Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Chapter 1

INTRODUCTION
The first description of Dyslexia appeared in 1896, authored by Dr. W. Pringle Morgan in Sussex,
England. The word Dyslexia is derived from the Greek word ‘dys’ meaning difficult and the word
‘lexis’ meaning words. Originally Dyslexia was called ‘reading blindness’.

Overview
Dyslexia is a learning disability that causes difficulties to read and understand written
language. Dyslexia is caused by a phonological processing problem whereby children who are
diagnosed with dyslexia have trouble manipulating the language. Dyslexic children also face
problems learning in regular classes. If this is not identified early, the children will be lacking in
their development thus losing their self-esteem.

It affects from 10% to 17.5% of the population in the U.S.A., 8.6% to 11% in Spain, and has a
considerable presence in web text. Competitive reading and writing are required in our education
system; therefore, school failure is associated with dyslexia, even if dyslexia is not related to
overall intelligence. Dyslexia, also known as ‘reading disorder’ is characterized by trouble with
reading despite normal intelligence.

Different people are affected to varying degrees. Problems may include difficulties in spelling
words, reading quickly, writing words, "sounding out" words in the head, pronouncing words when
reading aloud and understanding what one reads. Often these difficulties are first noticed at school.
People with dyslexia have higher rates of attention deficit hyperactivity disorder (ADHD),
developmental language disorders, and difficulties with numbers.

Symptoms of Dyslexia
The various symptoms of dyslexia include:
 Delayed speech.
 Difficulty in memorizing the alphabet.
 Delayed establishment of a dominant hand.
 Difficulty in sound pronunciation.
 Difficulty in pronouncing multisyllabic words.
 Dysgraphia - Difficulty in writing.
 Letter and word reversals in writing.

DEPT OF ECE, BGSIT BG Nagara Page 1

 Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Problem Statement

Dyslexia (pronunciation and talking) is the ability to physically move the tongue, lips, teeth
and jaw to produce sequences of speech sounds, which make up words and sentences that can be
easily understood, which others could interpret in order to express basic needs. Dyslexia disorders
involve difficulties in articulating specific types of sounds. It often involves substitution of one
sound for another, slurring of speech, or indistinct speech and it has become a major challenge in
the 21st century. Dyslexia problems reduce speech intelligibility and communication. It also affects
person's interpersonal communication, personality, social adaptive capability and learning ability.
Hence Therapy sessions are necessary for Dyslexia correction. In specific it is more important for
the children in the age group of 7-9 years, as it is a best time for detection and correction of
Dyslexia disorder. The therapy sessions help children to become clear confident communicators so
that they can become fully engaged in school and benefit from the curriculum, develop self-help
skills and independence for activities of daily living, actively participate in life experiences, and
build healthy social relationships.
In the sequence of therapy, language therapist subjectively utilizes clinical experience to
individualized assessment, treatment, and training. But assessing and treating children with
Dyslexia disorder is a major challenge for therapists, as it becomes difficult to analyze and select
the proper strategy for correction of Dyslexia in certain situations in spite of clinical experience.
Improper analysis of Dyslexia disorder may lead to wrong therapy strategy usage for correction of
Dyslexia and unsuccessful results. Hence articulation assessment and training system that is
therapist friendly (easy to handle and maintain) and which can support the therapists in analysis
and correction of Dyslexia (correction factors and suggestions given through visual cues which
gives the variations in the appearance of a graphic display which are intended to assist the therapist
in selecting the therapy strategy for Dyslexia correction more efficiently) is the imperious demands
of language therapist. Also, there is a need for automatic speech processing techniques that can be
used in a therapy system, which supports therapy sessions in the practice as well as tele-medical
therapy sessions. Thus, using such a system Dyslexia disorder can be treated easily.

