbalanced pixel numbers of background, thick and thin vessels. Extensive experiments were performed on several commonly used datasets (DRIVE, STARE and CHASEDB1) and some more challenging datasets (AV-WIDE, UoA-DR, RFMiD and UK Biobank). The ablation study shows that the proposed method achieves promising performance on maintaining the continuity of thin vessels and the comparative experiments demonstrate that our OCE-Net can achieve state-of-the-art performance on retinal vessel segmentation.
2022-12-28
Expert Systems with Applications (inconnu)
doi.org
arxiv.org
Picture Cards Versus Physical Examination: A Proof-of-Concept Study to Improve the SOSAS Survey Tool.
Adesoji Ademuyiwa
Benedict C. Nwomeh
Dan Poenaru
Justina O. Seyi-Olajide
Iyabo Y. Ademuyiwa
Tinuola O. Odugbemi
Ogechi Abazie
Oluwaseun A. Ladipo-Ajayi
Olufemi Bankole
Olumide A. Elebute
Babasola Okusanya
Felix M. Alakaloko
Eyitayo O. Alabi
Ayomide Makanjuola
Shailvi Gupta
Tu Tran
Amanda Onwuka A
Emily R. Smith
Riinu Pius
Ewen Harrison … (voir 1 de plus)
Christopher O. Bode
2022-12-26
Journal of Surgical Research (publié)
doi.org
Simulation-based power analysis could improve the design of clinical trials in Alzheimer's disease
Daniel Andrews
Douglas L. Arnold
Danilo Bzdok
Simon Ducharme
Howard Chertkow
D. Louis Collins
Clinical trials of new treatments in different progressive diseases use power analysis to determine the sample size needed for a trial to ob… (voir plus)tain a statistically significant estimate for an anticipated treatment effect. In trials with parallel designs, the standard power analysis approach is based on a two-sample t-test. For example, the standard t-test approach was used in determining the sample size for the Phase 3 trials of aducanumab, the first drug approved by the United States Food and Drug Administration (FDA) to potentially slow cognitive decline in early-stage Alzheimer’s disease. However, t-tests contain normality assumptions, and t-test-based power analyses do not implicitly factor in the uncertainty about anticipated treatment effects that arises due to inter-subject heterogeneity in disease progression. These limitations may lead to recommended sample sizes that are too small, potentially making a trial blind to a treatment effect that is truly present if the cohort’s endpoints are not normally distributed and/or the anticipated treatment effect is overestimated. To address these issues, we present a novel power analysis method that (1) simulates clinical trials in a progressive disease using real-world data, (2) accounts for inter-subject heterogeneity in disease progression, and (3) does not depend on normality assumptions. As a showcase example, we used our method to calculate power for a range of sample sizes and treatment effects in simulated trials similar to the Phase 3 aducanumab trials EMERGE and ENGAGE. As expected, our results show that power increases with number of subjects and treatment effect (here defined as the cohort-level percent reduction in the rate of cognitive decline in treated subjects vs. controls). However, inclusion of realistic inter-subject heterogeneity in cognitive decline trajectories leads to increased sample size recommendations compared to a standard t-test power analysis. These results suggest that the sample sizes recommended by the t-test power analyses in the EMERGE and ENGAGE Statistical Analysis Plans were possibly too small to ensure a high probability of detecting the anticipated treatment effect. Insufficient sample sizes could partly explain the statistically significant effect of aducanumab being detected only in EMERGE. We also used our method to analyze power in simulated trials similar the Phase 3 lecanemab trial Clarity AD. Our results suggest that Clarity AD was adequately powered, and that power may be influenced by a trial’s number of analysis visits and the characteristics of subgroups within a cohort. By using our simulation-based power analysis approach, clinical trials of treatments in Alzheimer’s disease and potentially in other progressive diseases could obtain sample size recommendations that account for heterogeneity in disease progression and uncertainty in anticipated treatment effects. Our approach avoids the limitations of t-tests and thus could help ensure that clinical trials are more adequately powered to detect the treatment effects they seek to measure.
2022-12-26
medRxiv (prépublication)
doi.org
How programmers find online learning resources
Deeksha M. Arya
Jin L.C. Guo
Martin P. Robillard
2022-12-23
Empirical Software Engineering (publié)
doi.org
FaithDial: A Faithful Benchmark for Information-Seeking Dialogue
Nouha Dziri
Ehsan Kamalloo
Sivan Milton
Osmar Zaiane
Mo Yu
Edoardo M. Ponti
Siva Reddy
The goal of information-seeking dialogue is to respond to seeker queries with natural language utterances that are grounded on knowledge sou… (voir plus)rces. However, dialogue systems often produce unsupported utterances, a phenomenon known as hallucination. To mitigate this behavior, we adopt a data-centric solution and create FaithDial, a new benchmark for hallucination-free dialogues, by editing hallucinated responses in the Wizard of Wikipedia (WoW) benchmark. We observe that FaithDial is more faithful than WoW while also maintaining engaging conversations. We show that FaithDial can serve as training signal for: i) a hallucination critic, which discriminates whether an utterance is faithful or not, and boosts the performance by 12.8 F1 score on the BEGIN benchmark compared to existing datasets for dialogue coherence; ii) high-quality dialogue generation. We benchmark a series of state-of-the-art models and propose an auxiliary contrastive objective that achieves the highest level of faithfulness and abstractiveness based on several automated metrics. Further, we find that the benefits of FaithDial generalize to zero-shot transfer on other datasets, such as CMU-Dog and TopicalChat. Finally, human evaluation reveals that responses generated by models trained on FaithDial are perceived as more interpretable, cooperative, and engaging.
