1
|
Tasaki S, Kim N, Truty T, Zhang A, Buchman AS, Lamar M, Bennett DA. Explainable deep learning approach for extracting cognitive features from hand-drawn images of intersecting pentagons. NPJ Digit Med 2023; 6:157. [PMID: 37612472 PMCID: PMC10447434 DOI: 10.1038/s41746-023-00904-w] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/20/2023] [Accepted: 08/10/2023] [Indexed: 08/25/2023] Open
Abstract
Hand drawing, which requires multiple neural systems for planning and controlling sequential movements, is a useful cognitive test for older adults. However, the conventional visual assessment of these drawings only captures limited attributes and overlooks subtle details that could help track cognitive states. Here, we utilized a deep-learning model, PentaMind, to examine cognition-related features from hand-drawn images of intersecting pentagons. PentaMind, trained on 13,777 images from 3111 participants in three aging cohorts, explained 23.3% of the variance in the global cognitive scores, 1.92 times more than the conventional rating. This accuracy improvement was due to capturing additional drawing features associated with motor impairments and cerebrovascular pathologies. By systematically modifying the input images, we discovered several important drawing attributes for cognition, including line waviness. Our results demonstrate that deep learning models can extract novel drawing metrics to improve the assessment and monitoring of cognitive decline and dementia in older adults.
Collapse
Affiliation(s)
- Shinya Tasaki
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA.
- Department of Neurological Sciences, Rush University Medical Center, Chicago, IL, USA.
| | - Namhee Kim
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
- Department of Neurological Sciences, Rush University Medical Center, Chicago, IL, USA
| | - Tim Truty
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
| | - Ada Zhang
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
| | - Aron S Buchman
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
- Department of Neurological Sciences, Rush University Medical Center, Chicago, IL, USA
| | - Melissa Lamar
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
- Department of Psychiatry and Behavioral Sciences, Rush University Medical Center, Chicago, IL, USA
| | - David A Bennett
- Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
- Department of Neurological Sciences, Rush University Medical Center, Chicago, IL, USA
| |
Collapse
|
2
|
Tasaki S, Kim N, Truty T, Zhang A, Buchman AS, Lamar M, Bennett DA. Interpretable deep learning approach for extracting cognitive features from hand-drawn images of intersecting pentagons in older adults. BIORXIV : THE PREPRINT SERVER FOR BIOLOGY 2023:2023.04.18.537358. [PMID: 37131841 PMCID: PMC10153174 DOI: 10.1101/2023.04.18.537358] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/04/2023]
Abstract
Hand drawing involves multiple neural systems for planning and precise control of sequential movements, making it a valuable cognitive test for older adults. However, conventional visual assessment of drawings may not capture intricate nuances that could help track cognitive states. To address this issue, we utilized a deep-learning model, PentaMind, to examine cognition-related features from hand-drawn images of intersecting pentagons. PentaMind, trained on 13,777 images from 3,111 participants in three aging cohorts, explained 23.3% of the variance in global cognitive scores, a comprehensive hour-long cognitive battery. The model’s performance, which was 1.92 times more accurate than conventional visual assessment, significantly improved the detection of cognitive decline. The improvement in accuracy was due to capturing additional drawing features that we found to be associated with motor impairments and cerebrovascular pathologies. By systematically modifying the input images, we discovered several important drawing attributes for cognition, including line waviness. Our results demonstrate that hand-drawn images can provide rich cognitive information, enabling rapid assessment of cognitive decline and suggesting potential clinical implications in dementia.