The Solution
The proposed Dyslexia assessment and therapy system to help language therapists and children with

Dyslexia disorders is achieved using MATLAB, which is more reliable and serve as an easy-to-

apply diagnostic tool compared to other tools in the field of automatic speech processing. This

system
DEPT OF ECE, BGSIT BG Nagara Page 2
 Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

provides improved Dyslexia quality assessment by therapists, provides visual cues using which

they

can guide the children easily and hence it is a user-friendly tool for them. At an initial step of

achieving the proposed system the data is acquired by the children without any Dyslexia disorder in

the age group of 7-8 years using for different microphones (Mobile, Laptop, Headphone, SLM:

Sound Level Meter). Then best suitable microphone is selected based on its frequency response for

providing the input to the system. Then reference is set using the recordings from the best suitable

microphone. The data is acquired by the children with Dyslexia disorder in the age group of 7-9

years by using the best suitable microphone. Later the recordings are compared with that of the

reference and necessary instructions are provided if there is a need for articulation correction. These

instructions are provided using visual cues, which makes easy for the therapist to analyze and guide

the children. Thus, therapists can use this system to guide Dyslexia disordered children in clinical

therapy sessions. This can also be used as articulation assessment systems in schools.

DEPT OF ECE, BGSIT BG Nagara Page 3

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Existing System
For many years, the importance of early identification and intervention for children
with dyslexia has been stressed. Accordingly, much research has been directed towards
establishing precursors of dyslexia in the preschool years in international studies of children
at family risk of reading problems. Although significant progress has been made in
identifying cognitive skills that predict literacy outcomes on a group basis, at the individual
level, it is much harder to make accurate predictions. In this light, the merits of costly
screening and assessment procedures must be questioned.
Disadvantages

The present system has certain major disadvantages. A few to be listed below.
 No accurate result.
 Detection of disease is very complicate.

Proposed System
Initially system will be provided with the dataset, system will preprocess the dataset i.e.
removal of unwanted data and produce useful information. Once the data has been
preprocessed by the system, classification will be one on the data with the help of decision
tree algorithm. After the classification is done the prediction on the disease will be provided
as final result to the user.
Advantages
 Produce accurate data.
 Less expensive
 Time consumption is less.

DEPT OF ECE, BGSIT BG Nagara Page 4

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Software and Hardware Requirement Specifications

Software Requirements
MATLAB
MATLAB (MATrix LABoratory) is a tool used to achieve the mentioned aim and
objectives of our proposed work. It is a multi-paradigm numerical computing environment
which allows matrix manipulations, plotting of functions and data, implementation of
algorithms etc. MATLAB is a tool for technical computation and visualization in an
integrated environment, e.g.,
 Math and computation.
 Algorithm development.
 Data acquisition.
 Modelling, simulation, and prototyping.
 Data analysis, exploration, and visualization.
 Scientific and engineering graphics.
 Application development, including graphical user interface building.

Features of MATLAB are as follows:

 It is a high-level language for numerical computation, visualization and

application development.
 It also provides an interactive environment for iterative exploration, design
and problem-solving.
 It provides vast library of mathematical functions for linear algebra, statistics,
Fourier analysis, filtering, optimization, numerical integration and solving
ordinary differential equations.
 It provides built-in graphics for visualizing data and tools for creating custom plots.
 MATLAB's programming interface gives development tools for improving code
quality, maintainability, and maximizing performance.
 It provides tools for building applications with custom graphical interfaces.
 It provides functions for integrating MATLAB based algorithms with external applications
and languages such as C, Java, .NET and Microsoft Excel.

DEPT OF ECE, BGSIT BG Nagara Page 5

 Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Hardware Requirements

Laptop (Dell inspiron 1545)

Figure 1.1: Laptop Specifications

SLM: Sound Level Meter (SLM 2250)

Figure 1.2 :SLM Specifications

DEPT OF ECE, BGSIT BG Nagara Page 6

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Headset (Creative H320)

Figure 1.3: Headset Specifications

Mobile (Samsung galaxy J7 max)

Figure 1.4 : Mobile Specifications

DEPT OF ECE, BGSIT BG Nagara Page 7

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Chapter 2

LITERATURE SURVEY

Detection of Dyslexia In The Early Stage Using Machine Learning

➢ Authors: Shreeram M K, Varun R, Vishrutha K J , Sowjanya K V, Shruthi K S
➢ Published Year: 2022

