Purpose: Parkinson’s Disease (PD) is the second most common form of neural degeneration and defined by the decay of dopaminergic cells in the substantia nigra. The current standard for diagnosing PD occurs once 80% of dopaminergic cells have decayed. The degradation of these cells has been shown to create thinning of the retina walls and retina microvasculature. This work serves to find machine learning techniques to provide PD diagnosis using non-invasive fundus eye images.
Materials and Methods: Two age and gender matched datasets where constructed using data from the UK Biobank (UKB) and data collected at the University of Florida (UF). The first dataset consists of 476 fundus eye images, 238 CN and 238 PD, sourced entirely from the UKB database. The second dataset, UF-UKB, consist of 100 images, 28 CN and 72 PD, collected at UF and 44 CN images from UKB. A second set of datasets, UKB-Green and UF-UKB-Green, were created using the green color channels to improve vessel segmentation. Vessel segmentation was performed using U-Net segmentation network. The vessel maps served as inputs to SVM classifying networks. Saliency maps were created to assess areas of interest for the networks.
Results: The top performing SVM network for the UKB and UKB-Green datasets were the sigmoid SVM networks which achieved accuracies of .698 and .719 respectively. Meanwhile the top performing networks for the UF-UKB and UF-UKB-Green datasets where the linear SVM networks which achieved accuracies of .821 and .857 respectively. The saliency maps indicate that the different networks focused on different vessel structures with the most successful networks focusing more on smaller vessels.
Conclusion: The results indicate that the machine learning networks can classify PD based on retina vasculature, with the key features being smaller blood vessels. The proposed methods further support the idea that changes in brain physiology can be observed in the eye. Machine learning networks can be applied to clinically available data and still provide accurate predictions
Clinical Relevance statement, not to exceed 200 characters: The work illustrates the feasibility of utilizing eye images as a potential method for diagnosing PD, opposed to the current method of using motor symptoms.
more »
« less
Image segmentation for neuroscience: lymphatics
Abstract A recent discovery in neuroscience prompts the need for innovation in image analysis. Neuroscientists have discovered the existence of meningeal lymphatic vessels in the brain and have shown their importance in preventing cognitive decline in mouse models of Alzheimer’s disease. With age, lymphatic vessels narrow and poorly drain cerebrospinal fluid, leading to plaque accumulation, a marker for Alzheimer’s disease. The detection of vessel boundaries and width are performed by hand in current practice and thereby suffer from high error rates and potential observer bias. The existing vessel segmentation methods are dependent on user-defined initialization, which is time-consuming and difficult to achieve in practice due to high amounts of background clutter and noise. This work proposes a level set segmentation method featuring hierarchical matting, LyMPhi, to predetermine foreground and background regions. The level set force field is modulated by the foreground information computed by matting, while also constraining the segmentation contour to be smooth. Segmentation output from this method has a higher overall Dice coefficient and boundary F1-score compared to that of competing algorithms. The algorithms are tested on real and synthetic data generated by our novel shape deformation based approach. LyMPhi is also shown to be more stable under different initial conditions as compared to existing level set segmentation methods. Finally, statistical analysis on manual segmentation is performed to prove the variation and disagreement between three annotators.
more »
« less
- Award ID(s):
- 1759802
- NSF-PAR ID:
- 10333236
- Date Published:
- Journal Name:
- Journal of Physics: Photonics
- Volume:
- 3
- Issue:
- 3
- ISSN:
- 2515-7647
- Page Range / eLocation ID:
- 035004
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract Background There is growing evidence indicating that a number of functional connectivity networks are disrupted at each stage of the full clinical Alzheimer’s disease spectrum. Such differences are also detectable in cognitive normal (CN) carrying mutations of AD risk genes, suggesting a substantial relationship between genetics and AD-altered functional brain networks. However, direct genetic effect on functional connectivity networks has not been measured. Methods Leveraging existing AD functional connectivity studies collected in NeuroSynth, we performed a meta-analysis to identify two sets of brain regions: ones with altered functional connectivity in resting state network and ones without. Then with the brain-wide gene expression data in the Allen Human Brain Atlas, we applied a new biclustering method to identify a set of genes with differential co-expression patterns between these two set of brain regions. Results Differential co-expression analysis using biclustering method led to a subset of 38 genes which showed distinctive co-expression patterns between AD-related and non AD-related brain regions in default mode network. More specifically, we observed 4 sub-clusters with noticeable co-expression difference, where the difference in correlations is above 0.5 on average. Conclusions This work applies a new biclustering method to search for a subset of genes with altered co-expression patterns in AD-related default mode network regions. Compared with traditional differential expression analysis, differential co-expression analysis yielded many more significant hits with extra insights into the wiring mechanism between genes. Particularly, the differential co-expression pattern was observed between two sets of genes, suggesting potential upstream genetic regulators in AD development.more » « less
-
