Early Detection of Cognitive Skill Impairment Using Deep Learning

Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and

GRU

Sunita Patil

and Swetta Kukreja

Amity School of Engineering and Technology, Amity University Maharashtra, Mumbai. Maharashtra -410206, India

Keywords: Cognitive Skill Impairment, Early Detection, Deep Learning Models, CNN, RNN, GRU, LSTM, GPT.

Abstract: Early detection of cognitive skill impairment is an important key in discovering the earliest possible

intervention and management. This paper presents a comparison of five deep learning-based models:

Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), Generative Pretrained

Transformer (GPT), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) applied on the

task of cognitive skill impairment classification. The best models are discussed and compared over several

key performance metrics, accuracy, F1-score, precision, recall (sensitivity), Matthews Correlation Coefficient

(MCC). The results show that the RNN outperforms all other models with an accuracy of 99.2%, GRU follows

closely with an accuracy of 98.7%, precision of 98.7%. The results of GPT and LSTM are almost similar with

accuracies of 98.5% and 98.8% but need more resources in memory to be used: 185 MB and 180 MB,

respectively. CNN did not lag as it had an accuracy is 98.5%, precision is 98.6% and combined with a memory

usage of 176 MB. Overall, the RNN emerged as the most efficient model, balancing high classification

accuracy with low memory consumption, and thus most suitable for real-time and resource-constrained

applications. This comparative analysis sets out the strengths and trade-offs of each model, providing valuable

insights for further development in this field of detecting cognitive impairment.

1 INTRODUCTION

Cognitive skills are used in various domains such as

education, workforce performance, and mental

health. To carry out routine tasks in daily life,

cognitive skill is essential. The term "cognitive skill"

describes a group of mental health abilities and

functions that allow people to interpret, process, and

use data. Memory, logical thinking, and problem

solving are examples of cognitive skills. Many mental

health disorders can lead to cognitive impairments.

For example, conditions like depression, anxiety, and

schizophrenia, dementia can affect cognitive

functions like problem-solving abilities, memory,

Communication, attention, concentration. This

cognitive impairment can affect everyday activities,

work performance, and overall quality of life.

Now a days Cognitive skill impairment problem

become more sensitive. Detecting cognitive skill loss

https://orcid.org/ 0009-0004-9828-5274

https://orcid.org/0000-0003-4754-8826

early allows for timely intervention and appropriate

treatment. Early detection can advantage to better

management and mediation schemes. Early

identification can help prevent further deterioration

and improve outcomes through targeted

interventions. Traditional assessment methods have

limitations in terms of personalization and

interactivity (Sanchez, Melo, et al. , 2022).

Traditional screening methods often depends on

clinical evaluations and interactive evaluations, but

these can be time-consuming, inconsistent and

Costly. Existing methods often shortage the required

correctness and specificity for early diagnosis,

specifically in the initial or mild phases of cognitive

decline skill.

Overall, the advantages of early identification

highlight the significance of routine cognitive testing,

particularly for those who are more vulnerable

because of age, family history, or health issues.

238

Patil, S. and Kukreja, S.

Early Detection of Cognitive Skill Impairment Using Deep Learning Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and GRU.

DOI: 10.5220/0013590000004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 238-245

ISBN: 978-989-758-763-4

Frequent examinations and screenings can help

promote proactive healthcare management and better

results for individuals with cognitive skill

impairments.

1.1 Motivation

The motivation behind developing cognitive skill

detection methods in the context of mental health is

to improve overall patient care and outcomes. By

identifying and addressing cognitive skill loss,

healthcare professionals can provide targeted

interventions, improve treatment planning, and

increase the value of life for those experiencing

mental health problems. Additionally, integrating

cognitive skill assessment into mental health care can

contribute to a more broad and holistic recognizing of

a person's well-being, promoting a more personalized

and effective approach to treatment.

Figure 1: Common sign of Cognitive Skill Impairments

1.2 Role of Deep Learning

Recent advancement in artificial intelligence,

exclusively deep learning, has proven ability for

improving diagnostic accuracy and effectiveness.

