Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

General General

Better Together: Using Multi-task Learning to Improve Feature Selection within Structural Datasets

ArXiv Preprint

There have been recent efforts to move to population-based structural health monitoring (PBSHM) systems. One area of PBSHM which has been recognised for potential development is the use of multi-task learning (MTL); algorithms which differ from traditional independent learning algorithms. Presented here is the use of the MTL, ''Joint Feature Selection with LASSO'', to provide automatic feature selection for a structural dataset. The classification task is to differentiate between the port and starboard side of a tailplane, for samples from two aircraft of the same model. The independent learner produced perfect F1 scores but had poor engineering insight; whereas the MTL results were interpretable, highlighting structural differences as opposed to differences in experimental set-up.

S. C. Bee, E. Papatheou, M Haywood-Alexander, R. S. Mills, L. A. Bull, K. Worden, N. Dervilis

2023-03-08

Cardiology Cardiology

Deep-Learning for Epicardial Adipose Tissue Assessment With Computed Tomography: Implications for Cardiovascular Risk Prediction.

In JACC. Cardiovascular imaging

BACKGROUND : Epicardial adipose tissue (EAT) volume is a marker of visceral obesity that can be measured in coronary computed tomography angiograms (CCTA). The clinical value of integrating this measurement in routine CCTA interpretation has not been documented.

OBJECTIVES : This study sought to develop a deep-learning network for automated quantification of EAT volume from CCTA, test it in patients who are technically challenging, and validate its prognostic value in routine clinical care.

METHODS : The deep-learning network was trained and validated to autosegment EAT volume in 3,720 CCTA scans from the ORFAN (Oxford Risk Factors and Noninvasive Imaging Study) cohort. The model was tested in patients with challenging anatomy and scan artifacts and applied to a longitudinal cohort of 253 patients post-cardiac surgery and 1,558 patients from the SCOT-HEART (Scottish Computed Tomography of the Heart) Trial, to investigate its prognostic value.

RESULTS : External validation of the deep-learning network yielded a concordance correlation coefficient of 0.970 for machine vs human. EAT volume was associated with coronary artery disease (odds ratio [OR] per SD increase in EAT volume: 1.13 [95% CI: 1.04-1.30]; P = 0.01), and atrial fibrillation (OR: 1.25 [95% CI:1.08-1.40]; P = 0.03), after correction for risk factors (including body mass index). EAT volume predicted all-cause mortality (HR per SD: 1.28 [95% CI: 1.10-1.37]; P = 0.02), myocardial infarction (HR: 1.26 [95% CI:1.09-1.38]; P = 0.001), and stroke (HR: 1.20 [95% CI: 1.09-1.38]; P = 0.02) independently of risk factors in SCOT-HEART (5-year follow-up). It also predicted in-hospital (HR: 2.67 [95% CI: 1.26-3.73]; P ≤ 0.01) and long-term post-cardiac surgery atrial fibrillation (7-year follow-up; HR: 2.14 [95% CI: 1.19-2.97]; P ≤ 0.01).

CONCLUSIONS : Automated assessment of EAT volume is possible in CCTA, including in patients who are technically challenging; it forms a powerful marker of metabolically unhealthy visceral obesity, which could be used for cardiovascular risk stratification.

West Henry W, Siddique Muhammad, Williams Michelle C, Volpe Lucrezia, Desai Ria, Lyasheva Maria, Thomas Sheena, Dangas Katerina, Kotanidis Christos P, Tomlins Pete, Mahon Ciara, Kardos Attila, Adlam David, Graby John, Rodrigues Jonathan C L, Shirodaria Cheerag, Deanfield John, Mehta Nehal N, Neubauer Stefan, Channon Keith M, Desai Milind Y, Nicol Edward D, Newby David E, Antoniades Charalambos

2023-Jan-10

adipose tissue, atherosclerosis, atrial fibrillation, computed tomography, deep-learning, visceral fat

General General

Machine learning-aided scoring of synthesis difficulties for designer chromosomes.

In Science China. Life sciences

Designer chromosomes are artificially synthesized chromosomes. Nowadays, these chromosomes have numerous applications ranging from medical research to the development of biofuels. However, some chromosome fragments can interfere with the chemical synthesis of designer chromosomes and eventually limit the widespread use of this technology. To address this issue, this study aimed to develop an interpretable machine learning framework to predict and quantify the synthesis difficulties of designer chromosomes in advance. Through the use of this framework, six key sequence features leading to synthesis difficulties were identified, and an eXtreme Gradient Boosting model was established to integrate these features. The predictive model achieved high-quality performance with an AUC of 0.895 in cross-validation and an AUC of 0.885 on an independent test set. Based on these results, the synthesis difficulty index (S-index) was proposed as a means of scoring and interpreting synthesis difficulties of chromosomes from prokaryotes to eukaryotes. The findings of this study emphasize the significant variability in synthesis difficulties between chromosomes and demonstrate the potential of the proposed model to predict and mitigate these difficulties through the optimization of the synthesis process and genome rewriting.

