Towards Parkinson’s Disease Prognosis Using Self-Supervised Learning and Anomaly Detection

Jiang, Hongchao; Lim, Wei Yang Bryan; Ng, Jer Shyuan; Wang, Yu; Cheng, Ying; Chen, Miao

doi:10.1109/icassp39728.2021.9414840

Cited by 5 publications

(6 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are well-constructed infrastructures for epilepsy seizure (e.g., the TUH EEG Corpus [76]) but very limited public datasets on biomedical time series-based AD or ASD. It is worth mentioning that the PD dataset adopted by [67] is collected by smartphone when participants conduct different activities (e.g., memory, tapping, voice, and walking) [77], which is different from other reported papers that involved neurological disorder diagnosis. In [67], the main indicator is not EEG but human behavior data from accelerometers and gyroscopes that measures acceleration and angular velocity, respectively.…”

Section: Medical Applicationsmentioning

confidence: 94%

“…They can be easily gathered by accelerometers and gyroscopes embedded in numerous devices such as smartphones and smartwatches. Two studies [39,67] adopted both signals as input, and another two works [64,65] only take the acceleration. Acceleration is one of the most popular and most affordable signals in human activity recognition (which may or may not relate to healthcare), we believe there will be more publications on acceleration analysis with contrastive learning.…”

Section: Icu Signalsmentioning

confidence: 99%

“…However, the development of research in this branch is strictly constrained by the availability of data. For example, while Alzheimer's dementia (AD), Parkinson's disease (PD), Autism spectrum disorder (ASD), depression, and epilepsy seizure are all widely-spread neurological disorders, we found five studies on seizure detection [20,31,[40][41][42], one on Parkinson's disease detection [67] but zero on other diseases. The authors have consulted several experts in neuroscience and computer science, and note that the most potential reason for the imbalance across neurological diseases is the data availability.…”

Section: Medical Applicationsmentioning

confidence: 99%

See 2 more Smart Citations

Self-Supervised Contrastive Learning for Medical Time Series: A Systematic Review

Liu

Alavi

Li³

et al. 2023

Sensors

View full text Add to dashboard Cite

Medical time series are sequential data collected over time that measures health-related signals, such as electroencephalography (EEG), electrocardiography (ECG), and intensive care unit (ICU) readings. Analyzing medical time series and identifying the latent patterns and trends that lead to uncovering highly valuable insights for enhancing diagnosis, treatment, risk assessment, and disease progression. However, data mining in medical time series is heavily limited by the sample annotation which is time-consuming and labor-intensive, and expert-depending. To mitigate this challenge, the emerging self-supervised contrastive learning, which has shown great success since 2020, is a promising solution. Contrastive learning aims to learn representative embeddings by contrasting positive and negative samples without the requirement for explicit labels. Here, we conducted a systematic review of how contrastive learning alleviates the label scarcity in medical time series based on PRISMA standards. We searched the studies in five scientific databases (IEEE, ACM, Scopus, Google Scholar, and PubMed) and retrieved 1908 papers based on the inclusion criteria. After applying excluding criteria, and screening at title, abstract, and full text levels, we carefully reviewed 43 papers in this area. Specifically, this paper outlines the pipeline of contrastive learning, including pre-training, fine-tuning, and testing. We provide a comprehensive summary of the various augmentations applied to medical time series data, the architectures of pre-training encoders, the types of fine-tuning classifiers and clusters, and the popular contrastive loss functions. Moreover, we present an overview of the different data types used in medical time series, highlight the medical applications of interest, and provide a comprehensive table of 51 public datasets that have been utilized in this field. In addition, this paper will provide a discussion on the promising future scopes such as providing guidance for effective augmentation design, developing a unified framework for analyzing hierarchical time series, and investigating methods for processing multimodal data. Despite being in its early stages, self-supervised contrastive learning has shown great potential in overcoming the need for expert-created annotations in the research of medical time series.

show abstract

Section: Medical Applicationsmentioning

confidence: 94%

Section: Icu Signalsmentioning

confidence: 99%

Section: Medical Applicationsmentioning

confidence: 99%

See 1 more Smart Citation

Self-Supervised Contrastive Learning for Medical Time Series: A Systematic Review