Collapse
|
3
|
DeBroff J, Omer N, Cohen B, Giladi N, Kestenbaum M, Shirvan JC, Cedarbaum JM, Gana‐Weisz M, Goldstein O, Orr‐Urtreger A, Mirelman A, Thaler A. The Influence of GBA and LRRK2 on Mood Disorders in Parkinson's Disease. Mov Disord Clin Pract 2023; 10:606-616. [PMID: 37070047 PMCID: PMC10105114 DOI: 10.1002/mdc3.13722] [Citation(s) in RCA: 4] [Impact Index Per Article: 4.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/20/2022] [Revised: 02/14/2023] [Accepted: 03/03/2023] [Indexed: 03/14/2023] Open
Abstract
Background Mood disorders have emerged as major non-motor comorbidities in Parkinson's disease (PD) even at the prodromal stage of the disease. Mutations in the LRRK2 and GBA genes are common among Ashkenazi Jews, with more severe phenotype reported for GBA-PD. Objective To explore the association between genetic status and mood related disorders before and after diagnosis of PD and the association between mood-related medications, phenotype, and genetic status. Methods Participants were genotyped for mutations in the LRRK2 and GBA genes. State of depression, anxiety and non-motor features were evaluated using validated questionnaires. History of mood disorders prior to diagnosis of PD and use of mood-related medications were assessed. Results The study included 105 idiopathic PD (iPD), 55 LRRK2-PD and 94 GBA-PD. Scores on mood related questionnaires and frequency of depression and anxiety before diagnosis were similar between the groups (p>0.05). However, more GBA-PD patients used mood related medications before PD diagnosis than LRRK2-PD and iPD (16.5% vs 7.1% and 8.2%, p=0.044). LRRK2-PD and GBA-PD receiving mood-related medications at time of assessment had worse motor and non-motor phenotype compared to those that did not (p<0.05). LRRK2-PD receiving mood related-medications at time of assessment, scored higher on mood-related questionnaires compared to LRRK2-PD not receiving such medications (p<0.04). Conclusions Prodromal GBA-PD are more frequently treated with mood related-medications despite equal rates of reported mood-related disorders, while LRRK2-PD with mood-related disorders experience high rates of anxiety and depression despite treatment, attesting to the need of more precise assessment and treatment of these genetic subgroups.
Collapse
Affiliation(s)
| | - Nurit Omer
- Sackler School of MedicineTel‐Aviv University
- Movement Disorders UnitNeurological Institute, Tel‐Aviv Medical Center
- Laboratory of Early Markers of NeurodegenerationNeurological Institute, Tel‐Aviv Medical Center
| | - Batsheva Cohen
- Laboratory of Early Markers of NeurodegenerationNeurological Institute, Tel‐Aviv Medical Center
| | - Nir Giladi
- Sackler School of MedicineTel‐Aviv University
- Movement Disorders UnitNeurological Institute, Tel‐Aviv Medical Center
- Sagol School of NeuroscienceTel‐Aviv University
| | - Meir Kestenbaum
- Sackler School of MedicineTel‐Aviv University
- Neurology departmentMeir HospitalKfar‐SabaIsrael
| | | | | | - Mali Gana‐Weisz
- Genomic Research Laboratory for NeurodegenerationTel‐Aviv Medical CenterTel‐AvivIsrael
| | - Orly Goldstein
- Genomic Research Laboratory for NeurodegenerationTel‐Aviv Medical CenterTel‐AvivIsrael
| | - Avi Orr‐Urtreger
- Sackler School of MedicineTel‐Aviv University
- Sagol School of NeuroscienceTel‐Aviv University
- Genomic Research Laboratory for NeurodegenerationTel‐Aviv Medical CenterTel‐AvivIsrael
| | - Anat Mirelman
- Sackler School of MedicineTel‐Aviv University
- Laboratory of Early Markers of NeurodegenerationNeurological Institute, Tel‐Aviv Medical Center
- Sagol School of NeuroscienceTel‐Aviv University
| | - Avner Thaler
- Sackler School of MedicineTel‐Aviv University
- Movement Disorders UnitNeurological Institute, Tel‐Aviv Medical Center
- Laboratory of Early Markers of NeurodegenerationNeurological Institute, Tel‐Aviv Medical Center
- Sagol School of NeuroscienceTel‐Aviv University
| |
Collapse
|
4
|
Matusz EF, Price CC, Lamar M, Swenson R, Au R, Emrani S, Wasserman V, Libon DJ, Thompson LI. Dissociating Statistically Determined Normal Cognitive Abilities and Mild Cognitive Impairment Subtypes with DCTclock. J Int Neuropsychol Soc 2023; 29:148-158. [PMID: 35188095 PMCID: PMC11194727 DOI: 10.1017/s1355617722000091] [Citation(s) in RCA: 5] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 01/26/2023]
Abstract
OBJECTIVE To determine whether the DCTclock can detect differences across groups of patients seen in the memory clinic for suspected dementia. METHOD Patients (n = 123) were classified into the following groups: cognitively normal (CN), subtle cognitive impairment (SbCI), amnestic cognitive impairment (aMCI), and mixed/dysexecutive cognitive impairment (mx/dysMCI). Nine outcome variables included a combined command/copy total score and four command and four copy indices measuring drawing efficiency, simple/complex motor operations, information processing speed, and spatial reasoning. RESULTS Total combined command/copy score distinguished between groups in all comparisons with medium to large effects. The mx/dysMCI group had the lowest total combined command/copy scores out of all groups. The mx/dysMCI group scored lower than the CN group on all command indices (p < .050, all analyses); and lower than the SbCI group on drawing efficiency (p = .011). The aMCI group scored lower than the CN group on spatial reasoning (p = .019). Smaller effect sizes were obtained for the four copy indices. CONCLUSIONS These results suggest that DCTclock command/copy parameters can dissociate CN, SbCI, and MCI subtypes. The larger effect sizes for command clock indices suggest these metrics are sensitive in detecting early cognitive decline. Additional research with a larger sample is warranted.
