ABSTRACT

Neurological disorders significantly impact the health landscape in India, contributing

heavily to both non-communicable and communicable disease burdens. This paper explores
the epidemiology, pathology, aetiology, and pathogenesis of Parkinson's disease (PD) and
ataxia, two prevalent neurological conditions. Parkinson's disease, a common
neurodegenerative disorder characterized by dopaminergic neuron degeneration and Lewy
body formation, primarily affects the elderly and is more prevalent in men. The disease
progression and symptomatology, including motor dysfunctions like tremors and balance
issues, are discussed alongside genetic and environmental risk factors. Ataxia, defined by
impaired coordination due to cerebellar dysfunction, encompasses both inherited and
acquired forms. The clinical presentation, diagnostic challenges, and differential diagnoses
for both PD and ataxia are examined. The study emphasizes the need for early detection and
personalized treatment strategies to manage these complex disorders effectively. Additionally,
the paper provides a comprehensive overview of the shifts in neurological disease burdens in
India from 1990 to 2019, highlighting an increase in non-communicable neurological
disorders. This work underscores the importance of continued research and development in
neurogenetics, molecular pathology, and therapeutic interventions to improve patient
outcomes in these debilitating conditions.
1. INTRODUCTION
Neurological disorders encompass a vast array of conditions that affect the nervous system,
including the brain, spinal cord, and peripheral nerves. Neurological disorders, both fatal and
non-fatal, are among the leading contributors to the burden of non-communicable and
communicable diseases in India [1]. These disorders can arise from genetic, developmental,
metabolic, traumatic, infectious, inflammatory, or degenerative causes, leading to diverse
symptoms and complications. In the past three decades, most studies in India have shown a
high disease burden for specific diseases including stroke, epilepsy, headache, Parkinson’s
disease, and dementia,[3] mostly reported for the urban Indian population.

Parkinson’s disease (PD) is one of the most common neurodegenerative disorders It occurs
due the death of neurons, resulting in a decrease of dopamine levels in the brain. Low levels
of dopamine hamper communication between synapses, causing ineffective motor functions
[2]. While the progress of symptoms may vary from patient to patient, balance problems and
tremors are the most prevalent side-effects of dopaminergic neuron death. Unfortunately,
there is no cure for PD, hence Patients with Parkinson’s (PWP) rely on early detection and
tailored treatments to slow the progress of the disease. A comprehensive physical
examination and medical history of the patient should be part of the differential diagnosis for
Parkinson’s disease (PD). According to epidemiology, Parkinson’s disease majorly affects
elderly persons and frequency of affecting men is more as compared to women where the
worldwide burden of Parkinson’s disease (PD) increased more than twice in the past 20 years.
As part of the Global Burden of Diseases, Injuries, and Risk Factors Study (GBD) 2019, the
India State-Level Disease Burden Initiative Neurological Disorders Collaborators for the first
time have provided a comprehensive analysis of patterns and trends in neurological diseases
from 1990 to 2019,[4] in terms of changes in prevalence or incidence and DALYs with time,
in different Indian states. The Article highlights the increasing contribution of non-
communicable and injury-related neurological disorders, and the decreasing contribution of
communicable neurological disorders, to total DALYs in India. The contribution of non-
communicable neurological disorders was 4·0% (95% UI 3·2–5·0) in 1990, doubling to 8·2%
(6·6–10·2) in 2019; injury related neurological disorders contributed 0·2% (0·2–0·3) in 1990
and 0·6% (0·5–0·7) in 2019; and communicable neurological disorders contributed to 4·1%
(3·5–4·8) in 1990 and 1·1% (0·9–1·5) in 2019.
Deriving from the Greek word "a-taxis," which means "lack of order," ataxia is a term used to
describe a condition in which muscular movements cease to be voluntary. It may be a sign of
a main problem or one of several underlying illnesses. There are two main types of ataxias:
acquired and inherited. Both have different pathophysiological mechanisms and etiological
variables. Understanding the complexities of ataxia is crucial for developing effective
diagnostic tools and treatments. Ataxia, defined as impaired coordination of voluntary muscle
movement and is usually caused by cerebellar dysfunction or impaired vestibular or
proprioceptive afferent input to the cerebellum. Ataxia can also have a subacute onset, as
from infectious or immunologic disorders, which may have a limited window of therapeutic
opportunities. A prompt management strategy for treatable causes of ataxia can save the
patient’s life and result in a good long-term outcome. Ataxia can also be benign in largely
symptomatic disorders (eg, vestibular neuritis). With the advancement of neurogenetics, more
inherited causes of cerebellar ataxia can be diagnosed,[5] but many sporadic ataxias,
including those with a chronic and progressive course, still remain undiagnosed.