2022-12-22
Transactions of the Association for Computational Linguistics (publié)
doi.org
arxiv.org
Post-hoc Interpretability for Neural NLP: A Survey
Andreas Madsen
Siva Reddy
A. Chandar
2022-12-22
ACM Computing Surveys (publié)
doi.org
arxiv.org
Towards Continual Reinforcement Learning: A Review and Perspectives
Khimya Khetarpal
Matthew D Riemer
Irina Rish
Doina Precup
2022-12-21
Journal of Artificial Intelligence Research (publié)
doi.org
arxiv.org
Lazy vs hasty: linearization in deep networks impacts learning schedule based on example difficulty
Thomas George
Guillaume Lajoie
Aristide Baratin
Among attempts at giving a theoretical account of the success of deep neural networks, a recent line of work has identified a so-called `laz… (voir plus)y' training regime in which the network can be well approximated by its linearization around initialization. Here we investigate the comparative effect of the lazy (linear) and feature learning (non-linear) regimes on subgroups of examples based on their difficulty. Specifically, we show that easier examples are given more weight in feature learning mode, resulting in faster training compared to more difficult ones. In other words, the non-linear dynamics tends to sequentialize the learning of examples of increasing difficulty. We illustrate this phenomenon across different ways to quantify example difficulty, including c-score, label noise, and in the presence of easy-to-learn spurious correlations. Our results reveal a new understanding of how deep networks prioritize resources across example difficulty.
2022-12-19
TMLR (accepté)
doi.org
openreview.net
Meta-topologies define distinct anatomical classes of brain tumours linked to histology and survival
Julius M. Kernbach
Daniel Delev
Georg Neuloh
Hans Clusmann
Danilo Bzdok
Simon B. Eickhoff
Victor E. Staartjes
Flavio Vasella
Michael Weller
Luca Regli
Carlo Serra
Niklaus Krayenbühl
Kevin Akeret
The current World Health Organization classification integrates histological and molecular features of brain tumours. The aim of this study … (voir plus)was to identify generalizable topological patterns with the potential to add an anatomical dimension to the classification of brain tumours. We applied non-negative matrix factorization as an unsupervised pattern discovery strategy to the fine-grained topographic tumour profiles of 936 patients with neuroepithelial tumours and brain metastases. From the anatomical features alone, this machine learning algorithm enabled the extraction of latent topological tumour patterns, termed meta-topologies. The optimal part-based representation was automatically determined in 10 000 split-half iterations. We further characterized each meta-topology’s unique histopathologic profile and survival probability, thus linking important biological and clinical information to the underlying anatomical patterns. In neuroepithelial tumours, six meta-topologies were extracted, each detailing a transpallial pattern with distinct parenchymal and ventricular compositions. We identified one infratentorial, one allopallial, three neopallial (parieto-occipital, frontal, temporal) and one unisegmental meta-topology. Each meta-topology mapped to distinct histopathologic and molecular profiles. The unisegmental meta-topology showed the strongest anatomical–clinical link demonstrating a survival advantage in histologically identical tumours. Brain metastases separated to an infra- and supratentorial meta-topology with anatomical patterns highlighting their affinity to the cortico-subcortical boundary of arterial watershed areas.Using a novel data-driven approach, we identified generalizable topological patterns in both neuroepithelial tumours and brain metastases. Differences in the histopathologic profiles and prognosis of these anatomical tumour classes provide insights into the heterogeneity of tumour biology and might add to personalized clinical decision-making.
2022-12-19
Brain Communications (publié)
doi.org
BLOOM+1: Adding Language Support to BLOOM for Zero-Shot Prompting
Zheng Xin Yong
Hailey Schoelkopf
Niklas Muennighoff
Alham Fikri Aji
David Ifeoluwa Adelani
Khalid Almubarak
M. Saiful Bari
Lintang A. Sutawika
Jungo Kasai
Ahmed Baruwa
Genta Indra Winata
Stella Biderman
Dragomir R. Radev
Vassilina Nikoulina
The BLOOM model is a large publicly available multilingual language model, but its pretraining was limited to 46 languages. To extend the be… (voir plus)nefits of BLOOM to other languages without incurring prohibitively large costs, it is desirable to adapt BLOOM to new languages not seen during pretraining. In this work, we apply existing language adaptation strategies to BLOOM and benchmark its zero-shot prompting performance on eight new languages in a resource-constrained setting. We find language adaptation to be effective at improving zero-shot performance in new languages. Surprisingly, we find that adapter-based finetuning is more effective than continued pretraining for large models. In addition, we discover that prompting performance is not significantly affected by language specifics, such as the writing system. It is primarily determined by the size of the language adaptation data. We also add new languages to BLOOMZ, which is a multitask finetuned version of BLOOM capable of following task instructions zero-shot. We find including a new language in the multitask fine-tuning mixture to be the most effective method to teach BLOOMZ a new language. We conclude that with sufficient training data language adaptation can generalize well to diverse languages. Our code is available at https://github.com/bigscience-workshop/multilingual-modeling.
2022-12-18
ArXiv (prépublication)
doi.org
arxiv.org