➢ Proposed Idea: The main aim of our project is to encourage students to achieve one of
the primary milestone i.e,articulations of phonemes.To provide training in articulation of
phonemes by continual therapy imparted through MATLAB and to promote interactive
learning method during therapy sessions are the prime objectives.This Technical
intervention would be very user friendly for therapist during the therapy session as they
provide visual cues and also help giving feedback on improvement towards achieving the
correct articulation to the children. These visual cues will encourage articulation disordered
kid to learn in an easier way, unlike the traditional practice.Whole process was divided into
three phases. Initially the reliable and efficient microphone was selected for phoneme
recording purpose based on frequency response characteristics.Secondly, a data base of
normal was created using the recordings from the suitable microphone that was articulated
by the native Kannada speaking children with the age range of 7-8 years and without any
articulation disorder.Comparison of its frequency response characteristic with the reference
of normal children was obtained This visual feedback not only helps the therapist to
provide therapy in clinical sessions but also encourage the child in achieving the milestone
towards correct articulation very quickly in an easier way.

DEPT OF ECE, BGSIT BG Nagara Page 8

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Dyslexia Biomarker Detection using Machine Learning

 Authors:- Sangeetha S, Uma R ,Valarmathi R

 Published Year: 2021

 Proposed Idea: Dyslexia detection involves classification such as pre-processing,feature

extraction, and segmentation processes that are used to identify and classify the various
Dyslexia images. A machine learning approach is being used in this proposed system
which is used to detect and segment dyslexia in the brain region. The resulting outcome is
analyzed for tumor detection in terms of evaluation parameters such as; sensitivity,
specificity, and accuracy. Kaggle allows clients to discover and distribute informational
indexes, investigate and construct models in an electronic information science
environment, and other AI researchers and AI designs. By using programming tools like
MATLAB, the existing and the proposed technique are implemented through simulation
results, and the performance of the proposed technique is compared with theexisting
method to prove the effectiveness of the proposed algorithm.

DEPT OF ECE, BGSIT BG Nagara Page 9

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Detecting Dyslexia from Audio Records: An AI approach.

 Authors: Jim Radford, Gilles Richard, Hugo Richard and Mathieu Serrurier

 Published Year:2020

 Proposed Idea: Dyslexia impacts the individual’s ability to read, interferes with
academic achievements and may also have long term consequences beyond the learning
years.It is usually done via a lengthy battery of tests: human experts score these tests to
decide whether the child requires specific education strategies. This human assessment
can also lead to inconsistencies. That is why there is a strong need for earlier, simpler
screening of dyslexia. In this paper, we investigate the potential of modern Artificial
Intelligence in automating this screening. With this aim in mind and building upon
previous works, we have gathered a dataset of audio recordings, from both non-
dyslexicand dyslexic children. After proper preprocessing, we have applied diverse
machine learning algorithms in order to check if some hidden patterns are discoverable,
making adifference between dyslexic and non-dyslexic readers. Then, we built up our
own neuralnetwork which outperforms the other tested approaches. Our results suggests
the possibility to classify audio records as characteristic of dyslexia, leading to an
accurate and inexpensive dyslexia screening via non-invasive methods, potentially
reaching a large population for early intervention.

DEPT OF ECE, BGSIT BG Nagara Page 10

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Developmental Dyslexia Detection using Machine Learning

Techniques

 Authors:-Shahriar Kaiser
 Published Year:-29 February 2019
 Proposed Idea:-The detection of developmental dyslexia using machine learning
techniques uses four steps: (i) data collection,(ii) preprocessing, feature extraction and
feature selection,(iii) system training and classification, and (iv) performance evaluation.
The first step of dyslexia detection is to conduct a user study to collect user- data. For
conventional dyslexia detection techniques, psychologists examine behavioral aspects of
participants during standardized tests, such as reading and writing, phonological
awareness and working memory. Dyslexic individuals are identified basedon their poor
marks in those tests. In the second step collected data needs to be pre- processed and
filtered before being used in machine learning techniques. After feature selection, system
training and classification is conducted using machine learning algorithms. The dataset is
divided into training and testing parts. Existing literature usedMATLAB, WEKA, and
python based tools for performance evaluation. In this case, different metrics were used
for evaluating the performance of dyslexia detection techniques using machine learning
approaches.

