Scroll Top
10 Old Grimsbury Rd, Banbury OX16 3HG, UK

7COM1084 Specialism Research Report

Assignment Brief:

As an MSc research student, you are expected to understand how other areas of computer science relate to your chosen specialism. You must have a grasp of significant current open problems in other specialisms, as well as an understanding of how techniques and methods from other research areas can be applied within your specialism.

In this assignment you must provide an overview of two MSc Research specialism lectures (AI / Networking / Cyber Security / Software Engineering / Data Science) which are NOT the specialism you discussed in Assignment 1.

For each of these two areas, you must discuss the research question presented within the relevant specialist lecture for this specialism, propose some research approaches to investigate this question and identify further work you might undertake which builds on this. You must also explore your personal strengths in this area.

Solution:

Automating Dementia Recognition in Early Stages among British Sign Language Users

1. Introduction

Dementia can be affirmed to be a generalised terminology used for a condition when an individual deciphers impaired changes in personality, thinking, reasoning and also memory disorders. The said disease can be affirmed to have spread globally and near about 55 million persons around the globe reflects symptoms of dementia. Further, as observed by WHO, 60 per cent of the population in middle and low-income nations depicts such problems and dementia is turning out to be a worldwide issue for the healthcare industry (Who. Int, 2021). The underlying problem herein is that the healthcare and medical community are finding it much difficult to keep up with the care and treatment of dementia and thus, the cost of getting the care and appropriate treatment is very costly. Thus, the implied aim of introducing data science within the healthcare aspects is to aid the treatment of dementia through the use of automation and machine learning. In the views of Pellegrini et al., (2018), data science can be defined as the combined study of programming, coding, domain knowledge and the wisdom and expertise of mathematics and statistics to incur meaningful acumen from relevant data. Thus, this study would lay down the diverse discussion and aspects regarding the use of machine learning and algorithm analysis in an in-depth manner so as it may ideally allow dementia recognition in British Sign Language Users with low human intervention.

2. Open research question

Is it possible to automate the recognition of early stages of dementia among deaf signers of British Sign Language Users?

Automatic aspects of disease recognition can be affirmed to an exquisite amalgamation of the healthcare community with that of the data science niche. Currently, care for the deaf community through ADL, IoT and 2D video screening toolkits are not frequently available and also appears not to be very pockets friendly (Liang et al., 2020). Moreover, the care and treatment offered to such impaired deaf individuals are very uneven. Thusly, automation driven treatment of such individuals would open up the scope for better treatment of dementia among these persons and also the cost incurred for getting the treatment would be less comparatively. Further, the use of such an extensive approach using the field of data science would allow more exposure to the application of data science in daily life which in turn effectively enhance the skills and experience of diagnosis of the deaf community (Leung et al., 2020).

Problem Statement

Facial expression and trajectory tracking of real-time hand movement of British Sign Language users would allow automated dementia recognition.

3. Existing and related work

Machine learning and Automated object recognition

In the 21st century, the ambit and scope of machine learning and artificial intelligence has opened up a wide range of aspects for this world and starting from retail to healthcare to education, all sectors have been blessed with the boom of the concerning technological innovations. In the views of Bird et al., (2020), computer vision can be said to be the branch and classified category of Data Science, Artificial Intelligence and Machine learning that specialises in working on videos and digital images to comprehend and deduce the understanding of such contents and data. The involvement of algorithms and other statistical approaches allows machine learning which in turn leads to a path of complete automation. Herein, in dementia, as has been reported by Ooms and Spruit, (2020), the disease deciphers a few known symptoms based on which a caregiver reassures the relevant tests for dementia recognition.

Relevance of the use of Pattern to develop a recognition model for dementia in deaf patients

In the deaf community, many genres of sign languages have come up such as the ASL and the BSL. According to Kapetanios and Angelopoulou, 2020), these language when one uses can be stored in a video and analysis of such data within the video would allow the establishment of a pattern or sets of patterns that are used by BSL users. As per Stephan et al., (2021), in data science, these patterns of data or more precisely speaking set of algorithms are analysed and classified to identify the objects. This form of identification or automatic identification is done through the use of programming preferably via python or open scope models. In the views of Bansal et al., (2018), sharing data profiling and cleansing can also be referred to as a significant part of the procedure. It is pertinent to mention herein that intensive researches in this area have not been treated much as the problem or approach appears to be very ambiguous. Moreover, relative research of such genres can be found in abundance.

Facial expression and trajectory tracking of real-time

According to Spruit and Lytras, (2018), deaf individuals with the use of sign language use facial expressions as well and understanding the specific sign language used by that individual means understanding the movement of his hand along with the expressions. In the words of Russ et al., (2019), both facial expression and motion of the hand are used to recognise what a deaf person intends to say; likewise fracking such movements and change of expressions about real-time and various other models such as RGBD and Neural Network models would allow creating a data set of such relevant movements. This information subsequently could be then classified under object identification and likewise when the information is cross-checked through automation the resultant would be accurate. However, according to Bansal et al., (2018), the probability of success in real life has not been much visible or practical. Thus, this can be said to be a theoretical approach towards the identification of patterns and objects through machine learning and automation.