The ataxias may be broadly divided into those that are genetic (with or without a family
history) and those that are acquired/degenerative. ‘Sporadic’ ataxia implies there is no family
history. Acquired progressive ataxias can be immune mediated (eg, paraneoplastic
spinocerebellar degeneration, gluten ataxia), degenerative (eg, cerebellar variant of multiple
systems atrophy (type C)), caused by deficiency states (eg, vitamin B12, vitamin E, and so
on), toxicity (eg, alcohol-related ataxia, phenytoin), or associated with infections (HIV,
sporadic Creutzfeldt-Jakob disease, progressive multifocal leucoencephalopathy, and so on).
Inherited ataxias can have autosomal dominant, autosomal recessive, X-linked or
mitochondrial (maternal) inheritance. Metabolic disorders (eg, Niemann-Pick type C, Tay-
Sachs disease), even though ‘inherited’, can present as late-onset ataxia with no family
history, emphasising the need for careful clinical scrutiny and comprehensive and appropriate
laboratory testing [6].
2.LITERATURE SURVEY

H. Bonney, R. de Silva, and P. Giunti,this reference is an extensive manual that focuses on

the most effective clinical techniques for ataxia management. It was released in its third
edition by Ataxia UK and includes the most recent clinical developments and research results
to give medical practitioners the most recent recommendations for identifying and managing
ataxias.[6]

Draoui A, El Hiba O, Aimrane A, El Khiat A, Gamrani H. The translational research in

Parkinson's disease (PD) that fills the knowledge gap between lab findings and clinical
applications is covered in this reference. It illustrates the path from comprehending the
molecular mechanisms of Parkinson's disease (PD) in experimental settings to using these
discoveries in clinical settings. This include investigating illness models, putative therapeutic
targets, and creating novel therapies that can be put to the test in clinical trials.[9]

Crowley EK, Nolan YM, Sullivan AM. This study uses rodent models to investigate how
exercise affects Parkinson's disease symptoms, both motor and non-motor. It provides proof
that physical activity can have neuroprotective effects, which may impede the advancement
of disease and enhance quality of life. The review addresses the potential processes by which
exercise can produce these advantages, including increased mitochondrial activity, decreased
neuroinflammation, and enhanced neuroplasticity.[11]

Masato A, Sandre M, Antonini A, Bubacco L. In order to increase the efficacy of treatments

that target biogenic aldehydes, this research investigates methods for stratifying patients with
Parkinson's disease. Neurodegeneration may be exacerbated by biogenic aldehydes, which
are hazardous consequences of neurotransmitter metabolism. According to the study,
customized strategies based on patient stratification may improve the neuroprotective benefits
of drugs that scavenge these dangerous substances, thereby delaying the course of the
disease.[13]

Xicoy H, Klemann CJ, De Witte W, Martens MB, Martens GJ, Poelmans G.This study
investigates the genetic links between Parkinson’s disease and lipid metabolism. It identifies
specific genetic factors that influence both PD and blood lipid levels, suggesting a shared
etiology. These findings highlight the potential for lipid-related biomarkers in diagnosing and
understanding PD and open avenues for exploring lipid metabolism as a therapeutic target.
[15]