DEPT OF ECE, BGSIT BG Nagara Page 11

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Chapter 3
OBJECTIVES AND EXPECTED OUTCOMES
Objectives of the Proposed System
The main objective of our project is to provide a tool that will be helpful for a therapist
while rendering services for a child with articulation disorder and to visually note the
correction factors for the children with articulation problem. The whole project can be
divided into three phases and it is as follows:

 Determination of microphone characteristics (frequency response), creating a

database of prerecorded phoneme articulated by the normal children (7-8 years)
through best suitable microphone (selected based on its frequency of operation)
and considering it to be the reference.
 Recording phonemes articulated by children with Dyslexia disorder of a 7-9 years
age group, comparing the speech characteristics with the reference obtained in
previous step.
 Visual cue and comparison will give a correction factor during Dyslexia of
phonemes. Thus, suggestions will be given to correct the articulation for children
with articulation disorder based on visual cues by the therapists.

Microphones

The proposed work involves different types of microphones. A microphone is a

transducer that converts acoustic into an electrical signal. Several different types of
microphone are in use, which employ different methods to convert the air pressure variations
of a sound wave to an electrical signal. Microphones are used in many applications such as
telephones, hearing aids, public address systems for concert halls and public events, motion
picture production, live and recorded audio engineering, sound recording, two-way radios,
megaphones, radio and television broadcasting, and in computers for recording voice, speech
recognition, VoIP, and for non-acoustic purposes such as ultrasonic sensors or knock
sensors. The most common microphone types are: Dynamic microphone, Condenser
microphone and Ribbon microphone.

DEPT OF ECE, BGSIT BG Nagara Page 12

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Dynamic microphone

Dynamic microphone elements are made up of a diaphragm, voice coil, and magnet which
form a sound-driven electrical generator. Sound waves move the diaphragm/voice coil in a
magnetic field to generate the electrical equivalent of the acoustic sound wave.

Figure 3.1 Dynamic microphone

In above figure 3.1 shows The signal from the dynamic element can be used directly,
without the need for additional circuitry. This design is extremely rugged, has good
sensitivity and can handle the loudest possible sound pressure levels without distortion. The
dynamic has some limitations at extreme high and low frequencies. To compensate, small
resonant chambers are often used to extend the frequency range of dynamic microphones.

Condenser microphone

Figure 3.2 Condenser microphone

DEPT OF ECE, BGSIT BG Nagara Page 13

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

In above figure 3.2 shows Condenser microphone elements use a conductive diaphragm and an
electrically charged back plate to form a sound-sensitive―condenser(capacitor). Sound
waves move the diaphragm in an electric field to create the electrical signal. In order to use this
signal from the element, all condensers have active electronic circuitry, (often referred to as
the―preamp) either built into the microphone or in a separate pack. This means that condenser
microphones require phantom power or a battery to operate. (For a detailed explanation
of―phantom power, see the sidebar.) However, the condenser design allows for smaller
microphone elements, higher sensitivity and is inherently capable of smooth response across a very
wide frequency range. These circuits can handle a specified maximum signal level from the
condenser element, so a condenser microphone has a maximum sound level before its output starts
to be distorted. Some condensers have switchable pads or attenuators between the element and the
electronics to allow them to handle higher sound level.