Collapse
Affiliation(s)
- Emily F. Matusz
- Department of Geriatrics and Gerontology, New Jersey Institute for Successful Aging, School of Osteopathic Medicine, Rowan University, Stratford, NJ, USA
| | - Catherine C. Price
- Clinical and Health Psychology, University of Florida, Gainesville, FL, USA
| | - Melissa Lamar
- Department of Behavioral Sciences and the Rush Alzheimer’s Disease Center, Rush University Medical Center, Chicago, IL, USA
| | - Rod Swenson
- University of North Dakota School of Medicine and Health Sciences, Grand Forks, ND, USA
| | - Rhoda Au
- Boston University Schools of Medicine & Public Health, Boston, MA, USA
| | - Sheina Emrani
- Department of Psychology, Rowan University, Stratford, NJ, USA
| | | | - David J. Libon
- Department of Geriatrics and Gerontology, New Jersey Institute for Successful Aging, School of Osteopathic Medicine, Rowan University, Stratford, NJ, USA
- Department of Psychology, Rowan University, Stratford, NJ, USA
| | - Louisa I. Thompson
- Department of Psychiatry and Human Behavior, Alpert Medical School of Brown University, Providence, RI, USA
- Butler Hospital Memory & Aging Program, Providence, RI, USA
| |
Collapse
|
5
|
Schejter-Margalit T, Kizony R, Ben-Binyamin N, Hacham R, Thaler A, Maidan I, Mirelman A. Neural activation in the prefrontal cortex during the digital clock drawing test measured with functional near-infrared spectroscopy in early stage Parkinson's disease. Parkinsonism Relat Disord 2022; 105:9-14. [PMID: 36327601 DOI: 10.1016/j.parkreldis.2022.10.021] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 08/12/2022] [Revised: 10/16/2022] [Accepted: 10/19/2022] [Indexed: 11/30/2022]
Abstract
INTRODUCTION The clock drawing test (CDT) is a neuropsychological test for the screening of global cognitive functioning. The test requires use of multiple cognitive domains including executive functions, visuospatial abilities and semantic memory and can be a suitable tool for screening cognitive decline in participants in the early stages of Parkinson's Disease (PD). Behavioral performance on the CDT has been studied in depth, however, neural activation during real-time performance has not been extensively investigated. In this study we explored changes in prefrontal cortex (PFC) activation during the performance of CDT in participants with PD compared to healthy controls (HC) and assessed the correlations between PFC activation and CDT performance. METHODS The study included 60 participants, 29 PD and 31 HC participants whom performed a digital CDT (DCTclock) in conjunction with a Functional Near-Infrared Spectroscopy (fNIRS) system measuring neural activation in the PFC. RESULTS HbO2 signals derived from the fNIRS during the CDT revealed that PD participants showed more moderate slopes than the HC in the right hemisphere in the command (p = 0.042) and copy task (p = 0.009). Better score on the measurement of information processing correlated with steeper right hemisphere HbO2 slope in the copy task in the PD group (p = 0.003). CONCLUSION Our results reflect slower PFC activation in participants with PD which correlates with behavioral measures. In addition, the findings of the study indicate the importance of performing the CDT copy task condition that detect early cognitive decline in participants with PD.
