1
|
Demir B, Ayna Altuntaş S, Kurt İ, Ulukaya S, Erdem O, Güler S, Uzun C. Cognitive activity analysis of Parkinson's patients using artificial intelligence techniques. Neurol Sci 2024:10.1007/s10072-024-07734-y. [PMID: 39256279 DOI: 10.1007/s10072-024-07734-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/01/2023] [Accepted: 08/19/2024] [Indexed: 09/12/2024]
Abstract
PURPOSE The development of modern Artificial Intelligence (AI) based models for the early diagnosis of Parkinson's disease (PD) has been gaining deep attention by researchers recently. In particular, the use of different types of datasets (voice, hand movements, gait, etc.) increases the variety of up-to-date models. Movement disorders and tremors are also among the most prominent symptoms of PD. The usage of drawings in the detection of PD can be a crucial decision-support approach that doctors can benefit from. METHODS A dataset was created by asking 40 PD and 40 Healthy Controls (HC) to draw spirals with and without templates using a special tablet. The patient-healthy distinction was achieved by classifying drawings of individuals using Support Vector Machine (SVM), Random Forest (RF), and Naive Bayes (NB) algorithms. Prior to classification, the data were normalized by applying the min-max normalization method. Moreover, Leave-One-Subject-Out (LOSO) Cross-Validation (CV) approach was utilized to eliminate possible overfitting scenarios. To further improve the performances of classifiers, Principal Component Analysis (PCA) dimension reduction technique were also applied to the raw data and the results were compared accordingly. RESULTS The highest accuracy among machine learning based classifiers was obtained as 90% with SVM classifier using non-template drawings with PCA application. CONCLUSION The model can be used as a pre-evaluation system in the clinic as a non-invasive method that also minimizes environmental and educational level differences by using simple hand gestures such as hand drawing, writing numbers, words, and syllables. As a result of our study, preliminary preparation has been made so that hand drawing analysis can be used as an auxiliary system that can save time for health professionals. We plan to work on more comprehensive data in the future.
Collapse
Affiliation(s)
- Bahar Demir
- Department of Computational Sciences, Trakya University, Edirne, 22030, Turkey.
| | - Sinem Ayna Altuntaş
- Department of Computational Sciences, Trakya University, Edirne, 22030, Turkey
- Department of Biomedical Device Technology, Trakya University, Edirne, 22030, Turkey
| | - İlke Kurt
- Department of Computational Sciences, Trakya University, Edirne, 22030, Turkey
- Department of Biomedical Device Technology, Trakya University, Edirne, 22030, Turkey
| | - Sezer Ulukaya
- Department of Electrical and Electronics Engineering, Trakya University, Edirne, 22030, Turkey
| | - Oğuzhan Erdem
- Department of Electrical and Electronics Engineering, Trakya University, Edirne, 22030, Turkey
| | - Sibel Güler
- Department of Neurology, Yalova University Faculty of Medicine, Yalova, 77200, Turkey.
| | - Cem Uzun
- Department of Otorhinolaryngology, Head and Neck Surgery, Koç University School of Medicine, İstanbul, 34010, Turkey
| |
Collapse
|
2
|
Brissenden JA, Scerbak T, Albin RL, Lee TG. Motivational Vigor in Parkinson's Disease Requires the Short and Long Duration Response to Levodopa. Mov Disord 2024; 39:76-84. [PMID: 38062630 PMCID: PMC10842158 DOI: 10.1002/mds.29659] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/21/2023] [Revised: 09/27/2023] [Accepted: 10/26/2023] [Indexed: 02/01/2024] Open
Abstract
BACKGROUND Impaired movement vigor (bradykinesia) is a cardinal feature of Parkinson's disease (PD) and hypothesized to result from abnormal motivational processes-impaired motivation-vigor coupling. Dopamine replacement therapy (DRT) improves bradykinesia, but the response to DRT is multifaceted, comprising a short-duration response (SDR) and a long-duration response (LDR) only manifesting with chronic treatment. Prior experiments assessing motivation-vigor coupling in PD used chronically treated subjects, obscuring the roles of the SDR and LDR. METHODS To disambiguate the SDR and LDR, 11 de novo PD subjects (6 male [M]:5 female [F]; mean age, 67) were studied before treatment, after an acute levodopa (l-dopa) dose, and in both the practical "off" (LDR) and "on" (LDR + SDR) states after chronic stable treatment. At each visit, subjects were characterized with a standard battery including the Movement Disorder Society-Sponsored Revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS) and an incentivized joystick task to assess motor performance in response to varying rewards. RESULTS l-Dopa induced a robust SDR and LDR, with further improvement in the combined SDR + LDR state. At baseline, after acute treatment (SDR), and after LDR induction, subjects did not exhibit the normal increase in movement speed with increasing reward. Only in the combined SDR + LDR state was there restoration of motivation-vigor coupling. CONCLUSIONS Although consistent with prior results in chronically treated PD subjects, the significant improvement in motor performance observed with the SDR and LDR suggests that bradykinesia is not solely secondary to deficient modulation of motivational processes. © 2023 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.