Deep learning utilizes multi-layered neural networks,

has capabilities in identifying patterns and guessing

results across several areas. In the cognitive skill

impairments, deep learning models can examine

complex datasets to detect small signs of early

cognitive decline. Proposed research work

investigates the use of modern deep learning models,

particularly LSTM, GPT and GRU, RNN and CNN

to analyse structured data for predicting cognitive

skill impairment. The research aims to provide a

comparative study of these models to assess their

efficacy in early detection using a dataset with

demographic and health information of Person.

1.3 Significance of the Study

The significance of utilizing deep learning for early

discovery of cognitive impairments keeps in its

potential to modernise diagnostic methods. By

leveraging deep learning, the investigation aims to

better the accuracy and relevance of cognitive

impairment detects, hypothetically leading to earlier

interference and healthier management of

neurodegenerative diseases. Additionally, the vision

grown from this research could contribute to the

wider area of medical diagnostics. Moreover,

integrating deep learning into cognitive assessment

procedures can modernise screenings, less time-

consuming, lower diagnostic costs. As these models

advance and improve, they could also contribute to a

deeper knowledge of the underlying tools of

cognitive skill impairments, requiring insights that

could initiative future research and medical treatment

development.

1.4 Structure of the Paper:

The entire paper is arranged as following:

Literature Review: A review of current research

on cognitive impairment detection and the purpose of

deep learning in medical diagnostics.

Methodology: A complete narrative of the data

sources, deep learning model architecture, and

evaluation methods used in this study.

Results: Presentation and analysis of the model’s

performance.

Discussion: Interpretation of the results,

including implications for clinical practice and future

research directions.

Conclusion: Summary of findings, contributions

of the study, and recommendations for implementing

deep learning models in early detection of cognitive

impairments.

2 RELATED WORK

2.1 Cognitive Skill Impairment:

Identification and Obstacles

Cognitive function decline, extensive conditions like

dementia, Alzheimer’s disease, Parkinson's disease,

Early Detection of Cognitive Skill Impairment Using Deep Learning Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and

GRU

239

difficulties in both identification and treatment.

Existing diagnostic approaches usually utilize

neuropsychological assessments, clinical

assessments, and brain imaging methods. For

example, the Mini-Mental State Examination and the

Montreal Cognitive Assessment are commonly used

tools for screening and estimating cognitive abilities

(Julayanont, Phillips, et al. , 2012), (Rodriguez,

Smailagic, et al. , 2010). Disadvantage of MMSE are

limited assessment and Functional Impairment

(Larsena, Lomholta, et al. , 2007). Brain imaging

methods, like Magnetic Resonance Imaging (MRI) is

key for observing brain structure. But MRI can be

costly and may not detect small cognitive changes

(Rao, and, Aparn, 2023), Difficulty in segmenting

small, variable brain areas like the hippocampus

(Rao, and, Aparn, 2023).

2.2 Artificial Intelligence for Detecting

Cognitive Impairment

Current research has concentrated on utilizing deep

learning for the early detection of cognitive

impairments. Deep learning CNN model for

Alzheimer's disease classification used (Battineni,

Chintalapudi, et al. , 2021), this paper has limitation

Lack of comparison with other deep learning models.

Research on early detection of cognitive impairment

has habitually been trained in expert evaluations,

clinical assessments and cognitive tests. With the

rapid growths of machine learning, scientists have

shifted toward data-driven simulations. Deep learning

models, exceptionally in the domain of NLP and

organized data analysis, have shown potential in

extracting samples that could indicate cognitive skill

decline.

Existing methods works have employed various

techniques, including NLP (Shan, Zhang, et al. ,

2022) Random Forest(Niyas, Thiyagarajan, et al. ,

2023), Naïve Bayes Classification (Mayilvaganana,

Kalpanadevib, et al. , 2015), decision trees, support

vector machines (SVM) (Niyas, Thiyagarajan, et al. ,

2023), simple CNN(Battineni, Chintalapudi, et al. ,

2021). However, latest transformer deep learning

models like BERT and GPT, as well as RNN-based

architectures like GRU and LSTM, have not been

broadly compared in this specific task of cognitive

skill impairment detection.