Rocha EM, Keeney MT, Di Maio R, De Miranda BR, Greenamyre JT. The function of the
leucine-rich repeat kinase 2 (LRRK2) gene in idiopathic and familial Parkinson's disease is
the main topic of this article. Although the gene also contributes to the idiopathic form of the
disease, LRRK2 mutations are a prevalent genetic cause of Parkinson's disease (PD). The
functions of LRRK2, its role in Parkinson's disease etiology, and the discovery of LRRK2
inhibitors as possible treatment agents are reviewed in this work.[16]

S. Summa and associates, published in Data in Brief, the journalThis research introduces a
dataset that uses the Kinect sensor to record spatiotemporal gait metrics in people with ataxia.
Subjects with ataxia, a disorder marked by poor coordination, had their walking patterns
recorded using the Kinect, a motion-sensing device. The dataset contains precise parameters
that are essential for assessing anomalies in gait, such as timing, step width, and length. The
goal of the study is to enable future research in the diagnosis, comprehension, and treatment
of ataxia by making this extensive dataset available. The availability of such accurate and
quantitative data enables the construction of possible diagnostic instruments or treatment
plans, as well as more complex studies.[22]

A. S. Nunes, N. Kozhemiako, C. D. Stephen, J. D. Schmahmann, S. Khan, and A. S. Gupta

in this study investigate a novel approach to ataxia assessment by video analysis of finger
tapping activities. The paper presents an automated approach that analyzes movies of fingers
tapping to define the existence and determine the severity of ataxia using machine learning
algorithms. In the finger tapping test, a popular neurological evaluation, participants tap their
fingers repeatedly in unison; abnormalities from typical patterns may suggest ataxia. This
automated approach aims to provide an objective and quantifiable assessment tool, reducing
the subjectivity and variability inherent in traditional clinical evaluations. The development
of such technology could significantly enhance the diagnostic process and monitoring of
ataxia, offering more consistent and reliable measurements of disease progression.[23]
3. TYPES OF NEUROLOGICAL DISORDERS

3.1 PARKINSON'S DISEASE

3.1.1 EPIDEMIOLOGY

Parkinson’s disease is the second most common neurodegenerative disorder, affecting

approximately 1% of the population over the age of 60. The prevalence increases with age,
and the disease affects men slightly more often than women. Since the early 1800s, PD has
gained widespread recognition. Doctors gave the condition its name when it was first
documented [7]. PD, sometimes known as “paralysis agitans,” it is uncommon in young
adults, especially those under 40 [8]. An estimated one million Americans are thought to be
affected with Parkinson’s disease (PD), of which about 60,000 new cases are recorded
annually. 7–10 million people are estimated to have Parkinson’s disease (PD), affecting men
1.5 times more frequently than women. A population-based investigation of Medicare
beneficiaries revealed that the average prevalence of PD among people 65 and older was
1.6% [9].

3.1.2 PATHOLOGY, AETIOLOGY AND PATHOGENESIS

Parkinson's disease is primarily characterized by the degeneration of dopaminergic neurons in

the substantia nigra pars compacta, a region of the midbrain. This loss leads to a significant
reduction in dopamine levels in the striatum, impairing the regulation of movement.
Additionally, the presence of Lewy bodies, which are abnormal aggregates of the protein
alpha-synuclein, is a pathological hallmark of the disease. The earliest documented
pathological changes in PD [10] have been observed in the medulla oblongata/pontine
tegmentum and olfactory bulb. In these early stages—Braak stages 1 and 2—patients are pre-
symptomatic. As the disease advances—Braak stages 3 and 4—the substantia nigra, areas of
the midbrain and basal forebrain become involved. Finally, the pathological changes appear
in the neocortex.

Aggregation of LB contains a wide range of proteins including ubiquitin alpha-synuclein and

ubiquitin, which impair optimal neuron function. Aging and environmental stress, according
to new guidelines, both contribute to neuropathology. Environmental contamination (e.g.,
pesticides), the strain of the growing-old process, or misuse of pills causes a low-stage illness
inside the mind ("inflammation"), persistent. Cellular aging in neurons in the brain over time
is caused by this inflammatory process [11].

