Hojjat Adeli’s research while affiliated with The Ohio State University and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (528)


SimCLR (based on Chen et al., [31]) pretext and downstream models for a classification problem similar to the current study
Original EEG traces (green/thick-light line) and EEG traces after various augmentations (black/regular-dark line) based on Mohsenvand et al. [6]
Example of a participant’s 10-minute session 500 Hz records of 16 channels from 7 modalities during a WAUC session with high cognitive workload and moderate physical workload, along with four examples of randomly selected chunks of data with different temporal lengths of 5 sec, 10 sec, 15 sec, and 20 sec (based on Albuquerque et al. [24]).A: ACC, 3-Axis Accelerometry; B: BVP, Blood Volume Pulse; C: ECG, Electrocardiography; E: EEG, Electroencephalography; G: GSR, Galvanic Skin Response; R: RSP, Respiratory Rate; T: TMP, Skin Temperature. A number next to a modality corresponds to a channel number, e.g., E4 corresponds to the 4th\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$4^{th}$$\end{document} EEG channel
The combinatory pattern recognition multi-layer machine learning model
Combination (Comb) accuracies versus selection rate percentage of top 10 combinations corresponding to (a) experiment 1, time-domain investigation (Table 3), (b) experiment 1, frequency-domain investigation (Table 4), (c) experiment 2, time-domain investigation (Table 5), and (d) experiment 2, frequency-domain investigation (Table 6)

+3

Self-Supervised Learning for Near-Wild Cognitive Workload Estimation
  • Article
  • Publisher preview available

November 2024

·

43 Reads

·

6 Citations

Journal of Medical Systems

·

Lynne V. Gauthier

·

Hojjat Adeli

·

Feedback on cognitive workload may reduce decision-making mistakes. Machine learning-based models can produce feedback from physiological data such as electroencephalography (EEG) and electrocardiography (ECG). Supervised machine learning requires large training data sets that are (1) relevant and decontaminated and (2) carefully labeled for accurate approximation, a costly and tedious procedure. Commercial over-the-counter devices are low-cost resolutions for the real-time collection of physiological modalities. However, they produce significant artifacts when employed outside of laboratory settings, compromising machine learning accuracies. Additionally, the physiological modalities that most successfully machine-approximate cognitive workload in everyday settings are unknown. To address these challenges, a first-ever hybrid implementation of feature selection and self-supervised machine learning techniques is introduced. This model is employed on data collected outside controlled laboratory settings to (1) identify relevant physiological modalities to machine approximate six levels of cognitive-physical workloads from a seven-modality repository and (2) postulate limited labeling experiments and machine approximate mental-physical workloads using self-supervised learning techniques.

View access options

OLYMPIAD in ENGINEERING SCIENCE-OES2025

November 2024

·

261 Reads

·

Hojjat Adeli

·

·

[...]

·

NEW DEADLINE for abstract submission: 30 January 2025. --------------- Olympiad in Engineering Science (OES 2025) is an international congress and contest focused on showcasing and evaluating the latest advancements in Engineering Science. It’s a prestigious event where the world’s most influential scientific minds will meet to share knowledge and assess the novelty of the research presented. Key Highlights - Olympiad Medals will be awarded to the top 5 conference papers. - The 2nd Archimedes Medal will be presented for outstanding achievements in Engineering Science. - The inaugural Timoshenko Lifetime Achievement Medal in Mechanics of Solids will be awarded by the President of the Ukrainian Academy of Sciences. - Special Issues: Substantially extended versions of the most innovative papers presented at the conference will be considered for publication in special issues of Computer-Aided Civil and Infrastructure Engineering and Integrated Computer- Aided Engineering. - Publisher of the Conference Proceedings: Springer (Scopus Indexed)



A new epileptic seizure prediction model based on maximal overlap discrete wavelet packet transform, homogeneity index, and machine learning using ECG signals