Figure 2: Key Finding of Survey Conducted

2.3 Survey Design and Study Area

This research conducted in-person interviews with

different participants located in Senior Citizen Park,

Old age home centres and Hospitals. As shown in Fig

2 the survey is conducted for age between 50 to 65

in order to determine the necessity early detection of

cognitive skill impairment . It is observed that there

are 40 % peoples are somewhat able to prioritize tasks

but occusionally miss dedalines.It is also observed

that 30% people have trouble falling asleep or

excessive sleep.As Fig 2 also shows that 63 %

peoples are rarely able to concentration without

getting distracted.Overall observation of this fig is

there is need to early detection of cognitive skill

impairement.

3 PROPOSED METHODOLOGY

3.1 Dataset Preparation

The study included individuals aged 60 years or

older. Those who stated latest head damage in the past

few months or were confined to bed due to any

neurological disorder were excluded from the

research study. For dataset collection propsed

research work mainely focus on three categories i.e.

normal group,mild cognitive skill impairement group

and severe cognitive skill impairement group as

follows.

INCOFT 2025 - International Conference on Futuristic Technology

240

Figure 3: Sample Collection Categories

The dataset used in this study contains

demographic information, health-related features and

some questionnaires to conduct Cognitive Test

3.2 Dataset Description

The feature set of datasets as follows:

Table 1: Feature Set of Dataset

Variable Type Description

Age Numerical Respondent’s age

(

inte

value

)

Gender Categorica

Gender of the

respondent (e.g.,

Male, Female,

Other

)

Are you working

anywhere?

Categorica

Working status

(e.g., working,

Unemployed,

Student, Retired,

etc.

)

Are you Married? Binary Marital status (1:

Married, 0: Not

Married).

Are you BP

Patient?

Binary High blood pressure

status (1: Yes, 0:

No).

Are you Diabetic

Patient?

Binary Diabetes status (1:

Diabetic, 0: Non-

Diabetic

)

Do you have

Family History of

cognitive skill

impairment?

Binary Family history of

cognitive skill

impairment (1: Yes,

0: No).

Cognitive Test

Score

Numerical Score obtained from

test

Have you

diagnosed

depression?

Binary Depression

diagnosis (1: Yes, 0:

No).

Cognitive skill

Impairment Level

Target

(Categoric

al)

The level of

cognitive skill

impairment

Dataset is intended to design to measures

different cognitive skill such as memory, attention,

logical reasoning, and problem-solving. For the

cognitive test preparation total 15 questions are used

for different cognitive skill such as Problem solving,

logical Reasoning, Attention, Concentration and

Memory. Total marks of Cognitive test are 30.

3.3 Classification of Cognitive skill

Impairment Levels Based on

Cognitive test Scores

Initially person must attempt a cognitive test. The

Cognitive test includes 15 questions of 30 marks. The

cognitive skill impairment levels are classified based

on the obtained scores as depicted in the figure 4.

Figure 4: Classification of Cognitive skill Impairment

4 SYSTEM ARCHITECTURE

The historical data is separated into training and

testing dataset. Proposed Model is trained using Deep

learning algorithm which is ready to accept user data.

Incoming User data is applied to trained model which

is ready to predict cognitive skill impairment level.

Proposed model implementation and algorithm will

discuss in next section.

Figure 5: System Architecture

4.1 Algorithm

In this research work Deep learning algorithm was

randomly assigned to training 80% and testing 20%.

Early Detection of Cognitive Skill Impairment Using Deep Learning Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and

GRU

241

Model is built using training data and testing data are

used to measure the prediction error and overtraining

also. This paper employed several deep learning

architectures to process different types of data and

extract meaningful patterns for the early detection of

cognitive skill impairment. The algorithms used

include Recurrent Neural Networks (RNN), Gated

Recurrent Units (GRU), Generative Pretrained

Transformers (GPT), Long Short-Term Memory

Networks (LSTM), and Convolutional Neural

Networks (CNN). Each of these models was selected

for its specific ability to process and analyse

sequential, time-series, or structured data associated

with cognitive decline.

