Special Issue "Explainable and Augmented Machine Learning for Biosignals and Biomedical Images"
Deadline for manuscript submissions: 31 December 2021.
Interests: information theory; machine learning; deep learning; explainable machine learning; biomedical signal processing; brain computer interface; cybersecurity; computer vision; material informatics
Interests: brain informatics; data analytics; brain–machine interfacing; Internet of Healthcare Things
Special Issues and Collections in MDPI journals
Interests: Neuroinformatics and Neurocomputational Modelling; Artificial Intelligence; Machine Learning; advanced Deep Learning techniques; Spiking Neural Networks and applications in Brain Diseases and Cognitive Impairment (Dementia, MCI, Stroke); Mental Health Information Technology and Spatiotemporal Brain Data Modelling; Personalised Predictive Modelling of static and dynamic data streams; Data Science
Interests: environmental and biomedical instrumentation and measurements even using nanotechnology for devices; advanced signal processing; sensors and sensing systems; machine learning
Special Issues and Collections in MDPI journals
Special Issue in Sensors: Advances in Nanotechnology and Nano-Inspired Computing for Sensors
Special Issue in Sensors: Sensors and Sensor Systems for Hydrodynamics
Special Issue in Sensors: New Technology and Application of Optic Flow Sensors
Special Issue in Sensors: Sensors and Data Analysis Applied in Environmental Monitoring
In recent decades, machine learning (ML) techniques have been providing encouraging breakthroughs in the biomedical research field, reporting outstanding predictive and classification performance.
However, ML algorithms are often perceived as black boxes with no explanation about the final decision process. In this context, explainable machine learning (XML) techniques intend to “open” the black box and provide further insight into the inner working mechanisms underlying artificial intelligence algorithms. Hence, the goal of XML is to explain and interpret outcomes, predictions, decisions, and recommendations automatically achieved by ML models in order to create more comprehensible and transparent machine decisions.
In medical application, such additional understanding, alongside the augmented availability of medical/clinical data acquired from even more interconnected biosensors (based on the Internet of Things (IoT) paradigm) as well as the recent advances in augmented techniques (e.g., generative adversarial network) able to generate synthetic samples, could play a significant role for clinicians, specifically, in the final human decision.
The proposed Special Issue aims to collate innovative explainable ML-based approaches and augmented ML-based methodologies, as well as comprehensive survey papers, applied to problems in medicine and healthcare in order to develop the next generation of systems that can potentially lead to relevant advances in clinical and biomedical research.
Dr. Cosimo Ieracitano
Dr. Mufti Mahmud
Dr. Maryam Doborjeh
Dr. Aime' Lay-Ekuakille
Manuscript Submission Information
Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All papers will be peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.
Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.
Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2200 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.
- Artificial Intelligence
- Pattern recognition
- Explainable machine learning
- Explainable deep learning
- Augmented machine learning
- IoT and biosensors
- Sensing technology for biomedical applications
- Biomedical signal processing
- Biosignals (EEG, ECG, EMG, etc.)
- Imaging technology for biomedical applications
- Biomedical image processing
- Biomedical images (MRI, RX, PET, etc.)
The below list represents only planned manuscripts. Some of these manuscripts have not been received by the Editorial Office yet. Papers submitted to MDPI journals are subject to peer-review.
Title: Explainable transfer learning of brain signals using brain-inspired spiking neural networks
Authors: Nik Kasabov
Affiliation: FIEEE, FRSNZ, FINNS College of Fellows. Professor of Knowledge Engineering, School of Engineering, Comp. and Mathem. Sciences, Faculty DCT. Founding Director of KEDRI, Auckland University of Technology, Auckland 1010.
Title: Quantitative Interpretation of CNN in Depression Identification
Authors: Hengjin Ke1 and Fengqin Wang2 and Fang Hu3 and Xinhua Zhang4
Affiliation: 1 HuBei Polytechnic University and Wuhan University, 2 Hubei Normal University, 3 Huangshi Central Hospital, 4 Peking University People's Hospital
Abstract: Online EEG classification can accurately assess the brain status of patients with Major Depression Disable (MDD) and track their development status in time, which can minimize the risk of falling into danger and suicide. However, it remains a grand research due to (1) the embedded intensive noises and the intrinsic non-stationarity determined by the evolution of brain states, (2) the lack of effective decoupling of the complex relationship between brain region and neural network during the attack of brain diseases. This study design a CNN to classify the EEG data and then provide the quantitative interpretation why the classifier is suitable.
Title: Machine Learning Methods for Fear Level Classification based on Physiological Features
Authors: Livia Petrescu1, Catalin Petrescu2, Ana Oprea2, Oana Mitrut2, Gabriela Moise3, Alin Moldoveanu2, Florica Moldoveanu2
Affiliation: 1- Faculty of Biology, University of Bucharest, 050095 Bucharest, Romania. 2 - Faculty of Automatic Control and Computers, University Politehnica of Bucharest, 060042 Bucharest, Romania. 3 - Faculty of Letters and Sciences, Petroleum-Gas University of Ploiesti, 100680 Ploiesti, Romania
Abstract: Emotion classification is an emerging multi-disciplinary field which has drawn the attention of researchers due to its applicability in healthcare, neuroscience, psychology and human-computer interaction. The development of physiological monitoring devices, signal processing algorithms and artificial intelligence methods has recently enabled the expansion of automatic emotion recognition systems. This paper focuses on the binary classification of the emotion of fear, based on the physiological data and subjective responses stored in the DEAP dataset. We have performed a mapping between the discrete and dimensional emotional information considering the participants’ ratings and extracted a substantial set of features from the physiological data, which represented the input to various machine learning algorithms, accompanied by dimensionality reduction, feature selection and the tuning of the most relevant hyperparameters boosting classification accuracy. The methodology we approached included tackling different situations such as resolving the problem of having an imbalanced dataset, reducing overfitting and computing various metrics in order to obtain the most reliable classification scores. The results showed that the level of fear can be predicted very accurately by extracting the most relevant features from the physiological data and by searching for the best parameters which maximize the machine learning algorithms’ classification scores.