Liu

Alavi

Li³

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Self-supervised learning algorithms are commonly used in medical research for detecting irregularities in patients' records. They are successfully employed for detecting epileptic seizures [81], pulmonary diseases [82], Parkinson disease [83], and retinal diseases [84]. In addition, they are applied to different modalities of medical data, including Computed Tomography (CT) scans [85], 3D volumetric CT data [41], X-ray scans [86], optical coherence tomography (OCT) [87], Spectral Domain -optical coherence tomography images (SD-OCT) [88], and MRI images [89], [90].…”

Section: Application Domainsmentioning

confidence: 99%

Self-Supervised Anomaly Detection: A Survey and Outlook

Hojjati¹,

Ho²,

Armanfard³

2022

Preprint

View full text Add to dashboard Cite

Over the past few years, anomaly detection, a subfield of machine learning that is mainly concerned with the detection of rare events, witnessed an immense improvement following the unprecedented growth of deep learning models. Recently, the emergence of self-supervised learning has sparked the development of new anomaly detection algorithms that surpassed state-of-the-art accuracy by a significant margin. This paper aims to review the current approaches in self-supervised anomaly detection. We present technical details of the common approaches and discuss their strengths and drawbacks. We also compare the performance of these models against each other and other state-of-the-art anomaly detection models. Finally, we discuss a variety of new directions for improving the existing algorithms. 1

show abstract

“…This could be combined with the whole range of SGs included in the PGS, enriching the information from integrated sources. Apparently, in that case, the adopted machine learning would be extended to deep learning classification schemes, that could create FV in the embedding space for efficient representation of the health status of patients with PD (Jiang et al, 2021). Finally, the transfer of the PGS/iMAT to an immersive (e.g., virtual reality) environment is also foreseen, in an effort to evaluate the level of user engagement and performance under more experiential interaction settings.…”

Section: Limitations and Future Workmentioning

confidence: 99%

Machine Learning-Based Analysis of Digital Movement Assessment and ExerGame Scores for Parkinson's Disease Severity Estimation

et al. 2022

View full text Add to dashboard Cite

Neurodegenerative Parkinson's Disease (PD) is one of the common incurable diseases among the elderly. Clinical assessments are characterized as standardized means for PD diagnosis. However, relying on medical evaluation of a patient's status can be subjective to physicians' experience, making the assessment process susceptible to human errors. The use of ICT-based tools for capturing the status of patients with PD can provide more objective and quantitative metrics. In this vein, the Personalized Serious Game Suite (PGS) and intelligent Motor Assessment Tests (iMAT), produced within the i-PROGNOSIS European project (www.i-prognosis.eu), are explored in the current study. More specifically, data from 27 patients with PD at Stage 1 (9) and Stage 3 (18) produced from their interaction with PGS/iMAT are analyzed. Five feature vector (FV) scenarios are set, including features from PGS or iMAT scores or their combination, after also taking into consideration the age of patients with PD. These FVs are fed into three machine learning classifiers, i.e., K-Nearest Neighbor (KNN), Support Vector Machines (SVM), and Random Forest (RF), to infer the stage of each patient with PD. A Leave-One-Out Cross-Validation (LOOCV) method is adopted for testing the classification performance. The experimental results show that a high (>90%) classification accuracy is achieved from both data sources (PGS/iMAT), justifying the effectiveness of PGS/iMAT to efficiently reflect the motor skill status of patients with PD and further potentiating PGS/iMAT enhancement with a machine learning a part to infer for the stage of patients with PD. Clearly, this integrated approach provides new opportunities for remote monitoring of the stage of patients with PD, contributing to a more efficient organization and set up of personalized interventions.

show abstract

Towards Parkinson’s Disease Prognosis Using Self-Supervised Learning and Anomaly Detection

Cited by 5 publications

References 16 publications

Self-Supervised Contrastive Learning for Medical Time Series: A Systematic Review

Self-Supervised Contrastive Learning for Medical Time Series: A Systematic Review

Self-Supervised Anomaly Detection: A Survey and Outlook

Machine Learning-Based Analysis of Digital Movement Assessment and ExerGame Scores for Parkinson's Disease Severity Estimation

Contact Info

Product

Resources

About