4.1.1 Recurrent Neural Networks (RNN)

RNNs have been utilized to determine sequential

medical history that might identify early recognition

of cognitive impairment by just inspecting the

progression of cognitive decline based on prior health

conditions.

4.1.2 Gated Recurrent Units (GRU)

By utilizing gates, GRUs would be able to deal with

moderate sequential dependencies during the

progression of dementia. They have been applied in

patient time-series data that tracked symptoms and

medical histories over a period, maintaining a balance

between computational efficiency and performance.

4.1.3 Generative Pretrained Transformer

(GPT)

GPT was mainly in the process of processing natural

language data, which included patient responses in

chatbot interactions. The model analysed the

conversational data, thus detecting subtle changes in

linguistic signs indicating cognitive decline, such as

word choice, sentence structure, and flow of

conversations. It proved to be an essential component

in the detection of early cognitive impairments based

on free-text responses.

4.1.4 Long Short-Term Memory Networks

(LSTM)

LSTMs were applied to capture long-term patient

data. This model comes in handy if the data points

show some longer trends, like gradual cognitive

decline that spreads over months or years.

4.1.5 Convolutional Neural Networks (CNN)

CNNs have been adopted in this study for processing

structured demographic and health data, after

extracting the relevant features related to cognitive

impairment. The hierarchical nature of CNNs allows

automated detection of patterns in variables like age,

medical history, and lifestyle, contributing to the

overall predicting of risks toward potential cognitive

impairments.

5 RESULT AND DISCUSSION

Table 2: Performance Comparison

Algor

ith

Accu

racy

Preci

sion

Recall

(Sensi

tivity)

ore

Matth

ews

Correl

ation

Coeffi

cient

mor

Usa

RNN

0.99

2 0.992

0.9

92 0.987

103

GRU

0.98

0.987

0.9

0.979

145

GPT

0.98

0.988

0.9

0.977

185

LST

0.98

0.988

0.9

0.982

180

CNN

0.98

0.985

0.9

0.977

176

The performance of various deep learning

algorithms—RNN, GRU, GPT, LSTM, and CNN—

was evaluated based on several classification metrics:

accuracy, precision, recall (sensitivity), F1-score,

Matthews Correlation Coefficient (MCC), and

memory usage. These metrics provide a

comprehensive overview of the models'

effectiveness, efficiency, and suitability for detecting

cognitive skill impairment in individuals.

5.1 Model Performance

RNN had the best overall performance as its

accuracy, precision, recall (sensitivity), and F1-score

were reported at 99.2%, respectively. Moreover, the

model's MCC was at 0.987, which means a strong

correlation existed between the true labels and

predicted labels. The high recall and precision of the

RNN indicate that it provides an about-equitable

balance between false positives and false negatives,

making it a sound model for detecting early

impairment in cognitive skills. Memory usage by

INCOFT 2025 - International Conference on Futuristic Technology

242

RNN was measured to be 103 MB, which is relatively

moderate compared to the other models.