February 2024

·

62 Reads

·

21 Citations

Biomedical Signal Processing and Control

Epilepsy, a complex pathology with various etiological origins, is characterized by producing hyperexcitability in the brain, which can have multiple disruptive symptoms. It impacts about 40 million people worldwide, of which 20 to 30% have chronic and intractable seizures. Each seizure can create hazardous situations for patients resulting from fractures, burns, submersion accidents, and soft-tissue injuries. Therefore, a method capable of predicting a seizure with sufficient window time before its onset is highly desirable because it will allow the patient to locate a safe place or take appropriate precautionary actions. In this article, a novel method is presented through adroit integration of maximal overlap wavelet packet transform, homogeneity index, and a K-Nearest Neighbors classifier to predict an epileptic event twenty minutes before its onset using electrocardiogram (ECG) signals. The method's effectiveness for predicting an epileptic seizure is verified by employing a database provided by the Massachusetts Institute of Technology-Beth Israel Hospital (MIT-BIH), which includes seven patients with ten epileptic seizures. The results show that the proposed method effectively predicts an epileptic seizure 20 min prior to its onset with an accuracy of 93.25%.


Multiple Classification of Brain MRI Autism Spectrum Disorder by Age and Gender Using Deep Learning

January 2024

·

131 Reads

·

37 Citations

Journal of Medical Systems

The fact that the rapid and definitive diagnosis of autism cannot be made today and that autism cannot be treated provides an impetus to look into novel technological solutions. To contribute to the resolution of this problem through multiple classifications by considering age and gender factors, in this study, two quadruple and one octal classifications were performed using a deep learning (DL) approach. Gender in one of the four classifications and age groups in the other were considered. In the octal classification, classes were created considering gender and age groups. In addition to the diagnosis of ASD (Autism Spectrum Disorders), another goal of this study is to find out the contribution of gender and age factors to the diagnosis of ASD by making multiple classifications based on age and gender for the first time. Brain structural MRI (sMRI) scans of participators with ASD and TD (Typical Development) were pre-processed in the system originally designed for this purpose. Using the Canny Edge Detection (CED) algorithm, the sMRI image data was cropped in the data pre-processing stage, and the data set was enlarged five times with the data augmentation (DA) techniques. The most optimal convolutional neural network (CNN) models were developed using the grid search optimization (GSO) algorism. The proposed DL prediction system was tested with the five-fold cross-validation technique. Three CNN models were designed to be used in the system. The first of these models is the quadruple classification model created by taking gender into account (model 1), the second is the quadruple classification model created by taking into account age (model 2), and the third is the eightfold classification model created by taking into account both gender and age (model 3). ). The accuracy rates obtained for all three designed models are 80.94, 85.42 and 67.94, respectively. These obtained accuracy rates were compared with pre-trained models by using the transfer learning approach. As a result, it was revealed that age and gender factors were effective in the diagnosis of ASD with the system developed for ASD multiple classifications, and higher accuracy rates were achieved compared to pre-trained models.


Multiple Classification of Brain MRI Autism Spectrum Disorder by Age and Gender Using Deep Learning

August 2023

·

80 Reads

·

1 Citation

The fact that the rapid and definitive diagnosis of autism cannot be made today and that autism cannot be treated provides an impetus to look into novel technological solutions. To contribute to the resolution of this problem through multiple classifications by considering age and gender factors, in this study, two quadruple and one octal classifications were performed using a deep learning (DL) approach. Gender in one of the four classifications and age groups in the other were considered. In the octal classification, classes were created considering gender and age groups. In addition to the diagnosis of ASD (Autism Spectrum Disorders), another goal of this study is to find out the contribution of gender and age factors to the diagnosis of ASD by making multiple classifications based on age and gender for the first time. Brain structural MRI (sMRI) scans of participators with ASD and TD (Typical Development) were pre-processed in the system originally designed for this purpose. Using the Canny Edge Detection (CED) algorithm, the sMRI image data was cropped in the data pre-processing stage, and the data set was enlarged five times with the data augmentation (DA) techniques. The most optimal convolutional neural network (CNN) models were developed using the grid search optimization (GSO) algorism. The proposed DL prediction system was tested with the five-fold cross-validation technique. The accuracy rates acquired for all three CNN models designed to be utilized within the system were compared with successfull pre-trained CNN models through the transfer learning (TL) practice. As a result, it was revealed that age and gender factors were effective in the diagnosis of ASD with the system developed for ASD multiple classifications, and it was seen that higher accuracy rates were achieved than pre-trained models.