The identification of single gene defects in PD has focused interest on the ubiquitin-
proteasome system (UPS) as one potential candidate in the development of cell death [12]
The UPS is important for intracellular proteolysis and a large number of intracellular
processes that maintain the viability of cells. It does this by removing unwanted proteins that
are no longer required by the cell. Failure of the UPS leads to the abnormal aggregation of
proteins including a-synuclein which are a major component of lewy bodies. One of the first
sites for LB deposition in early PD is the olfactory bulb. It is, therefore, of interest that a
disturbance in smell and taste is often one of the earliest clinical features in PD raising the
possibility that LB formation may be integral for the activation of pathways leading to
neuronal dysfunction and death.

The link between UPS and neurodegeneration has been strengthened by the discovery of
mutations in genes which code for several ubiquitin-proteasome pathway proteins in PD.
Details about the pathophysiology of PD are shown in Figure. 1.

Destruction of dopaminergic neuronal cells in

the substantial nigra in the basal gangla

Depletion of dopamine stores

Degeneration of the dopaminergic

nigrostriatal pathway

Imbalance of excitatory (acetylcholino)

and inhibiting(dopamine)neurotransmitters in
the corpus striatum

Impairment of extrapyramidal tracts controlling

complex body movements
 Tremors Rigidity Bradykinesia

Fig. 1 Pathophysiology PD. (Parkinson’s disease is mainly characterized by the neuronal loss within
the SN of the brain, which causes motor and non-motor signs such as tremors, bradykinesia, and
stiffness.)

3.1.3 ROLE OF ENVIRONMENTAL, GENETIC, AND EPIGENETIC FACTORS

IN CAUSING PARKINSON’S DISEASE

The risk of Parkinson's disease is age. This pattern has important public health ramifications:
The number of Parkinson's disease (PD) patients is expected to increase by over 50% by 2030
due to both an aging population and an increase in life expectancy worldwide [13].
Environmental exposures are also risk factors for PD. These factors have been demonstrated
that significantly alter the risk of PD in a meta-analysis of individual capability threat
elements. The notion that smoking could offer defense against the illness has emerged due to
the elements that reduce the chance of Parkinson's disease (PD) associated with smoking. But
the findings of recent studies and thorough case-control studies showed that PD patients can
quit smoking more quickly and that the associations with smoking can be caused by a
decreased reactivity to nicotine in the prodromal stage of the disease. The results of at least
five plausible population-based research indicated a negative relationship between blood
pressure and the risk of Parkinson's disease (PD), a finding that may be more pronounced in
males than in women [14].

Even though there are many variables that could raise one's risk of Parkinson's disease (PD),
awareness of their intricate interactions is growing. For example, the study's circumstantial
findings indicated that the likelihood of Parkinson's disease (PD) was elevated by both brain
trauma exposure and Paraquat.[15] Additional investigation has revealed genetic influences
on environmental risk factors. For instance, the risk associated with consuming coffee is
influenced by single-nucleotide polymorphisms in CYP1A2, which codes for the isoform of
Cytochrome P450 that causes metabolism of GRIN2A, which codes for a portion of the N-
methyl-D-aspartate (NMDA) receptor. Furthermore, the form of a polymorphism combined
with a repeat dinucleotide in the alpha-synuclein protein gene promoter of SNCA (synuclein
gene) influences the risk of Parkinson's disease (PD) associated with head trauma [16].The
various risk factors of parkinson’s disease are shown in the Figure 2.

Fig.2 Risk factors for PD (Parkinson's disease is a central nervous system illness that
frequently results in stiffness, bradykinesia, and tremors.)

3.2 ATAXIA DISEASE

3.2.1 EPIDEMIOLOGY
Many illnesses can cause ataxia, such as autoimmune diseases, infections, genetic problems,
and chemical exposures. Based on the particular underlying etiology, ataxia has varying
prevalences. As an illustration, one in 50,000 people have Friedreich’s ataxia, a prevalent
genetic ataxia.