Zheng Yan, Song Kai, Xie Ze-Xiong, Han Ming-Zhe, Guo Fei, Yuan Ying-Jin

2023-Mar-03

artificial chromosome, chemical synthesis, machine learning, synthetic biology

General General

Defining Selective Neuronal Resilience and Identifying Targets of Neuroprotection and Axon Regeneration Using Single-Cell RNA Sequencing: Computational Approaches.

In Methods in molecular biology (Clifton, N.J.)

We describe a computational workflow to analyze single-cell RNA-sequencing (scRNA-seq) profiles of axotomized retinal ganglion cells (RGCs) in mice. Our goal is to identify differences in the dynamics of survival among 46 molecularly defined RGC types together with molecular signatures that correlate with these differences. The data consists of scRNA-seq profiles of RGCs collected at six time points following optic nerve crush (ONC) (see companion chapter by Jacobi and Tran). We use a supervised classification-based approach to map injured RGCs to type identities and quantify type-specific differences in survival at 2 weeks post crush. As injury-related changes in gene expression confound the inference of type identity in surviving cells, the approach deconvolves type-specific gene signatures from injury responses by using an iterative strategy that leverages measurements along the time course. We use these classifications to compare expression differences between resilient and susceptible subpopulations, identifying potential mediators of resilience. The conceptual framework underlying the method is sufficiently general for analysis of selective vulnerability in other neuronal systems.

Butrus Salwan, Sagireddy Srikant, Yan Wenjun, Shekhar Karthik

2023

Machine learning, Optic nerve crush, Retinal ganglion cells, Single-cell RNA-sequencing, Supervised classification

General General

Using serial dependence to predict confidence across observers and cognitive domains.

In Psychonomic bulletin & review

Our perceptual system appears hardwired to exploit regularities of input features across space and time in seemingly stable environments. This can lead to serial dependence effects whereby recent perceptual representations bias current perception. Serial dependence has also been demonstrated for more abstract representations, such as perceptual confidence. Here, we ask whether temporal patterns in the generation of confidence judgments across trials generalize across observers and different cognitive domains. Data from the Confidence Database across perceptual, memory, and cognitive paradigms was reanalyzed. Machine learning classifiers were used to predict the confidence on the current trial based on the history of confidence judgments on the previous trials. Cross-observer and cross-domain decoding results showed that a model trained to predict confidence in the perceptual domain generalized across observers to predict confidence across the different cognitive domains. The recent history of confidence was the most critical factor. The history of accuracy or Type 1 reaction time alone, or in combination with confidence, did not improve the prediction of the current confidence. We also observed that confidence predictions generalized across correct and incorrect trials, indicating that serial dependence effects in confidence generation are uncoupled to metacognition (i.e., how we evaluate the precision of our own behavior). We discuss the ramifications of these findings for the ongoing debate on domain-generality versus domain-specificity of metacognition.

Mei Ning, Rahnev Dobromir, Soto David

2023-Mar-07

Cognition, Confidence, Machine learning, Memory, Metacognition, Perception

General General

Extreme Learning Machine (ELM) Method for Classification of Preschool Children Brain Imaging.

In Journal of autism and developmental disorders ; h5-index 76.0

Brain tumors are formed due to the abnormal growth of the cells that get multiplied and become in uncontrollable perspective. Tumors can damage brain cells by pressing down the skull, which consociate begins in, which negatively affects human health. In advanced stages, a brain tumor is a more dangerous infection that cannot be relieved. Brain tumor detection and early prevention are necessary in today's world. An extreme learning machine (ELM) is a widely adopted algorithm in machine learning. It is proposed to use classification models in brain tumor imaging. This classification is based on the techniques implemented: Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GAN). CNN efficiently solves the convex optimization problem and is faster, requiring less human effort. The algorithmic architecture of GAN uses two neural networks, pocking one against the other. These networks are implemented in various fields to classify brain Tumor images. The present study mainly aims to introduce a new proposed classification system of preschool children brain imaging with Hybrid Convolutional Neural Networks and with the techniques of GAN. The proposed technique is compared with the existing hybrid-CNN and hybrid-GAN techniques. The outcomes are encouraging because the loss is deduced, and the accuracy facet increases. The proposed system achieved a training accuracy of 97.8% and a validation accuracy of 89%. The outcomes of the studies show that ELM in the platform of GAN for preschool children brain imaging classification has achieved higher predictive performance than the traditional classification mechanisms in increasingly complex scenarios. Time elapsed for training brain images samples finds inference value for training samples and time elapsed value increased by 28.9855%. Approximation ratio for cost based on probability, finding Approximation ratio for low probability range is increased by 88.1%. The combination of CNN, GAN and hybrid-CNN, hybrid-GAN, and hybrid CNN + GAN, compared with the proposed hybrid system, increased Detection latency for low range learning rate by 3.31%.

Li Deming, De Li, Keqing Li, Gjoni Gazmir

2023-Mar-07

Deep learning, Image classification, Neural networks, Tumor detection