Collapse
Affiliation(s)
- Tamara Schejter-Margalit
- Laboratory for Early Markers of Neurodegeneration, Center for the Study of Movement, Cognition and Mobility, Tel Aviv Medical Center, Tel Aviv, Israel; Occupational Therapy Department, University of Haifa, Haifa, Israel.
| | - Rachel Kizony
- Occupational Therapy Department, University of Haifa, Haifa, Israel; Occupational Therapy, Sheba Medical Center, Tel Hashomer, Israel
| | | | - Roni Hacham
- Laboratory for Early Markers of Neurodegeneration, Center for the Study of Movement, Cognition and Mobility, Tel Aviv Medical Center, Tel Aviv, Israel
| | - Avner Thaler
- Laboratory for Early Markers of Neurodegeneration, Center for the Study of Movement, Cognition and Mobility, Tel Aviv Medical Center, Tel Aviv, Israel; Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Israel
| | - Inbal Maidan
- Laboratory for Early Markers of Neurodegeneration, Center for the Study of Movement, Cognition and Mobility, Tel Aviv Medical Center, Tel Aviv, Israel; Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Israel
| | - Anat Mirelman
- Laboratory for Early Markers of Neurodegeneration, Center for the Study of Movement, Cognition and Mobility, Tel Aviv Medical Center, Tel Aviv, Israel; Sackler Faculty of Medicine and Sagol School of Neuroscience, Tel Aviv University, Israel
| |
Collapse
|
6
|
Libon DJ, Swenson R, Lamar M, Price CC, Baliga G, Pascual-Leone A, Au R, Cosentino S, Andersen SL. The Boston Process Approach and Digital Neuropsychological Assessment: Past Research and Future Directions. J Alzheimers Dis 2022; 87:1419-1432. [PMID: 35466941 DOI: 10.3233/jad-220096] [Citation(s) in RCA: 10] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/15/2022]
Abstract
Neuropsychological assessment using the Boston Process Approach (BPA) suggests that an analysis of the strategy or the process by which tasks and neuropsychological tests are completed, and the errors made during test completion convey much information regarding underlying brain and cognition and are as important as overall summary scores. Research over the last several decades employing an analysis of process and errors has been able to dissociate between dementia patients diagnosed with Alzheimer's disease, vascular dementia associated with MRI-determined white matter alterations, and Parkinson's disease; and between mild cognitive impairment subtypes. Nonetheless, BPA methods can be labor intensive to deploy. However, the recent availability of digital platforms for neuropsychological test administration and scoring now enables reliable, rapid, and objective data collection. Further, digital technology can quantify highly nuanced data previously unobtainable to define neurocognitive constructs with high accuracy. In this paper, a brief review of the BPA is provided. Studies that demonstrate how digital technology translates BPA into specific neurocognitive constructs using the Clock Drawing Test, Backward Digit Span Test, and a Digital Pointing Span Test are described. Implications for using data driven artificial intelligence-supported analytic approaches enabling the creation of more sensitive and specific detection/diagnostic algorithms for putative neurodegenerative illness are also discussed.