Figure 6:. Model’s Performance Evaluation Analysis

GRU does similarly with an accuracy of 98.7%, a

precision of 98.7%, a recall of 98.7%, and F1-score

of 98.7%. MCC value is 0.979, again very good but

slightly less than that of RNN. Thus, despite these

minor differences, GRU gets close to being identical

to the results obtained by RNN for all classification

metrics. However, the reported memory usage for

GRU was surprisingly negative (-145 MB), which

might indicate some anomaly in measurement or

system-specific issues. Further investigation into the

model's memory consumption would be required to

address such discrepancy. For GPT, the result was

slightly lower than for RNN and GRU, at 98.5%

accuracy, 98.3% precision, 98.8% recall, and 98.5%

F1-score, with the MCC at 0.977, also indicating a

drop in robustness of the model as compared to the

RNN and GRU. Memory usage for GPT was higher,

at 185 MB, which is within expectations due to its

large transformer architecture. Even though the

performance remains strong, the higher memory

requirement might limit the applicability in resource-

constrained environments’ is basically similar with

GPT where accuracy of the model remained at 98.8%,

precision was at 98.9%, recall was also at 98.8%, and

F1-score at 98.8%. The MCC that comes with the

model is 0.982, which reflects strong classification

reliability. However, on the other hand, memory

consumption of 180 MB is hefty, which demonstrates

that a model as efficient as the LSTM may consume

many resources. CNN was able to attain an accuracy

of 98.5%, precision of 98.6%, recall of 98.5%, and

F1-score of 98.5%. MCC of 0.977 for CNN matches

well with GPT, hence putting CNN on par in terms of

classification quality with other models. But memory

usage at 176 MB is reasonable but stays above the

RNN which might make CNN less preferable for

deployment in memory-limited environments.

6 CONCLUSION AND FUTURE

SCOPE

Each of these algorithms was selected based on its

ability to model different aspects of the data related

to cognitive skill impairment. While RNN, GRU, and

LSTM excel in handling sequential data, GPT focuses

on capturing complex linguistic patterns from

conversational data, and CNN is effective in

extracting features from structured medical data.

Together, these models provide a comprehensive

approach to early detection of cognitive impairment

by leveraging both time-series and text-based patient

information.

The future scope of this research would be

expansion of the datasets for diverse populations and

deployment of the models in telemedicine and mobile

health apps will bridge the gap between research and

practice, ensuring equitable and ethical use.

ACKNOWLEDGEMENTS

We wish to thank Dr. Bhushan Chaudhari,

Psychiatrist, Dr. D. Y. Patil Medical College,

Hospital and Research Centre, Pimpri, Pune. for the

guidance, constructive comments, at all stages of this

research work. His expertise and unconditional

support have shaped the direction and quality of this

work.

REFERENCES

M. Mayilvaganana , D. Kalpanadevib,“Cognitive Skill

Analysis for Students through Problem Solving Based

on Data Mining Techniques”, Available online at

www.sciencedirect.com, Procedia Computer Science

47 ( 2015 ) 62 – 75

Montserrat Mateos-Sanchez 1, , Amparo Casado Melo 2 ,

Laura Sánchez Blanco 2 and Ana M. Fermoso García 1

“Chatbot, as Educational and Inclusive Tool for People

with Intellectual Disabilities”, Sustainability 2022, 14,

1520. https://doi.org/10.3390/su14031520

Rohit Binu Mathew, Sandra Varghese, Sera Elsa Joy,

Swanthana Susan Alex,“Chatbot for Disease Prediction

and Treatment Recommendation using Machine

Learning”, Proceedings of the Third International

Conference on Trends in Electronics and Informatics

(ICOEI 2019) IEEE Xplore Part Number: CFP19J32-

ART; ISBN: 978-1-5386-9439-8.

Eileen Bendiga Benjamin Erbb Lea Schulze-Thuesinga

Harald Baumeistera,“The Next Generation: Chatbots in

Clinical Psychology and Psychotherapy to Foster

Mental Health – A Scoping Review”,

Early Detection of Cognitive Skill Impairment Using Deep Learning Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and

GRU

243

Verhaltenstherapie 2022;32(suppl 1):64–76, August

20, 2019

Tamizharasi B., Jenila Livingston L.M. and S.