Ribbon microphone

Figure 3.3 Ribbon microphone

In above figure 3.3 shows Ribbon microphone elements, a variation of the dynamic microphone
operating principle, consist of a thin piece of metal, typically corrugated aluminum, suspended
between two magnetic pole pieces. As with moving-coil dynamics, no additional circuitry or
powering is necessary for operation, however, the output of ribbon microphones tends to be quite
low. Depending on the gain of the mixer or recording device to which the microphone is
connected, additional pre-amplification may be necessary. Note that ribbon microphones are not as
ruggedas moving-coil dynamic microphones. The ribbon element itself is typically no more than a
few microns thick, and can be deformed by a strong blast of air, or by blowing into the
microphone. Also, phantom power applied to the ribbon

DEPT OF ECE, BGSIT BG Nagara Page 14

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

microphone could be harmful. Ribbon microphones are highly regarded in studio recording for
their ―warmth and good low frequency response.
Microphone
SI No Output Ruggedness Sound Quality Throw
Types

1. Ribbon Poorest Poorest Best 2nd Best

2. Dynamic 2nd Best Best Poorest Best

3. Condenser Best Poor 2nd Close 2nd Poorest

Table no 3.4 Comparison between different microphones

The above table no 3.4 shows When choosing a microphone for recording one should consider the
characteristics of a microphone to get the best possible recording and should be carefully analyzed
before opting it in any application. Because some microphones are built for a purpose while others
are built for a more general use.

Expected Outcomes

Figure 3.5 Laptop ( Phoneme ―a‖ ) Figure 3.6 Mobile ( Phoneme ―a‖ )

Figure 3.7 Headset ( Phoneme ―a‖ ) Figure 3.8 SLM ( Phoneme ―a‖ )

DEPT OF ECE, BGSIT BG Nagara Page 15

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Chapter 4

METHODOLOGY

The proposed Dyslexia assessment and therapy system to help language therapists and
children with Dyslexia disorders is achieved using MATLAB, which is more reliable and serve as
an easy-to-apply diagnostic tool compared to other tools in the field of automatic speech
processing. This system provides improved Dyslexia quality assessment by therapists, provides
visual cues using which they can guide the children easily and hence it is a user-friendly tool for
them. At an initial step of achieving the proposed system the data is acquired by the children
without any Dyslexia disorder in the age group of 7-8 years using for different microphones
(Mobile, Laptop, Headphone, SLM: Sound Level Meter). Then best suitable microphone is
selected based on its frequency response for providing the input to the system. Then reference is set
using the recordings from the best suitable microphone. The data is acquired by the children with
Dyslexia disorder in the age group of 7-9 years by using the best suitable microphone. Later the
recordings are compared with that of the reference and necessary instructions are provided if there
is a need for articulation correction. These instructions are provided using visual cues, which
makes easy for the therapist to analyze and guide the children. Thus, therapists can use this system
to guide Dyslexia disordered children in clinical therapy sessions. This can also be used as
articulation assessment systems in schools.

Algorithms used in project

Cepstrum algorithm

Cepstrum method is a widely used algorithm. Cepstrum of a signal is obtained by

taking the Inverse Fourier transform (IFT) of the logarithm of the spectrum of that signal.
Mathematically it is given by the equation (1).

c[n] = 𝐹−1{log(|F{x[n]}|)} ………(1)

Where, x[n] is the sampled speech signal, F indicates its Fourier transform, and c[n] are the
Cepstrum coefficients. Cepstrum method reads a 40ms window segment of the down
sampled voice signal using 'audio read' function in MATLAB. Then the signal is multiplied
by a hamming window. Fast Fourier Transformation (FFT) of this windowed frame

DEPT OF ECE, BGSIT BG Nagara Page 16

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

gives the

DEPT OF ECE, BGSIT BG Nagara Page 17

 Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

spectrum of the speech signal in frequency domain. Taking Inverse Fourier Transformation.

(IFT) of logarithm of the spectrum gives Cepstrum in quefrency domain. Once in the quefrency

domain, the pitch can be estimated by determining the peak of the Cepstrum which
represents pitch lag. The lag at which there is the most energy represents the dominant
frequency in the log spectrum and thereby it gives the pitch frequency. A flow diagram of
Cepstrum method for pitch detection is shown in Figure 4.1.

speech x[n] X[k] log(X[k]) c[n]