3.2.1 PATHOLOGY, AETIOLOGY AND PATHOGENESIS

A lesion that interferes with the sensory transmission to the cerebellum may be the cause of
ataxia. Spinal ataxia or sensory ataxia may result from this disease. Cerebellar ataxia is
brought on by a disruption in cerebral impulses from the cerebellum. A consequence of both
of the aforementioned disorders is spinocerebellar ataxia. They are caused by the CAG repeat
on chromosomes and are autosomal dominant.
The most prevalent hereditary ataxia is Friedreich's ataxia. Its inheritance pattern is
autosomal recessive. The frataxin gene is implicated. Peripheral nerve axons are
degenerating, and sensory cells are disappearing. Patients typically appear in their first or
second decade of life. There are multiple problems, including ataxia of the gait, sensory loss,
pescavus, spastic extensor plantar reflexes, atrophy of the extremities, and cardiomyopathy.
Patients may also have hearing loss, vision loss, or diabetes mellitus.[17][18]

Depending on the location of the lesion, characteristic findings are as follows:

 Lesions in the lateral cerebellum cause symptoms on the same side as the lesion
(ipsilateral), whereas diffuse lesions cause generalized symptoms.
 Lesions in the cerebellum hemisphere cause limb ataxia.
 Lesions in the vermis cause truncal, gait ataxia with sparing of the limbs.
 Lesions at vestibule-cerebellar areas cause disbalance, vertigo and gait ataxia.

4. MATERIALS AND METHODS

To improve the classification of Parkinson's disease and ataxia severity prediction, a novel
feature selection is introduced in this section. This feature selection includes three categories:
progressive ataxia (PA), chronic ataxia (CA), and healthy (H) participant. Figure 3 shows the
suggested illness classification architecture. It is claimed that the classification task is a multi-
label classification problem.

Training
set
Ensemble
Disease classification Features & labels Feature Selection model
dataset (Parkinsons & Testing
Ataxia) set

Feature importance ranking

Evaluation metrics by minimizing learning error
and construct ataxia severity
classification prediction
model

Fig 3 Generic classification strategy for Parkinson's and ataxia severity and kind utilizing the
Ensemble model
Using machine learning approaches, this methodology describes a procedure for categorizing
disorders, namely Parkinson's and ataxia. A dataset containing data on patients with
Parkinson's and ataxia is the starting point of the process. Numerous labels (disease
categories) and features(attributes) are included in this collection. Features and labels make
up the dataset's division.
Labels are the output variables that indicate the disease classification (Parkinson's or Ataxia),
while features are the input values that define each patient (e.g., age, symptoms, test results).
Choosing the dataset's most pertinent features is the task of this stage. Feature selection aids
in decreasing overfitting, enhancing model performance, and lowering the dimensionality of
the data. A training set and a testing set are created by dividing the chosen features and labels
into two subsets.

The testing set is used to assess the machine learning model's performance, whereas the
training set is used to train it. The training data is utilized to create an ensemble model.
Ensemble models enhance prediction robustness and overall performance by combining
many machine learning techniques. Techniques like bagging, boosting, and stacking are
frequently used in groups. Based on how much each feature helps to reduce learning mistake,
its significance is ranked. This stage aids in determining which characteristics have the
greatest predictive power for the illness. Furthermore, a classification prediction is developed
for the ataxia severity. In order to assess the model's efficacy, a performance study is
ultimately carried out. In order to evaluate how effectively the model works in categorizing
the diseases, a number of metrics are evaluated, including accuracy, precision, recall, and F1-
score. Using this methodology, Parkinson's and ataxia are reliably classified, the severity of
ataxia is predicted, and pertinent features are chosen and ranked. An ensemble model is
trained, the model's performance is assessed, and the illness classification dataset is pre-
processed.