Collapse
Affiliation(s)
- James A Brissenden
- Department of Psychology, University of Michigan, Ann Arbor, Michigan, USA
| | - Teresa Scerbak
- Department of Neurology, University of Michigan, Ann Arbor, Michigan, USA
| | - Roger L Albin
- Department of Neurology, University of Michigan, Ann Arbor, Michigan, USA
- Neurology Service and Geriatric Research Education and Clinical Center, Veteran Affairs Ann Arbor Healthcare System, Ann Arbor, Michigan, USA
| | - Taraz G Lee
- Department of Psychology, University of Michigan, Ann Arbor, Michigan, USA
| |
Collapse
|
3
|
Kehnemouyi YM, Petrucci MN, Wilkins KB, Melbourne JA, Bronte-Stewart HM. The Sequence Effect Worsens Over Time in Parkinson's Disease and Responds to Open and Closed-Loop Subthalamic Nucleus Deep Brain Stimulation. JOURNAL OF PARKINSON'S DISEASE 2023:JPD223368. [PMID: 37125563 DOI: 10.3233/jpd-223368] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 05/02/2023]
Abstract
BACKGROUND The sequence effect is the progressive deterioration in speech, limb movement, and gait that leads to an inability to communicate, manipulate objects, or walk without freezing of gait. Many studies have demonstrated a lack of improvement of the sequence effect from dopaminergic medication, however few studies have studied the metric over time or investigated the effect of open-loop deep brain stimulation in people with Parkinson's disease (PD). OBJECTIVE To investigate whether the sequence effect worsens over time and/or is improved on clinical (open-loop) deep brain stimulation (DBS). METHODS Twenty-one people with PD with bilateral subthalamic nucleus (STN) DBS performed thirty seconds of instrumented repetitive wrist flexion extension and the MDS-UPDRS III off therapy, prior to activation of DBS and every six months for up to three years. A sub-cohort of ten people performed the task during randomized presentations of different intensities of STN DBS. RESULTS The sequence effect was highly correlated with the overall MDS-UPDRS III score and the bradykinesia sub-score and worsened over three years. Increasing intensities of STN open-loop DBS improved the sequence effect and one subject demonstrated improvement on both open-loop and closed-loop DBS. CONCLUSION Sequence effect in limb bradykinesia worsened over time off therapy due to disease progression but improved on open-loop DBS. These results demonstrate that DBS is a useful treatment of the debilitating effects of the sequence effect in limb bradykinesia and upon further investigation closed-loop DBS may offer added improvement.
Collapse
Affiliation(s)
- Yasmine M Kehnemouyi
- Stanford University School of Medicine, Department of Neurology and Neurological Sciences, Stanford, CA, USA
- Stanford University School of Engineering, Department of Bioengineering, Stanford, CA, USA
| | - Matthew N Petrucci
- Stanford University School of Medicine, Department of Neurology and Neurological Sciences, Stanford, CA, USA
- Stanford University School of Engineering, Department of Bioengineering, Stanford, CA, USA
| | - Kevin B Wilkins
- Stanford University School of Medicine, Department of Neurology and Neurological Sciences, Stanford, CA, USA
| | - Jillian A Melbourne
- Stanford University School of Medicine, Department of Neurology and Neurological Sciences, Stanford, CA, USA
| | - Helen M Bronte-Stewart
- Stanford University School of Medicine, Department of Neurology and Neurological Sciences, Stanford, CA, USA
- Stanford University School of Medicine, Department of Neurosurgery, Stanford, CA, USA
| |
Collapse
|
4
|
The significance of micrographia as a clinical feature of Parkinson's disease and underlying pathophysiology. Neurol Sci 2023; 44:1791-1793. [PMID: 36593420 DOI: 10.1007/s10072-022-06590-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/09/2022] [Accepted: 12/27/2022] [Indexed: 01/04/2023]
|
5
|
David FJ, Rivera YM, Entezar TK, Arora R, Drane QH, Munoz MJ, Rosenow JM, Sani SB, Pal GD, Verhagen-Metman L, Corcos DM. Encoding type, medication, and deep brain stimulation differentially affect memory-guided sequential reaching movements in Parkinson's disease. Front Neurol 2022; 13:980935. [PMID: 36324383 PMCID: PMC9618698 DOI: 10.3389/fneur.2022.980935] [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: 06/29/2022] [Accepted: 09/26/2022] [Indexed: 11/13/2022] Open
Abstract