DFT log IDFT

Figure 4.1 Flow diagram for Pitch detection using Cepstrum method

Once in the cepstral domain, the pitch can be estimated by picking the peak of the
resulting signal within a certain range. The Cepstrum is given in term of ―quefrency‖
which, besides being a terrible name, represents pitch lag. Therefore, the lag at which there is
the most energy represents the dominant frequency in the log magnitude spectrum thereby
giving you the pitch. There are of course some caveats to this approach. First of all, pitch and
fundamental frequency are not actually the same thing, so depending on which peak your
algorithm picks, you may be getting F0 (the fundamental) of FI(one of the formants).
Secondly, the Cepstrum is time shift variant. Therefore, you cannot just apply this method
blindly. Instead, you need to precisely line up your time domain windows such that they start
and stop exactly over a voiced speech segment. This is not a trivial task as most VADs often
have errors and thus your Cepstrum will suffer from phase ambiguity. To get around this
problem, we can use the differential Cepstrum and its variants such as the mean differential
Cepstrum. This method is widely used and represents an important step in understanding the
usefulness of this second Fourier domain.

Cepstrum method is used to determine the fundamental frequency of human speech

and to analyze radar signal returns. Cepstrum pitch determination is particularly effective
because the effects of the vocal excitation (pitch) and vocal tract (formants) are additive in
the logarithm of the power spectrum and thus clearly separate. The Cepstrum is a
representation used in homomorphic signal processing, to convert signals combined by
convolution (such as a source and filter) into sums of their cepstrum, for linear separation.
DEPT OF ECE, BGSIT BG Nagara Page 18
Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Decision tree

Decision Trees are a type of Supervised Machine Learning where the data is
continuously split according to a certain parameter. The tree can be explained by two entities,
namely decision nodes and leaves. The leaves are the decisions or the final outcomes. And the
decision nodes are where the data is split.
Decision Tree algorithm belongs to the family of supervised learning algorithms. Unlike other
supervised learning algorithms, decision tree algorithm can be used for solving regression and
classification problems.

Step1: Place the best attribute of the dataset at the root of the tree. To that we have calculate
Entropy
Entropy, also called as Shannon Entropy is denoted by H(S) for a finite set S, is the measure of the
amount of uncertainty or randomness in data.
Step2: Split the training set into subsets. Subsets should be made in such a way that each subset
contains data with the same value for an attribute.ie Make that attribute a decision node and breaks
the dataset into smaller subsets by considering the trained dataset.
Now the next step is to choose the attribute that gives us highest possible Information Gain
Step3: Use the classified values from previous algorithm and predict the intrusion type. Repeat
step 1and step 2 on each subset until you find leaf nodes in all the branches of the tree.

DEPT OF ECE, BGSIT BG Nagara Page 19

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Select K Algorithm

Feature selection is a technique where we choose those features in our data that contribute
most to the target variable. In other word we choose the best predictors for the target variable. The
classes in the learn. feature selection module can be used for feature selection/dimensionality
reduction on sample sets, either to improve estimators’ accuracy scores or to boost their
performance on very high-dimensional datasets.

Step-1: Select the number K to decide the number of clusters.

Step-2: Select random K points or centroids. (It can be other from the input dataset).

Step-3: Assign each data point to their closest centroid, which will form the predefined K clusters.

Step-4: Calculate the variance and place a new centroid of each cluster.

Step-5: Repeat the third steps, which means reassign each datapoint to the new closest centroid of
each cluster.

Step-6: If any reassignment occurs, then go to step-4 else go to FINISH.

Step-7: The model is ready.

DEPT OF ECE, BGSIT BG Nagara Page 20

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Chapter 5

SYSTEM ANALYSIS AND DESIGN

System Analysis

Block Diagram

MICROPHONE
MOBILE
MATLAB CHARACTERISTICS

LAPTOP BEST SUITABLE

Cepstrum MICROPHONE
HEADSET Algorithm
COMPARISION AND
SLM CORRECTION OF
ARTICULATION

Figure 5.1 Generalized Block Diagram

The above Figure 5.1 shows proposed idea is to determine suitable microphone for recording
different phonemes. Analyzing microphone characteristics and comparing the recorded
phonemes from the articulation disordered children (7-9 years) with that of the normal
children (7-8 years). After acquisition and comparison of data, the necessary instructions are
displayed with the help of GUI (Graphic User Interface), where in the therapists will analyze
and guide the disordered children to correct their articulation. Data acquisition is performed
using various microphones(Ex: Mobile, Laptop, Headset, SLM: Sound Level Meter).
PRAAT tool is used for the editing of speech samples and MATLAB is used for analysis and
comparison of phonemes. The results are microphone characteristics (Frequency response),
best suitable microphone for recording, comparison and correction of articulation.