5. COMPARATIVE ANALYSIS OF DISEASE (PARKINSON’S

& ATAXIA)

This section compares the performance of the standard Ensemble model with generic
methodology-based disease classification. Additionally, the model is contrasted with the
Parkinson's and ataxia severity classification models that are already in use, which are ML
[19] and DL-based [20]. Using the Anaconda Python framework, the proposed and
alternative ataxia and Parkinson's severity classification models are implemented. The dataset
used in the experiment was gathered from [21], [22]. The study employs a Parkinson's and
ataxia severity classification dataset [23], sourced from [22]. The dataset has three classes,
namely healthy, PA, and CA. Consequently, the classification task transforms into a multi-
label classification problem. The Microsoft Kinect software development kit, tripod, and
Kinect v2 are used to create the gait characteristic dataset. Further information about the
dataset can be found in reference [22]. Metrics like accuracy, sensitivity, specificity,
precision, and F-measure are employed to verify the effectiveness of the classification
method.

5.1 SPECIFICITY VS SENSITIVITY MEASURE

This section provides performance study of specificity and sensitivity metrics in classifying
whether a person is ataxic or healthy. A total of five current machine learning algorithms
[20], including random forest (RF), ADABoost, radial basis function (RBF) support vector
machine (RBF-SVM), linear regression (LR), and XGB, as well as one deep learning method
[22], are compared with the suggested ensemble. Figure 4 displays the graphical results of
specificity and sensitivity.

5.2 ACCURACY PERFORMANCE

This section offers a performance analysis of the indicators used to determine if an individual
is ataxic or healthy. Eight current machine learning algorithms [20], including RF,
ADABoost, RBF-SVM, LR, XGB, Bayes, SVM, two-layer neural network (TLNN), and one
DL neural network method [22], are compared with the proposed FSR-XGB. Figure 5
displays the accuracy graphic result. Figure 5 indicates that the ensemble performs better than
any of the currently available ML and DL-based ataxic person identification techniques.
 Specificity & Sensitivity
Accuracy
1.01
1 0.993
0.95
0.98
0.89 0.958
0.96
0.83 0.94
RF AD- RBF- LR DL XGB FSR- 0.94 0.93 0.93 0.933
ABOO SVM XGB 0.92 0.91
ST 0.9
Spe 0.92 0.93 0.93 0.96 0.971 0.900 0.988 0.88
cifici 2
ty 0.86
RF ST VM LR DL B B
Sen 0.96 0.96 0.88 0.88 0.946 0.951 0.997 O S XG - XG
BO F- R
sitiv 8 A RB FS
ity AD

Fig 4 Specificity vs sensitivity performance of different Fig 5 Accuracy performance of different ML and DL
ML and DL algorithms algorithms

CONCLUSION
Conclusively, ataxia and Parkinson's disease are intricate neurological conditions
characterized by a variety of pathologies, etiologies, and clinical manifestations. To improve
patient outcomes and quality of life, it is essential to comprehend the underlying mechanisms
and practical therapeutic solutions for these illnesses. New therapy strategies will be
developed as a result of ongoing investigation into the genetic, molecular, and environmental
components causing these illnesses.

REFERENCES
1. GBD 2016 Neurology Collaborators

Global, regional, and national burden of neurological disorders, 1990–2016: a systematic

analysis for the Global Burden of Disease Study 2016.Lancet Neurol. 2019; 18: 459-480.

2. Braak, H., Braak, E. (2000) “Pathoanatomy of Parkinson’s disease” J Neurol 247, II3–II10.
https://doi.org/10.1007/PL00007758.

3. Nadig R, Namapally USK, Sarma GRK, Mathew T. Outpatient burden of neurological

disorders: a prospective evaluation of 1500 patients. Neurol India 2019; 67: 708–13.

4. India State-Level Disease Burden Initiative Neurological Disorders Collaborators. The

burden of neurological disorders across the states of India: the Global Burden of Disease
Study 1990–2019. Lancet Glob Health 2021; published online July 14.
https://doi.org/10.1016/ S2214-109X(21)00164-9.

5. Nemeth AH, Kwasniewska AC, Lise S, et al. Next generation sequencing for molecular
diagnosis of neurological disorders using ataxias as a model. Brain 2013;136(pt
10):3106Y3118. doi:10.1093/ brain/awt236.