Vision-based real-time structural vibration measurement through deep-learning-based detection and tracking methods

April 2023

·

174 Reads

·

71 Citations

Engineering Structures

Structural vibration measurement is crucial in structural health monitoring and structural laboratory tests. Traditional contact type sensors are usually required to be attached to the test specimens, which may be difficult to install, and may affect the structural properties and response. Non-contact type wireless sensors are usually expensive and require specialized workers to install and operate. In recent years, vision-based tracking methods for structural vibration measurement have gained increasing interests due to their high accuracy, non-contact feature and low cost. However, traditional vision-based tracking algorithms are susceptible to external environmental conditions such as illumination and background noise. In this paper, two real-time methods, YOLOv3-tiny and YOLOv3-tiny-KLT, are proposed to track structural motions. In the first method, YOLOv3-tiny is established based on the YOLOv3 architecture to localize customized markers where structural displacements are directly determined from the bounding boxes generated. The second method, YOLOv3-tiny-KLT, is a more advanced method which combines the YOLOv3-tiny detector and the traditional KLT tracking algorithm. The pretrained YOLOv3-tiny is deployed to localize the targets automatically, which will then be tracked by Kanade‐Lucas‐Tomasi algorithm. YOLOv3-tiny is intended to provide baseline vibration measurement when the KLT tracking gets lost. The proposed methods were implemented for the videos of shake table tests on a two-storey steel structure. Parametric studies were conducted for the YOLOv3-tiny-KLT method to examine its sensitivity to the tracking parameters. The results show that the proposed method is capable of achieving real-time speed and high accuracy, when compared with the traditional displacement sensors including linear variable differential transducer (LVDT) and String Pots. It is also found that the combined YOLOv3-tiny-KLT approach achieves higher accuracy than YOLOv3-tiny only method, and higher robustness than KLT only method against illumination changes and background noise.


AutoEncoder Filter Bank Common Spatial Patterns to decode Motor Imagery from EEG

February 2023

·

91 Reads

·

37 Citations

IEEE Journal of Biomedical and Health Informatics

The present paper introduces a novel method, named AutoEncoder-Filter Bank Common Spatial Patterns (AE-FBCSP), to decode imagined movements from electroencephalography (EEG). AE-FBCSP is an extension of the well-established FBCSP and is based on a global (cross-subject) and subsequent transfer learning subject-specific (intra-subject) approach. A multi-way extension of AE-FBCSP is also introduced in this paper. Features are extracted from high-density EEG (64 electrodes), by means of FBCSP, and used to train a custom AE, in an unsupervised way, to project the features into a compressed latent space. Latent features are used to train a supervised classifier (feed forward neural network) to decode the imagined movement. The proposed method was tested using a public dataset of EEGs collected from 109 subjects. The dataset consists of right-hand, left-hand, both hands, both feet motor imagery and resting EEGs. AE-FBCSP was extensively tested in the 3-way classification (right hand vs left hand vs resting) and also in the 2-way, 4-way and 5-way ones, both in cross- and intra-subject analysis. AE-FBCSP outperformed standard FBCSP in a statistically significant way (p > 0.05) and achieved a subject-specific average accuracy of 89.09% in the 3-way classification. The proposed methodology performed subject-specific classification better than other comparable methods in the literature, applied to the same dataset, also in the 2-way, 4-way and 5-way tasks. One of the most interesting outcomes is that AE-FBCSP remarkably increased the number of subjects that responded with a very high accuracy, which is a fundamental requirement for BCI systems to be applied in practice.