Memory-guided movements, vital to daily activities, are especially impaired in Parkinson's disease (PD). However, studies examining the effects of how information is encoded in memory and the effects of common treatments of PD, such as medication and subthalamic nucleus deep brain stimulation (STN-DBS), on memory-guided movements are uncommon and their findings are equivocal. We designed two memory-guided sequential reaching tasks, peripheral-vision or proprioception encoded, to investigate the effects of encoding type (peripheral-vision vs. proprioception), medication (on- vs. off-), STN-DBS (on- vs. off-, while off-medication), and compared STN-DBS vs. medication on reaching amplitude, error, and velocity. We collected data from 16 (analyzed n = 7) participants with PD, pre- and post-STN-DBS surgery, and 17 (analyzed n = 14) healthy controls. We had four important findings. First, encoding type differentially affected reaching performance: peripheral-vision reaches were faster and more accurate. Also, encoding type differentially affected reaching deficits in PD compared to healthy controls: peripheral-vision reaches manifested larger deficits in amplitude. Second, the effect of medication depended on encoding type: medication had no effect on amplitude, but reduced error for both encoding types, and increased velocity only during peripheral-vision encoding. Third, the effect of STN-DBS depended on encoding type: STN-DBS increased amplitude for both encoding types, increased error during proprioception encoding, and increased velocity for both encoding types. Fourth, STN-DBS was superior to medication with respect to increasing amplitude and velocity, whereas medication was superior to STN-DBS with respect to reducing error. We discuss our findings in the context of the previous literature and consider mechanisms for the differential effects of medication and STN-DBS.
Collapse
Affiliation(s)
- Fabian J. David
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Yessenia M. Rivera
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Tara K. Entezar
- School of Integrative Biology, University of Illinois at Urbana-Champaign, Urbana-Champaign, IL, United States
| | - Rishabh Arora
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Quentin H. Drane
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Miranda J. Munoz
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Joshua M. Rosenow
- Department of Neurological Surgery, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Sepehr B. Sani
- Department of Neurosurgery, Rush University Medical Center, Chicago, IL, United States
| | - Gian D. Pal
- Department of Neurology, Rutgers University, New Brunswick, NJ, United States
| | - Leonard Verhagen-Metman
- Department of Neurology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| | - Daniel M. Corcos
- Department of Physical Therapy and Human Movement Sciences, Northwestern University Feinberg School of Medicine, Chicago, IL, United States
| |
Collapse
|
6
|
Associations between resting-state functional connectivity changes and prolonged benefits of writing training in Parkinson's disease. J Neurol 2022; 269:4696-4707. [PMID: 35420350 DOI: 10.1007/s00415-022-11098-8] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/31/2021] [Revised: 03/21/2022] [Accepted: 03/23/2022] [Indexed: 10/18/2022]
Abstract
BACKGROUND Our earlier work showed that automaticity and retention of writing skills improved with intensive writing training in Parkinson's disease (PD). However, whether this training changed the resting-state networks in the brain and how these changes underlie retention of motor learning is currently unknown. OBJECTIVE To examine changes in resting-state functional connectivity (rs-FC) and their relation to behavioral changes immediately after writing training and at 6 week follow-up. METHODS Twenty-five PD patients underwent resting-state fMRI (ON medication) before and after 6 weeks writing training. Motor learning was evaluated with a dual task paradigm pre- and post-training and at follow-up. Next, pre-post within-network changes in rs-FC were identified by an independent component analysis. Significant clusters were used as seeds in ROI-to-ROI analyses and rs-FC changes were correlated with changes in behavioral performance over time. RESULTS Similar to our larger cohort findings, writing accuracy in single and dual task conditions improved post-training and this was maintained at follow-up. Connectivity within the dorsal attentional network (DAN) increased pre-post training, particularly with the right superior and middle temporal gyrus (rS/MTG). This cluster also proved more strongly connected to parietal and frontal areas and to cerebellar regions. Behavioral improvements from pre- to post-training and follow-up correlated with increased rs-FC between rS/MTG and the cerebellum. CONCLUSIONS Training-driven improvements in dual task writing led to functional reorganization within the DAN and increased connectivity with cerebellar areas. These changes were associated with the retention of writing gains and could signify task-specific neural changes or an inability to segregate neural networks.
Collapse
|