Collapse
Affiliation(s)
- David J Libon
- New Jersey Institute for Successful Aging, Rowan University, School of Osteopathic Medicine, NJ, USA
| | - Rod Swenson
- Department of Psychiatry and Behavioral Science, University of North Dakota School of Medicine and Health Sciences, Grand Forks, ND, USA
| | - Melissa Lamar
- Rush Alzheimer's Disease Center and the Department of Psychiatry and Behavioral Sciences, Rush University Medical Center, Chicago, IL, USA
| | - Catherine C Price
- Department of Clinical and Health Psychology, University of Florida, Gainesville, FL, USA
| | - Ganesh Baliga
- Department of Computer Science, Rowan University, Glassboro, NJ, USA
| | - Alvaro Pascual-Leone
- Hinda and Arthur Marcus Institute for Aging Research and Deanna and Sidney Wolk Center for Memory Health, Hebrew Senior Life, Boston, MA, USA.,Department of Neurology, Harvard Medical School, Boston, MA, USA.,Guttmann Brain Health Institute, Barcelona, Spain
| | - Rhoda Au
- Departments of Anatomy & Neurobiology and Neurology; Framingham Heart Study, Slone Epidemiology Center and Alzheimer's Disease Research Center, Boston University School of Medicine, Boston, MA, USA.,Department of Epidemiology, Boston University School of Public Health, Boston, MA, USA
| | - Stephanie Cosentino
- Department of Neurology, Taub Institute and Sergievsky Center, Cognitive Neuroscience Division, Columbia University Medical Center, New York, NY, USA
| | - Stacy L Andersen
- Department of Medicine, Boston University School of Medicine, Boston, MA, USA
| |
Collapse
|
7
|
Tao X, Zhou H, Mo D, Zhang W, Chang Z, Zeng Y, Luo Y, Wu S, Tang W, Yang C, Wang Q. Erythrocytes Are an Independent Protective Factor for Vascular Cognitive Impairment in Patients With Severe White Matter Hyperintensities. Front Aging Neurosci 2022; 14:789602. [PMID: 35250538 PMCID: PMC8894857 DOI: 10.3389/fnagi.2022.789602] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/05/2021] [Accepted: 01/27/2022] [Indexed: 11/13/2022] Open
Abstract
Background and Purpose: Hemoglobin is one of the main proteins in erythrocytes. There are significant correlations between low hemoglobin and white matter hyperintensities (WMH) and cognitive impairment. This study explored whether erythrocytopenia has predictive value for vascular cognitive impairment (VCI) in patients with WMH. Method: We conducted a cross-sectional study of 302 patients, including 62 with cerebral small vessel disease and 240 with stroke. Basic demographic data and fasting blood were collected. First, all patients were divided into normal cognition (NC), mild VCI (mVCI), and severe VCI (sVCI) groups (subgroups later) based on cognitive behavior scores. Second, all patients were divided into mild WMH (mWMH) and severe WMH (sWMH) groups based on Fazekas scores. The differences in blood markers between different groups or subgroups with different cognitive levels were analyzed by univariate analysis. Then, binary logistic regression was used to analyze the diagnostic value of erythrocyte counts for VCI in the sWMH group, and ordinal logistic regression was used to analyze the predictive value of multiple variables for different cognitive levels. Results: Univariate analysis showed that erythrocytes, hemoglobin, high-sensitivity C-reactive protein, retinol binding protein and prealbumin were potential blood markers for different cognitive levels in sWMH patients. Among them, erythrocytopenia has good predictive value for the diagnosis of mVCI (AUC = 0.685, P = 0.008) or sVCI (AUC = 0.699, P = 0.003) in patients with sWMH. Multivariate joint analysis showed that erythrocytes were an independent protective factor reducing the occurrence of VCI in patients with sWMH (OR = 0.633, P = 0.045). Even after adjusting for age, there was still a significant difference (P = 0.047). Conclusion: Erythrocytes are an independent protective factor for VCI in patients with sWMH. Promoting hematopoietic function may have potential value for prevention of cognitive decline in patients with cerebrovascular disease.
Collapse
Affiliation(s)
- Xi Tao
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
- Department of Neurological Rehabilitation, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, China
| | - Hang Zhou
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
| | - Danheng Mo
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
- Department of Neurological Rehabilitation, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, China
| | - Wenjie Zhang
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
| | - Zihan Chang
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
| | - Yiheng Zeng
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
| | - Yuqi Luo
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
| | - Siyuan Wu
- Department of Neurological Rehabilitation, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, China
| | - Wenjing Tang
- Department of Neurological Rehabilitation, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, China
| | - Chen Yang
- Department of Neurological Rehabilitation, Hunan Provincial People’s Hospital, The First Affiliated Hospital of Hunan Normal University, Changsha, China
| | - Qing Wang
- Department of Neurology, Zhujiang Hospital, Southern Medical University, Guangzhou, China
- *Correspondence: Qing Wang
| |
Collapse
|