Figure 1. Overview of SHM steps in offshore and marine structures
Figure 4. Vibration-based SHM based on structural measurements only
Figure 5. Vibration-based SHM based on both structural and environmental measurements
Figure 6. Framework for fatigue assessment and data management of OWTs (adapted from Martinez-Luengo et al. (2019))
Figure 7. Digital twin approach of SHM
State of the art in structural health monitoring of offshore and marine structures

January 2023

·

1,125 Reads

·

82 Citations

Maritime Engineering

The present paper deals with state of the art in Structural Health Monitoring (SHM) methods in offshore and marine structures. Most of the SHM methods have been developed for onshore infrastructures. Few works are available to implement SHM technologies in offshore and marine structures. This paper aims to fill this gap and highlight the challenges in implementing SHM methods in offshore and marine structures. The present work categorizes the available techniques for establishing SHM models in oil rigs, offshore wind turbine structures, subsea systems, vessels, pipelines etc. Besides, the capabilities of the proposed ideas in the recent publications are classified into three main categories: a) the Model-Based, b) the Vibration-Based, and c) the Digital Twin methods. Recently developed novel signal processing and machine learning algorithms have been reviewed, and their abilities have been discussed. Developed methods in Vision-Based and Population-Based approaches have also been presented and discussed. The present paper aims to provide a guideline for selecting and establishing SHM in offshore and marine structures.


Diagnostic of autism spectrum disorder based on structural brain MRI images using, grid search optimization, and convolutional neural networks

January 2023

·

133 Reads

·

70 Citations

Biomedical Signal Processing and Control

In this study, an automatic autism diagnostic model based on sMRI is proposed. This proposed model consists of two basic stages. The first stage is the preprocessing stage, which consists of removing unclear images, identifying the edges of the images by applying the canny edge detection (CED) algorithm, cropping them to the size required by the system, and finally enlarging the images five times with data augmentation. The data augmentation method should not affect the discrimination in the images such as coloring, and also since it is applied to both groups of autism spectrum disorders (ASD) and typical development (TD), it is performed with care not to cause any manipulation in the data. In the second stage, the grid search optimization (GSO) algorithm is applied to the deep convolutional neural networks (DCNN) used in the system to have optimal hyperparameters. As a result, the proposed diagnostic method of ASD based on sMRI achieves an outstanding success rate of 100%. The reliability of the proposed model is validated by testing with five-fold cross-validation, and its superiority is demonstrated by comparing it with recent studies and widely-used pre-trained models.


Citations (81)


... In recent years, tremendous progress has been made in machine learning (ML) and deep learning (DL) in computational mechanics. Numerous robust approaches have been proposed and integrated into broad applications, including pattern recognition (Rafiei et al. 2024;Alam et al. 2020), engineering (Adeli & Yeh 1989), structural designs , modeling (Haghighat et al. 2021), and structural monitoring (Malekloo et al. 2022), which contribute to structural stability, reliability, efficiency, and smart structures. Pereira et al. (2020) proposed a finite element machine classifier for supervised pattern recognition in multivariate data, which adopted basis functions to form probabilistic functions. ...

Reference:

Geometry physics neural operator solver for solid mechanics
Self-Supervised Learning for Near-Wild Cognitive Workload Estimation

Journal of Medical Systems

... Similarly, the gyroscopic moment effect on the dynamic response of FWTs is also of great importance, with several researchers dedicating their efforts to this study [15,16]. In addition, Mostafa et al. [17] and Nematbakhsh et al. [18] have given the calculation formula for gyroscopic moment. ...

Gyroscopic effects of the spinning rotor-blades assembly on dynamic response of offshore wind turbines

Journal of Wind Engineering and Industrial Aerodynamics

... This VMD framework obtained lower computational complexity. Nogay and Adeli (2024) introduced the CNN model for the automatic diagnosis of ASD. In the data pre-processing stage the MRI image data was cropped by the Canny Edge Detection algorithm (CED). ...

Multiple Classification of Brain MRI Autism Spectrum Disorder by Age and Gender Using Deep Learning

Journal of Medical Systems

... Indeed, multifractal spectral features have proven their efficacy in supporting machine learning across various domains, from physiology to perception, action, and cognition. For example, the incorporation of multifractal features into machine-learning algorithms has enabled the discrimination between healthy, interictal, and seizure activities [94], automatic seizure [95][96][97] and intention [98] detection using EEG signals, early detection of diabetic retinopathy using macular images [99,100], Alzheimer's disease diagnosis via functional magnetic resonance imaging (fMRI) images [101][102][103], texture discrimination of hepatocellular carcinoma in histopathological images [104], fatigue assessment using EMG signals [105,106], glioma detection in brain MRI scans [107], and classification of breast cancer from ultrasound images [108]. Thus, ample evidence suggests that multifractal features could enhance the ability of machine-learning models to detect the rich texture of cascade-like nonlinear interactions across scales. ...

