A convolutional oculomotor representation to model parkinsonian fixational patterns from magnified videos

Acosta, Isail Salazar; Pertuz, Said; Contreras, William; Martínez, Fabio

doi:10.1007/s10044-020-00922-4

Cited by 7 publications

(5 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As comparison with standard convolutional nets, we implement the complete ResNet-18 CNN, achieving only an accuracy of 94.26%, with a relative more complex architecture and a total of 11.7M parameters. As baseline comparison, we compared with a machine learning approach that classifies SPD matrices constructed from convolutional responses [17]. Under the same data validation conditions, this approach reports an accuracy of 87.7%, 10% less than the proposed end-to-end ConvSPD 4th-Block model.…”

Section: Evaluation and Resultsmentioning

confidence: 99%

“…These requirements are rarely realistic in clinical scenarios with large pattern variability, and where annotated examples are difficult to get, above all in the task of discovering motor anomalies associated with PD. To avoid such challenging training scheme, some works have used a collection of convolutional responses extracted from the first layers of a Convolutional Neural Network (CNN) architecture previously trained on a general natural image classification problem [17]. Subsequently, different works propose pooling methods to compact these representations in low dimension descriptors using Symmetric Positive Definite (SPD) matrices that summarize feature statistics [17,2].…”

Section: Introductionmentioning

confidence: 99%

“…To avoid such challenging training scheme, some works have used a collection of convolutional responses extracted from the first layers of a Convolutional Neural Network (CNN) architecture previously trained on a general natural image classification problem [17]. Subsequently, different works propose pooling methods to compact these representations in low dimension descriptors using Symmetric Positive Definite (SPD) matrices that summarize feature statistics [17,2]. However, these matrices belong to a Riemannian manifold, making it necessary to design proper methods regarding the geometric structure [11,3].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

An Oculomotor Digital Parkinson Biomarker from a Deep Riemannian Representation

Olmos

Manzanera

Carrillo

2022

Pattern Recognition and Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

Parkinson's disease (PD) is characterized by motor alterations and associated with dopamine neurotransmitters degeneration, affecting 3 % of the population over 65 years of age. Today, there is no definitive biomarker for an early diagnosis and progression characterization. Recently, oculomotor alterations have shown promising evidence to quantify PD patterns. Current capture and oculomotor setups however require sophisticated protocols, limiting the analysis to coarse measures that poorly exploit alterations and restrict their standard use in clinical environments. Computational based deep learning strategies today bring a robust alternative by discovering in video sequences hidden patterns associated to the disease. However, these approaches are dependent on large training data volumes to cover the variability of patterns of interest. This work introduces a novel strategy that exploits data geometry within a deep Riemannian manifold, withstanding data scarcity and discovering oculomotor PD hidden patterns. First, oculomotor information is encoded as symmetric matrices that capture second order statistics of deep features computed by a convolutional scheme. These symmetric matrices then form an embedded representation, which is decoded by a Riemannian network to discriminate Parkinsonian patients w.r.t a control population. The proposed strategy, evaluated on a fixational eye experiment, proves to be a promising approach to represent PD patterns.

show abstract

Section: Evaluation and Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

An Oculomotor Digital Parkinson Biomarker from a Deep Riemannian Representation

Olmos

Manzanera

Carrillo

2022

Pattern Recognition and Artificial Intelligence

Self Cite

View full text Add to dashboard Cite

show abstract

“…These fundamentals have been explored to propose video strategies that recover and learn eye fixation patterns, making possible the representation of disease in weakly controlled scenarios [31]. In this case an approach proposed in previous work [31] achieved an average accuracy of 95%, in a population with 13 control subjects and 13 patients. Despite the remarkable results, this approach is limited by focusing on eye analysis and, losing other signs that may complement disease characterization.…”

Section: Discussionmentioning

confidence: 99%

“…Recently, some spatiotemporal relationships have been captured from the analysis of raw video sequences, computing dense pixelwise motion-based representations that allow to characterize and distinguish some abnormal motion patterns on a particular population of study [28][29][30]. New markerless video schemes have also been introduced to magnify ocular video patterns, emphasizing fixation patterns [31]. These approaches have demonstrated capabilities to discriminate among Parkinson and Control population, but their analysis is dependent on video segmentation, manually performed from the video sequences.…”

Section: Fixational Oculomotor Patternsmentioning

confidence: 99%

A multimodal Parkinson quantification by fusing eye and gait motion patterns, using covariance descriptors, from non-invasive computer vision

Archila¹,

Manzanera

Martínez³

2022

Computer Methods and Programs in Biomedicine

Self Cite

View full text Add to dashboard Cite

Quantification of Parkinsonian unilateral involvement from ocular fixational patterns using a deep video representation

Olmos,

Valenzuela,

Martínez

2023

Health Technol.

View full text Add to dashboard Cite

Purpose Lateralisation of motor symptoms is a prevalent characteristic of Parkinson’s disease (PD). Hence, unilateral involvement is crucial for personalized treatments and measuring therapy effectiveness. Nonetheless, most motor symptoms, including lateralization, are mainly evident at advanced stages of the disease. Recently, ocular fixation instability emerged as a promising PD biomarker with a high sensitivity to discriminate PD. We hypothesize that unilateral involvement can be recovered from the assessment and quantification of PD-related ocular abnormalities. Methods This method proposes a computer-based strategy to quantify PD lateralization from ocular fixation patterns. The method follows a markerless strategy fed by slices with spatiotemporal eye movement information. A deep convolutional model was used to discriminate between PD and a control population. Additionally, model prediction probabilities were analyzed to select the dominant eye associated with unilateral involvement. Results The proposed approach reports an average accuracy of 91.92% classifying PD. Interestingly, using the dominant side, the approach achieves an average PD prediction probability of 93.3% (95% CI: [91.61,95.07]), evidencing capabilities to capture the most affected side. Besides, the reported results strongly correlate with the disease, even for patients categorized at early stages. A low-dimensional projection tool was used to support the classification results by finding a 2d space that eases the discrimination among classes. Conclusions The strategy is sensitive to detecting and classifying PD fixational patterns and determining the side with major impairments. This approach may be a potential tool to support the characterization of the disease and as an alternative to defining personalized treatments.

show abstract

A convolutional oculomotor representation to model parkinsonian fixational patterns from magnified videos

Cited by 7 publications

References 42 publications

An Oculomotor Digital Parkinson Biomarker from a Deep Riemannian Representation

An Oculomotor Digital Parkinson Biomarker from a Deep Riemannian Representation

A multimodal Parkinson quantification by fusing eye and gait motion patterns, using covariance descriptors, from non-invasive computer vision

Quantification of Parkinsonian unilateral involvement from ocular fixational patterns using a deep video representation

Contact Info

Product

Resources

About