The Generalized Block diagram which describes the detailed procedure used to record,
analyze and interpret the comparative measures required for the articulation disorder therapy
purpose

DEPT OF ECE, BGSIT BG Nagara Page 21

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

High Level Design

System Architecture

Three-tier (layer) is a client-server architecture in which the user interface, business

process (business rules) and data storage and data access are developed and maintained as
independent modules or most often on separate platforms.
The Architecture of Application is based on three-tier architecture. The three logical tiers are:

 Presentation tier - HTML Forms, Images.

 Middle tier – Python classes.
 Data tier- SQL
The main reason for considering three-tier architecture for the Application is as follows:
Flexibility
• Management of data is independent from the physical storage support,
• Maintenance of the business logic is easier,
• Migration to new graphical environments is faster
• If there is a minor change in the business logic, we don’t have to install the entire
system in individual user’s PCs.
Reusability
 Reusability of business logic is greater for the presentation layer. As this component
is developed and tested, we can use it in any other project and would be helpful for
future use.
Security
 More secured architecture since the client cannot access the database directly.

Tier Architecture
A 3-tier architecture is a type of software architecture which is composed of three
“tiers” or “layers” of logical computing. They are often used in applications as a specific type
of client- server system. 3-tier architectures provide many benefits for production and
development environments by modularizing the user interface, business logic, and data
storage layers. Doing so gives greater flexibility to development teams by allowing them to
update a specific part of an application independently of the other parts. This added flexibility
can improve overall time- to-market and decrease development cycle times by giving
development teams the ability to replace or upgrade independent tiers without affecting the
other parts of the system.

DEPT OF ECE, BGSIT BG Nagara Page 22

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Data Flow Diagram

Dataset Input

Dyslexia

Detection

Figure 5.3: Level 0 dfd

In the above Figure 5.4 shows level 0 DFD, at first dataset will be gathered and this dataset
will be given as input to the system and once the dataset has been loaded system will give the
disease detection as output.

Figure 5.4: User level 1 dfd

In the above Figure 5.5 shows level 1 DFD, at the start of the system user will gather the
dataset and give the dataset as input to the system once the data has been given system will
preprocess the data andperform segmentation and based on that cluster will be created after
this feature comparison will take place and in the end prediction about the dyslexia will be
given to the user.

DEPT OF ECE, BGSIT BG Nagara Page 23

 Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

Flow Chart
The flowchart of the process is depicted in the Figure 5.6. The inputs to the program are

audio file that contains the reference phoneme and the audio file that contains the phoneme

articulatedby the child with articulation disorder. After taking the inputs, the frequency response is

determined for both the inputs. Then the frequency response of both the inputs is compared.

Based on the comparison results the instructions are provided using which the therapistcan guide the

child to correct the articulation if correction is necessary. The conditions and their respective output

are: If the frequency response of phoneme articulated by child with articulation disorder is equal to

the reference phoneme, then audio output: Excellent is given. If not equal, then one more condition

is applied, which checks whether the frequency response of the articulation disordered child is within

the range. If it is within the range then audio output: Partially correct is given else audio output: ―Poor

articulation, try again is given.