A new epileptic seizure prediction model based on maximal overlap discrete wavelet packet transform, homogeneity index, and machine learning using ECG signals
  • Citing Article
  • February 2024

Biomedical Signal Processing and Control

... By utilizing video-image-based sensing methods, structural motion signals can be extracted without physical sensors, allowing for a dense network of contactless sensors across the entire structure [13]. For example, Pan et al. [14] developed a deep learning-based YOLOv3-tiny-KLT algorithm that accurately measures structural motion while mitigating the efects of illumination changes and background noise. Oliveira et al. [15] used the open video platform YouTube to flter and analyze SHM data in response to seismic waves, providing insights into wave propagation and its efects on the built environment. ...

Vision-based real-time structural vibration measurement through deep-learning-based detection and tracking methods
  • Citing Article
  • April 2023

Engineering Structures

... To explore the brain activity associated with motor imagery tasks, we created topographic brain maps using EEG and fNIRS data separately. The EEG data analysis employed the filter band common spatial pattern (FBCSP), a technique frequently used to enhance the detection of motor imagery patterns by filtering EEG data into specific frequency bands and extracting common spatial patterns related to motor imagery [34][35][36]. The resulting topographical maps visually represent the spatial distribution of brain regions involved in motor imagery tasks. ...

AutoEncoder Filter Bank Common Spatial Patterns to decode Motor Imagery from EEG
  • Citing Article
  • February 2023

IEEE Journal of Biomedical and Health Informatics

... Offshore pipelines have a significant role in the global energy infrastructure, allowing the movement of oil and gas from underwater fields to processing plants and markets. However, these pipelines are subject to constant threats to their integrity for a variety of environmental and operational reasons; corrosion has emerged as one of the major issues associated with these pipelines [1], [2]. Corrosion is a phenomenon that transpires naturally, exacerbated by the harsh marine surroundings and highly reactive chemical processes, resulting in the deterioration of materials by inducing thinning of the pipeline wall and compromising structural stability. ...

State of the art in structural health monitoring of offshore and marine structures

Maritime Engineering

... Modeling with Hifoo assumes infinite control knowledge. Hinfinite control is a common approach for vibration control [13][14][15][16]. Hifoo is related to Hinfinity control, which aims to minimize the effects of disturbances and uncertainties in a system. ...

Engineering Applications of Artificial Intelligence New adaptive robust ∞ control of smart structures using synchrosqueezed wavelet transform and recursive least-squares algorithm
  • Citing Article
  • December 2022

Engineering Applications of Artificial Intelligence

... In another study classifying individuals with ASD, researchers achieved a classification accuracy of 76.67% in a sample of IQ-matched typically-developing individuals and 178 ASD individuals. Researchers caution that guided ML analyses of brain imaging data may be limited by small participant numbers [11][12][13]. Acceptable classification precision in ML research was achieved with population samples of fewer than 100 participants, with higher accuracy (above 90%) only attained in studies involving dozens of participants. Classification accuracy suffers significantly when data comes from multiple sites and larger population samples [15,16]. ...

Diagnostic of autism spectrum disorder based on structural brain MRI images using, grid search optimization, and convolutional neural networks
  • Citing Article
  • January 2023

Biomedical Signal Processing and Control

... The area mostly affected by the 2016-2017 earthquakes involves the territories of Abruzzo, Lazio, Umbria and Marche. Reliable earthquake characterization is important to better understand the seismic effects on the territory [90]. Therefore, a summary of the most important features of the main shocks of the Central Italy seismic sequence is given below One month after the first shock in August, numerous groups from different Italian universities, through an agreement of DPC, MiBACT and ReLUIS, carried out an assessment of the usability and damage of churches through the A-DC form. ...

Time-Frequency Signal Analysis of Earthquake Records1
  • Citing Chapter
  • September 2022