Rajkumar,“Building a Medical Chatbot using Support

Vector Machine Learning Algorithm” National

Science, Engineering and Technology Conference

(NCSET) 2020, doi:10.1088/1742-

6596/1716/1/012059

Iris Rawtaer1 , MBBS, MMed, MCI; Rathi Mahendran2 ,

MBBS, MMed, FAMS; Ee Heok Kua2 , MBBS, MD,

FRCPsych; Hwee Pink Tan3 , PhD; Hwee Xian Tan3 ,

PhD; Tih-Shih Lee4 , MD, PhD; Tze Pin Ng2 , MBBS,

PhD “Early Detection of Mild Cognitive Impairment

With In-Home Sensors to Monitor Behavior Patterns in

Community-Dwelling Senior Citizens in Singapore:

Cross-Sectional Feasibility Study”, Med Internet Res

2020 | vol. 22 | iss. 5 | e16854 | p. 1

Eliane M. Boucher, Nicole R. Harake, Haley E. Ward,

Sarah Elizabeth Stoeckl, Junielly Vargas, Jared Minkel,

Acacia C. Parks & Ran Zilca,” Artificially intelligent

chatbots in digital mental health interventions: a

review”, Expert Review Of Medical Devices 2021,

VOL. 18, NO. S1, 37-49

https://doi.org/10.1080/17434440.2021.2013200.

Ines Hungerbuehler1 , PhD; Kate Daley 1 , BSc, MSc,

DClinPsych; Kate Cavanagh2 , BA, DPhil,

DClinPsych; Heloísa Garcia Claro3,4 , BSN, MSc,

PhD; Michael Kapps1 , BA ,“Chatbot-Based

Assessment of Employees’ Mental Health: Design

Process and Pilot Implementation”, JMIR Form Res

2021 | vol. 5 | iss. 4 | e21678 | p. 1.

Abid Hassan , 1 M. D. Iftekhar Ali , 1 Rifat Ahammed , 1

Sami Bourouis , 2 and Mohammad Monirujjaman Khan

1,” evelopment of NLP-Integrated Intelligent Web

System for EMental Health”, Computational and

Mathematical Methods in Medicine Volume 2021,

Article ID 1546343, 20 pages

https://doi.org/10.1155/2021/1546343

Muhammed Niyas K. P1 Thiyagarajan P2 ,“A Systematic

Review on Early Prediction of Mild Cognitive

Impairment to Alzheimers Using Machine Learning

Algorithms”, International Journal of Intelligent

Networks, 29 March 2023.

A. Revathi , 1 R. Kaladevi , 2 Kadiyala Ramana , 3 Rutvij

H. Jhaveri , 4 Madapuri Rudra Kumar , 3 and M.

Sankara Prasanna Kumar 3 ,“Early Detection of

Cognitive Decline Using Machine Learning Algorithm

and Cognitive Ability Test”, Hindawi Security and

Communication Networks Volume 2022, Article ID

4190023, 13 pages

https://doi.org/10.1155/2022/4190023.

Ashir Javeed1,2 Ana Luiza Dallora2 · Johan Sanmartin

Berglund2 · Arif Ali3 · Liaqata Ali4 · Peter

Anderberg2, ,“Machine Learning for Dementia

Prediction: A Systematic Review and Future Research

Directions”, Journal of Medical Systems (2023) 47:17

https://doi.org/10.1007/s10916-023-01906-7.

Francesco Salis , Diego Costaggiu and Antonella Mandas

“Mini-Mental State Examination: Optimal Cut-Off

Levels for Mild and Severe Cognitive Impairment”,

Department of Medical Sciences, and Public Health,

University of Cagliari, SS 554 Bvio Sestu, 09042

Monserrato, Italy

Rahul Singhal,“Early detection of dementia using Machine

Learning techniques”, Journal of IMS Group ISSN No.

0973-824X Volume 19, No.01, January-June, 2022, pp.

1- 6.

Prabod Rathnayaka , Nishan Mills , Donna Burnett ,

Daswin De Silva * , Damminda Alahakoon and Richard

Gray “A Mental Health Chatbot with Cognitive Skills

for Personalised Behavioural Activation and Remote

Health Monitoring”, Sensors 2022, 22, 3653.

https://doi.org/10.3390/s22103653

https://www.mdpi.com/journal/sensors.

“Mental Health Assessment for the Chatbots”, Yong Shan1

, Jinchao Zhang1 , Zekang Li23, Yang Feng23, Jie

Zhou1, rXiv:2201.05382v1 [cs.CL] 14 Jan 2022.