4. Research approach

The research approach focuses to identify the various methods and sets of principles that need to be adhered to when conducting the entire research (Mukherjee, 2019). The relevant data for the ideal completion of the said research would be collected by a secondary method of data collection. It means data and information would be incurred and allowed only based on the works of others, such as published books, articles, journals, newspapers and similar others. The research approach for this study preferably would be deductive research approach as a secondary method of data collection is involved herein. The research philosophy in such a context is defined as the belief that is adhered to while exploring the data and analysing the same. Among the four major types of research philosophies, the best-suited one of this study would be interpretivism as herein the belief of data collection method is subject and limited to interpreting the data incurred through the work of others.

Further, research design can be affirmed to be a framework or outlining structure that reflects the various methods and techniques that would be obeyed by in this research (Devi, 2017). Herein for this study, explanatory research design would be the best fit. Furthermore, this research would allow understanding the various ways in which automation for dementia recognition at a diverse stage would be possible. Another advantage that would be pointed out through this study is that the relevance and importance of data and information and the use of pattern, coding and programming mainly offer scientific approaches towards real-life problem solving within the healthcare sector. The primary disadvantage that is prima facie herein is the limitation of the data collection method. The concerned study typically focuses on using a secondary method of data collection and thus the scope of interpretation and exposure to a prolonged and thorough analysis of the problem statement would not be achieved. It is pertinent to mention that the time constraints would also be a factor in adding to the disadvantage of the study.

5. Personal investment

To explore appropriate answers to the concerned research, the question reflects the pathways of identification of AI and automation as can be used for determining early stages of dementia among deaf individuals. Investigating the said aspects of the research problem would allow one to scout the potential of care offered to the deaf population and nullify the unevenness at once (Mohajan, 2018). The concerned research aims to contribute not just to the field of data science but also strives to address a larger problem about mankind serving better aspects of caregiving for the healthcare industry.

In terms of personal strengths and experiences, it needs to be mentioned that as the problem identified here, is faced in real life and implementation of such research may be fruitful for others appears to me as much more challenging than considering a hypothetical aspect for research. In this regard, the acquired expertise in investigation and deduction would allow me to ideally search for adequate answers for this research.

6. References

Bansal, D., Chhikara, R., Khanna, K. and Gupta, P., 2018. Comparative analysis of various machine learning algorithms for detecting dementia. Procedia computer science132, pp.1497-1502.

Bird, J.J., Ekárt, A. and Faria, D.R., 2020. British sign language recognition via late fusion of computer vision and leap motion with transfer learning to american sign language. Sensors20(18), p.5151.

Devi, P.S., 2017. Research methodology: A handbook for beginners. Notion Press.

Kapetanios, E. and Angelopoulou, A., 2020, May. Machine Learning for Enhancing Dementia Screening in Ageing Deaf Signers of British Sign Language. In LREC 2020, 9th Workshop on the Representation and Processing of Sign Languages: Sign Language Resources in the Service of the Language Community, Technological Challenges and Application Perspectives (pp. 135-138). European Language Resources Association (ELRA).

Leung, C.K., Fung, D.L., Mushtaq, S.B., Leduchowski, O.T., Bouchard, R.L., Jin, H., Cuzzocrea, A. and Zhang, C.Y., 2020, August. Data science for healthcare predictive analytics. In Proceedings of the 24th Symposium on International Database Engineering & Applications (pp. 1-10).

Liang, X., Angelopoulou, A., Kapetanios, E., Woll, B., Al Batat, R. and Woolfe, T., 2020, August. A multi-modal machine learning approach and toolkit to automate recognition of early stages of dementia among British sign language users. In European Conference on Computer Vision (pp. 278-293). Springer, Cham.

Mohajan, H.K., 2018. Qualitative research methodology in social sciences and related subjects. Journal of Economic Development, Environment and People7(1), pp.23-48.

Mukherjee, S.P., 2019. A guide to research methodology: An overview of research problems, tasks and methods.

Ooms, R. and Spruit, M., 2020. Self-service data science in healthcare with automated machine learning. Applied Sciences10(9), p.2992.

Pellegrini, E., Ballerini, L., Hernandez, M.D.C.V., Chappell, F.M., González-Castro, V., Anblagan, D., Danso, S., Muñoz-Maniega, S., Job, D., Pernet, C. and Mair, G., 2018. Machine learning of neuroimaging for assisted diagnosis of cognitive impairment and dementia: a systematic review. Alzheimer’s & Dementia: Diagnosis, Assessment & Disease Monitoring10, pp.519-535.

Russ, T.C., Woelbert, E., Davis, K.A., Hafferty, J.D., Ibrahim, Z., Inkster, B., John, A., Lee, W., Maxwell, M., McIntosh, A.M. and Stewart, R., 2019. How data science can advance mental health research. Nature human behaviour3(1), pp.24-32.

Spruit, M. and Lytras, M., 2018. Applied data science in patient-centric healthcare: Adaptive analytic systems for empowering physicians and patients.

Stephan, B., Julovich, D.A., Bracy, D. and Nguyen, J., 2021. Rule out Screening for Undiagnosed Dementia and Alzheimer’s Disease Using an EHR Based Machine Learning Solution. SMU Data Science Review5(1), p.5.

Who.Int, 2021, https://www.who.int/news-room/fact-sheets/detail/dementia. Accessed 13 Nov 2021.

Related Posts

Leave a comment

× WhatsApp Us