6. Bonney H, de Silva R, Giunti P. Management of the ataxias: towards best clinical practice.
3rd edn. Ataxia UK, 2016.

7. Skidmore FM, Monroe WS, Hurt CP, Nicholas AP, Gerstenecker A, Anthony T, Jololian L,
Cutter G, Bashir A, Denny T, Standaert D (2022) The emerging postural instability phenotype
in idiopathic Parkinson disease. NPJ Parkinson’s Disease 8(1):28

8. Xu Y, Cai X, Qu S, Zhang J, Zhang Z, Yao Z, Huang Y, Zhong Z (2020) Madopar

combined with acupuncture improves motor and non-motor symptoms in Parkinson’s disease
patients: a multicenter randomized controlled trial. Eur J Integr Med 1(34):101049

9. Draoui A, El Hiba O, Aimrane A, El Khiat A, Gamrani H (2020) Parkinson’s disease: from

bench to bedside. Revue Neurol 176(7–8):543

10. Betarbet R, Sherer TB, Greenamyre JT (2005) Ubiquitin-proteasome system and

Parkinson’s diseases. Exp Neurol, 191 (Suppl 1), 17 –27. Review.
11. Crowley EK, Nolan YM, Sullivan AM (2019) Exercise as a therapeutic intervention for
motor and non-motor symptoms in Parkinson’s disease: evidence from rodent models. Prog
Neurobiol 172:2–22

12. Betarbet R, Sherer TB, Greenamyre JT (2005) Ubiquitin-proteasome system and

Parkinson’s diseases. Exp Neurol, 191 (Suppl 1), 17 –27. Review.

13. Masato A, Sandre M, Antonini A, Bubacco L (2021) Patients stratifcation strategies to

optimize the efectiveness of scavenging biogenic aldehydes: towards a neuroprotective
approach for Parkinson’s disease. Curr Neuropharmacol 19(10):1618

14. Gao C, Liu J, Tan Y, Chen S (2020) Freezing of gait in Parkinson’s disease:
pathophysiology, risk factors and treatments. Trans Neurodegener 9:1–22

15. Xicoy H, Klemann CJ, De Witte W, Martens MB, Martens GJ, Poelmans G (2021) Shared
genetic etiology between Parkinson’s disease and blood levels of specifc lipids. NPJ
Parkinson’s Disease 7(1):23

16. Rocha EM, Keeney MT, Di Maio R, De Miranda BR, Greenamyre JT (2022) LRRK2 and
idiopathic Parkinson’s disease. Trends Neurosci 45(3):224–236

17. Pandolfo M. Friedreich ataxia. Arch Neurol. 2008 Oct;65(10):1296-303. [PubMed]

18. Bastian AJ. Mechanisms of ataxia. Phys Ther. 1997 Jun;77(6):672-5. [PubMed]

19. O. Vyšata et al., “Classification of ataxic gait,” Sensors, vol. 21, no. 16, p. 5576, Aug.
2021, doi: 10.3390/s21165576.

20. H. I. Fawaz, G. Forestier, J. Weber, L. Idoumghar, and P. A. Muller, “Deep learning for
time series classification: a review,” Data Mining and Knowledge Discovery, vol. 33, no. 4,
pp. 917–963, Jul. 2019, doi: 10.1007/s10618-019-00619-1.

21. M. Serrao et al., “Dataset on gait patterns in degenerative neurological diseases,” Data in
Brief, vol. 16, pp. 806–816, Feb. 2018, doi: 10.1016/j.dib.2017.12.022.

22. S. Summa et al., “Spatio-temporal parameters of ataxia gait dataset obtained with the
Kinect,” Data in Brief, vol. 32, p. 106307, Oct. 2020, doi: 10.1016/j.dib.2020.106307.
23. A. S. Nunes, N. Kozhemiako, C. D. Stephen, J. D. Schmahmann, S. Khan, and A. S.
Gupta, “Automatic classification and severity estimation of ataxia from finger tapping
videos,” Frontiers in Neurology, vol. 12, p. 795258, Feb. 2022, doi:
10.3389/fneur.2021.795258.