NAME: MUKUDZEISHE MARIANO

REG NUMBER: R211793B

PROGRAM: HBIOENG
COURSE: BIOMEDICAL SIGNAL PROCESSING
LECTURER: Mr Zhou

ASSIGNMENT 1
Neural networks originated from research through the
understanding of how the human brain works. The first artificial
neural networks were developed in the 1940s and 1950s and were
based on the concept of perceptrons, which were modelled on the
human brain's neurons. In 1980s when advancements in computing
power and parallel processing made more complex neural networks
practical. Since then, the field of neural networks has continued to
advance rapidly, with new architectures, algorithms, and
applications being developed. Artificial Neural Network (ANN) is a
signal processing system composed of a large number of simple
processing elements which are interconnected by direct links and
which cooperate to perform parallel distributed processing in order
to solve a desired computational task. Neural networks process
information in a similar way the human brain does. ANN is inspired
by the way the biological nervous systems. A neural network consist
of interconnected group of nodes, inspired by a simplification of
neurons in the brain. In the diagram on figure 1 each circular node
represent an artificial neuron an arrow represents a connection from
the output of one artificial neuron to the input of another.

Figure 1
A Neural network receives signals then processes them and can
pass the signal to the neurons connected to it. The "signal" at a
connection is a real number, and the output of each neuron is
computed by some non-linear function of the sum of its inputs. The
connections are called edges. Neurons and edges typically have a
weight that adjusts as learning proceeds. The weight increases or
decreases the strength of the signal at a connection. Neurons may
have a threshold such that a signal is sent only if the aggregate
signal crosses that threshold. Typically, neurons are aggregated into
layers. Different layers may perform different transformations on
their inputs. Signals travel from the first layer (the input layer), to
the last layer (the output layer), possibly after traversing the layers
multiple times. Neural networks are also now being used in many
applications such as speech recognition, cardiology, and neurology
and also used in sensory waveform.
SPEECH RECOGNITION
The field of neural networks has grown rapidly in recent years.
Nowdays neural networks are now being used in speech recognition.
In speech language recognition, neural networks are used to
accurately identify and transcribe speech from audio recordings.
This is achieved by training the neural network on a large dataset of
pre-labeled speech data, and the network learns to identify the
different speech sounds, patterns and structures. The trained model
can then be used to transcribe audio in real-time, which has
applications in areas such as speech-to-text tools, voice assistants,
and automated transcription services. Most speech recognition
research has centered on stochastic models, in particular the use of
hidden Markov models (HMMs) (Deller J.R.Jr, Proakis J.G, Hansen J.H
1993). Neural networks perform very well at learning phoneme
probability from highly parallel audio input, while Markov models
can use the phoneme observation probabilities that neural networks
provide to produce the likeliest phoneme sequence or word. This is
at the core of a hybrid approach to natural language understanding.
The idea for using neural networks as a classifier stems comes from
the fact that neural networks within the human brain are used for
speech recognition. Neural networks are used in aspects of speech
recognition such as phoneme classification, isolated word
recognition, audiovisual speech recognition, audiovisual Markov
models (HMMs) (Deller J.R.Jr, Proakis J.G, Hansen J.H 1993.

Cardiology
Neural networks can be used to analyse ECG signals and identify
underlying abnormalities. The network can be trained on a large
dataset of labelled ECG signals to learn different features and
patterns of normal and abnormal signals. This enables the model to
accurately classify and diagnose different types of arrhythmias,
such as atrial fibrillation or ventricular tachycardia. This allows for
earlier detection, improved treatment plans and better patient
outcomes. Neural networks are capable of identifying relations in
input data that are not easily apparent with current common
analytic techniques. In cardiology neural networks are used in
diagnosis and treatment of coronary artery diseases and myocardial
infarction thereby neural networks have a strong impact on
cardiovascular medicine. Neural networks will drive improved
patient care because physicians will be able to interpret more data
in greater depth than ever before. Reinforcement learning
algorithms will become companion physician aids, unobtrusively
assisting physicians and streamlining clinical care. Advances in
unsupervised learning will enable far greater characterization of
patients’ disorders and ultimately lead to better treatment selection
and improved outcomes. Artificial neural networks have now being
used to identify patients with cardiac chest pain and more
specifically those with acute myocardial infarction.
Ectrocardiography). Artificial neural network technology is now
being used for automated analysis and interpretation of ECGs
(Bortolan et al. 2017). Cardiac image analysis. Artificial neural
networks can be used in the fields of cardiac image processing and
interpretation. They are being used to process images, to facilitate
complex pattern recognition, which will aid classification of images
into clinically relevant categories.

Neurology

Neural networks are used to model and understand the complex

interactions between neurons in the brain. Researchers can use
these models to simulate the behaviour of neurons, study the
effects of drugs or disease, and develop new treatments for
neurological disorders. Radiological image classification and
segmentation can be performed in help of neural networks efforts.
Image classification and segmentation tasks can be perfumes by
deep learning due to the high-dimensional nature of neuroimaging
data which is unfavourable to manual analysis. Deep learning has
been applied to functional brain mapping and correlational studies
using functional magnetic resonance imaging data for tasks such as
prediction of postoperative seizure. Diagnostic prognostication with
neural networks using multiple data types, including lab values,
images, notes, has been used to assign disease risk. The application
of deep learning in medicine also involves the analysis of imaging
modalities, especially those for the detection of Alzheimer’s disease
(AD) and other cognitive impairments. The application of deep
learning to neuroimaging big data will help develop computer-aided
diagnosis of neurological diseases. Pattern recognition using deep
learning can extract features of neuroimaging signals unique to
various neurological diseases, leading to better diagnoses.
Computer-aided diagnosis is crucial to improve treatment strategies
for neurological diseases

Sensory waveform

In sensory waveform analysis, neural networks are used to analyze

waveforms generated by sensors that measure various physical
phenomena such as sound, light, or pressure. By analyzing these
waveforms, neural networks can recognize patterns and identify
specific features that correspond to particular events or conditions.
This technique is widely used in the field of signal processing for
applications such as fault detection and condition monitoring.
Sensory waveform analysis is used in a wide range of applications,
including audio signal processing, image processing, and sensor
data analysis. For example, neural networks can be used to classify
different types of sounds or images, detect faults in machines or
vehicles, or monitor environmental conditions such as temperature
and humidity. Neural networks are particularly useful in these
applications because they can learn to recognize complex patterns
and features that are difficult or impossible for humans to detect.

References

 Hinton, G.; Deng, L.; Yu, D.; Dahl, G.E.; Mohamed, A.R.;
Jaitly, N.; Senior, A.; Vanhoucke, V.; Nguyen, P.;
Sainath, T.N.; et al. Deep neural networks for acoustic
modeling in speech recognition: The shared views of
four research groups. IEEE Signal Process. Mag. 2012,
29, 82–97. [Google Scholar] [CrossRef]

 Yu, D.; Li, J. Recent progresses in deep learning based

acoustic models. IEEE/CAA J. Autom. Sin. 2017, 4, 396–
409. [Google Scholar] [CrossRef