DEPT OF ECE, BGSIT BG Nagara Page 24

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

START

Equal? Yes
Reference phoneme audio file Articulation disordered child‘s
as input phoneme audio file as input

No
Determine the frequency response Determine the frequency response

No

Yes

articulation, try

STOP

Figure5.5: Flow chart

DEPT OF ECE, BGSIT BG Nagara Page 25

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

CONCLUSION

As the proposed idea was to determine suitable microphone for recording different
phonemes, SLM is selected as the best suitable microphone. It is because of its better
frequency response, good sound quality and less noise. One can also choose headset for their
work as it is cost effective compared to SLM. One more objective of our work was to
compare the recorded speech from children suffering from articulation disorder with the
normal. On comparison the necessary instructions will be given to correct the articulation.
The audio output along with the message box is used to instruct the children and this helps
the therapist by making their job easier. Thus the results satisfied the objectives of our work.
This work will be helpful for the researcher in choosing the right microphone for recording
and further studies. It can be used as a therapeutic tool by providing visual cues that helps in
getting attention of the child while giving therapy. It provides improved articulation quality
assessment. Since the tools used for extraction of sample and analysis of articulation are
PRAAT and MATLAB respectively. It also helps to store data in the system that can be used
in further case study. Hence it is a user-friendly tool for therapists.

DEPT OF ECE, BGSIT BG Nagara Page 26

Detection Of Dyslexia in Early Stage Using Machine Learning 2022-23

REFERENCES
[1] Masooda Modak, Ketan Ghotane, Siddhanth V, Nachiket Kelkar, Aravind Iyer,
PrachiG,” Detection of Dyslexia using Eye Tracking Measures”, International Journal of
Innovative Technology and Exploring Engineering (IJITEE) ISSN: 2278-3075, Volume-8,
Issue - 6S4, April 2019.
[2] Luz Rello and Miguel Ballesteros, “Detecting Readers with Dyslexia Using Machine
Learning with Eye Tracking Measures “, w4A ‘18, Article No. 16,
Doi/10.1145/2745555.2746644, association of computing machinery publication,2018
[3] Alex Frid Edmond J. and ZviaBreznitz Edmond J. ,“An SVM Based Algorithm for
Anal and Discrimination of Dyslexic Readers from Regular Readers Using ERPs “,2017
IEEE27 Convention of Electrical and Electronic s Engineers in Israel,2017
[4] Anuradha Kar AND Peter Corcoran, “A Review and Analysis of Eye-Gaze Estimation
Systems, Algorithms and Performance Evaluation Methods in Consumer Platforms”, IEEE
access. Doi 10.1109/access.2017.2735633,2019.
[5] Maria M. Suarez-Alvarez, Duc-Truong Pham ,Mikhail Y. Prostov and Yuriy I. Prostov
, “Statistical approach to normalization of feature vectors and clustering of mixed datasets
“, proceedings of the Royal society ,2016
[6] RehmanUllahKhan1,Julia Lee Ai Cheng, Oon Yin Bee, Machine Learning and
Dyslexia Diagnostic and Classification System (DCS) for Kids with Learning Disabilities”,
Internati Journal of Engineering & Technology ,2016.
[7] Khaled Tawfik, ―Towards The Development of Computer Aided Speech Therapy
Tool in Arabic Language Using Artificial Intelligence‖, Department of Computing &
Information Systems, Cardiff School of Management, Cardiff Metropolitan University,
April 2016.
[8] Hung-Yu Su, Chun-Hsien Wu and Pei-Jen Tsai, ―Automatic Assessment of
articulation disorders using Confident Unit-Based Model Adaptation‖,1-4244-
1484- 9/08/$25.00©2008IEEE, ICASSP 2008.
[9] Yeou-Jiunn Chen, Jing- Wei Huang, Hui-Mei Yang, Yi-Hui Lin and Jiunn-Liang
W―Development of Articulation Assessment and training system with Speech Recognition
and Articulation training strategies selection‖, 1-4244-0728-1/07/$20.00 ©2007 IEEE,
ICASSP 2007.
[10] Yeou-Jiunn Chen, Chung-Hsien Wu, Jiunn-Liang Wu, Hui-Mei Yang, Chih-Chang Chen
and Shan-Shan Ju, ―An Articulation Training System with Intelligent Interface
and Multimode Feedbacks to Articulation Disorders‖, 2009 International Conference.
DEPT OF ECE, BGSIT BG Nagara Page 27
DEPT OF CSE, BGSIT Page. 40