Mrs. Smita Desai1 ; Ms. Bharati Ambali2 ,“Early Detection

of Alzheimer’s Using EEG”, , International Journal of

Computer Science and Mobile Computing, Vol.12

Issue.3, March- 2023, pg. 34-39.

James C. L. Chow 1,2,* , Leslie Sanders 3 and Kay Li

4,“Design of an Educational Chatbot Using Artificial

Intelligence in Radiotherapy”, AI 2023, 4, 319–332.

https://doi.org/10.3390/ai4010015.

Wei Ying Tana,b, Carol Hargreavesc , Christopher Chend,e

and Saima Hilala,d,e,“A Machine Learning Approach

for Early Diagnosis of Cognitive Impairment Using

The authors. Published by IOS Press. This is an Open

Access article distributed under the terms of the

Creative Commons Attribution-Non-commercial

License (CC BY-NC 4.0)

Christos Karapapas and Christos Goumopoulos “Mild

Cognitive Impairment Detection Using Machine

Learning Models Trained on Data Collected from

Serious Games”, Appl. Sci. 2021, 11, 8184.

https://doi.org/10.3390/app11178184

https://www.mdpi.com/journal/applssci.

Ryoma A Kawakami Douglas W. Scharre Xia Ning

,“Detection of cognitive impairment from eSAGE

cognitive data using machine learning”, Alzheimer’s

Alzheimer’s Association.

wileyonlinelibrary.com/journal/alz 1 of 5

https://doi.org/10.1002/alz.063690.

F Yildiz Kaya , Ozlem erden aki , Ufuk Can , Eda Derle

“Validation of Montreal Cognitive Assessment and

Discriminant Power of Montreal Cognitive Assessment

Subtests in Patients With Mild Cognitive Impairment

and Alzheimer Dementia in Turkish Population”,

Journal of Geriatric Psychiatry and Neurology

27(2),February 2014.

Julayanont, P., Phillips, N., Chertkow, H., and Nasreddine,

Z.S. The Montreal Cognitive Assessment (MoCA):

Concept and Clinical Review. To appear in A.J. Larner

(Ed.), Cognitive Screening Instruments: A Practical

Approach. Springer-Verlag, pp. 111-152.

, Kirsten Schultz-Larsena,b,* Rikke Kirstine Lomholta ,

Svend Kreinerc “Mini-Mental Status Examination: A

INCOFT 2025 - International Conference on Futuristic Technology

244

short form of MMSE was as accurate as the original

MMSE in predicting dementia”, Journal of Clinical

Epidemiology 60 (2007) 260e-267

Ingrid Arevalo‐Rodriguez ,Nadja Smailagic ,Marta Roqué

i Figuls ,Agustín Ciapponi, Erick Sanchez‐Perez Antri

Giannakou ,Olga L Pedraza ,Xavier Bonfill Cosp

,“Mini‐Mental State Examination (MMSE) for the

detection of Alzheimer's disease and other dementias in

people with mild cognitive impairment (MCI)”

https://doi.org/10.1002/14651858.CD010783.pub2

Battula Srinivasa Rao And Mudiyala Aparn “A Review on

Alzheimer’s Disease Through Analysis of MRI Images

Using Deep Learning Techniques”, Digital Object

Identifier 10.1109/ACCESS.2023.3294981,IEEE.

Gopi Battineni a , Nalini Chintalapudi b , Francesco

Amenta c and Enea Traini ,“Deep Learning Type

Convolution Neural Network Architecture for

Multiclass Classification of Alzheimer’s Disease”, In

Proceedings of the 14th International Joint Conference

on Biomedical Engineering Systems and Technologies

(BIOSTEC 2021) - Volume 2: BIOIMAGING, pages

209-215 ISBN: 978-989-758-490-9

Early Detection of Cognitive Skill Impairment Using Deep Learning Models: A Comparative Analysis of CNN, RNN, GPT, LSTM and

GRU

245