Background: Recent reports have raised concern about the risk of vessel wall injury (VWI) when pulling out current laser-cut stent retrievers with active strut apposition to the vessel walls. Development of braided stroke thrombectomy-assist devices for use in conjunction with aspiration systems may be gentler in the internal carotid (ICA) and basilar vessels (with regards to radial force) compared to existing laser-cut stent retrievers. Methods: Radial force (RF) bench testing was performed using a radial compression station (Blockwise Engineering, Phoenix, AZ). The average total radial force (RF) in Newtons (N) generated (average of 3 readings) in vessel diameters (d) (Range 3.25 to 4.00mm) seen in proximal LVOs of the anterior circulation (such as in the internal carotid artery - ICA), and vessel diameters (d) (Range 2.50 to 3.25mm) seen in the posterior circulation (such as in the basilar artery - BA) was measured. The Solitaire Platinum Revascularization Device (Medtronic, Irvine, CA) was used as the predicate device. All thrombectomy and thrombectomy-assist devices were compared in terms of the RF being higher or lower (%) to the predicate device. Results: The results of the radial force testing are shown in the table below. The total radial force (RF) of the SHELTER® Retriever (part of Insera System, Insera Therapeutics, Inc., Dallas, TX), a braided thrombectomy-assist device is significantly lower (@ d=2.5mm: 58%) than the predicate device (@ d=2.5mm: 100%) and other laser-cut stent retrievers (@ d=2.5mm: 103% to 152%). Thrombectomy devices with lower OD had higher radial forces than larger devices. Conclusion: Novel braided stroke thrombectomy-assist devices for use in conjunction with aspiration systems have lower radial force compared to existing laser-cut stent retrievers in the ICA and BA vessel diameters. Further studies in-vivo need to assess the impact of lower radial force on minimizing VWI. Funding Source: This study was funded in part by a research grant (NSF Award: 1819491; PI: Vallabh Janardhan, MD) from the National Science Foundation (NSF). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Conference Proceeding: This paper was presented in part at the 2018 Annual Meeting of the Society of Vascular & Interventional Neurology (SVIN), November 14-17, 2018 in San Diego, CAmore » « less
-
null (Ed.)Introduction: Alzheimer’s disease (AD) causes progressive irreversible cognitive decline and is the leading cause of dementia. Therefore, a timely diagnosis is imperative to maximize neurological preservation. However, current treatments are either too costly or limited in availability. In this project, we explored using retinal vasculature as a potential biomarker for early AD diagnosis. This project focuses on stage 3 of a three-stage modular machine learning pipeline which consisted of image quality selection, vessel map generation, and classification [1]. The previous model only used support vector machine (SVM) to classify AD labels which limited its accuracy to 82%. In this project, random forest and gradient boosting were added and, along with SVM, combined into an ensemble classifier, raising the classification accuracy to 89%. Materials and Methods: Subjects classified as AD were those who were diagnosed with dementia in “Dementia Outcome: Alzheimer’s disease” from the UK Biobank Electronic Health Records. Five control groups were chosen with a 5:1 ratio of control to AD patients where the control patients had the same age, gender, and eye side image as the AD patient. In total, 122 vessel images from each group (AD and control) were used. The vessel maps were then segmented from fundus images through U-net. A t-test feature selection was first done on the training folds and the selected features was fed into the classifiers with a p-value threshold of 0.01. Next, 20 repetitions of 5-fold cross validation were performed where the hyperparameters were solely tuned on the training data. An ensemble classifier consisting of SVM, gradient boosting tree, and random forests was built and the final prediction was made through majority voting and evaluated on the test set. Results and Discussion: Through ensemble classification, accuracy increased by 4-12% relative to the individual classifiers, precision by 9-15%, sensitivity by 2-9%, specificity by at least 9-16%, and F1 score by 712%. Conclusions: Overall, a relatively high classification accuracy was achieved using machine learning ensemble classification with SVM, random forest, and gradient boosting. Although the results are very promising, a limitation of this study is that the requirement of needing images of sufficient quality decreased the amount of control parameters that can be implemented. However, through retinal vasculature analysis, this project shows machine learning’s high potential to be an efficient, more cost-effective alternative to diagnosing Alzheimer’s disease. Clinical Application: Using machine learning for AD diagnosis through retinal images will make screening available for a broader population by being more accessible and cost-efficient. Mobile device based screening can also be enabled at primary screening in resource-deprived regions. It can provide a pathway for future understanding of the association between biomarkers in the eye and brain.more » « less
-
Introduction: Recent reports have raised concern about the risk of vessel wall injury (VWI) when pulling out current laser-cut stent retrievers during active strut apposition to the vessel walls.1-4 Development of braided thrombectomy-assist devices for use in conjunction with aspiration systems may be gentler (lower radial force) and more optimized for vessel diameters seen in proximal LVOs and distal LVOs. Methods: Bench testing of radial force (RF) was performed using a radial compression station. The total radial force (RF) in Newtons (N) generated in vessel diameters (d) (Range 2.25 to 3mm) seen in proximal LVOs (~M1), and vessel diameters (d) (Range 1.5 to 2.24mm) seen in distal LVOs (~M2) was measured. Radial Force of less than or equal to 1N was grouped as “low” and radial force greater than 1N was grouped as “high” for this analysis. Results: The total radial force (RF) of all laser-cut stent retrievers (with distal outer diameter OD in mm) studied namely Solitaire Platinum (6.0), Solitaire 2 (4.0), Trevo ProVue (4.0), Baby Trevo (3.0), Capture L (3.0) were all higher in the M2 vessels (>1N) compared to M1 vessels (<1N), whereas the total radial force (RF) of the braided thrombectomy-assist devices namely SHELTER® Retriever (6.0) were uniformly low in both the M1 (<1N) and M2 (<1N) vessels. Conclusion: Choosing a stent retriever with lower OD does not translate to lower radial force. As a result, sizing of stent retrievers and thrombectomy-assist devices to target vessels should not only factor the OD of the devices but also the total radial force in the target vessel diameter. Novel braided thrombectomy-assist devices for use in conjunction with aspiration systems have lower radial force compared to existing laser-cut stent retrievers in the M1 and M2 vessel diameters. Further studies in-vivo need to assess the impact of lower radial force on minimizing VWI. Funding Source: This study was funded in part by a research grant (NSF Award: 1819491; PI: Vallabh Janardhan, MD) from the National Science Foundation (NSF). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Conference Proceeding: This paper was presented in part at the 2018 Annual Meeting of the Society of Vascular & Interventional Neurology (SVIN), November 14-17, 2018 in San Diego, CAmore » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1088/2515-7647/ac050e
