Agarwal, Ankur

The major objective of this dissertation was to create a framework which is used for medical image diagnosis. In this diagnosis, we brought classification and diagnosing of diseases through an Artificial Intelligence based framework, including COVID, Pneumonia, and Melanoma cancer through medical images. The algorithm ran on multiple datasets. A model was developed which detected the medical images through changing hyper-parameters.
The aim of this work was to apply the new transfer learning framework DenseNet-201 for the diagnosis of the diseases and compare the results with the other deep learning models. The novelty in the proposed work was modifying the Dense Net 201 Algorithm, changing hyper parameters (source weights, Batch Size, Epochs, Architecture (number of neurons in hidden layer), learning rate and optimizer) to quantify the results. The novelty also included the training of the model by quantifying weights and in order to get more accuracy. During the data selection process, the data were cleaned, removing all the outliers. Data augmentation was used for the novel architecture to overcome overfitting and hence not producing false absurd results the computational performance was also observed. The proposed model results were also compared with the existing deep learning models and the algorithm was also tested on multiple datasets.

Hospital readmission rates are considered to be an important indicator of quality of care because they may be a consequence of actions of commission or omission made during the initial hospitalization of the patient, or as a consequence of poorly managed transition of the patient back into the community. The negative impact on patient quality of life and huge burden on healthcare system have made reducing hospital readmissions a central goal of healthcare delivery and payment reform efforts.
In this study, we will be proposing a framework on how the readmission analysis and other healthcare models could be deployed in real world and a Machine learning based solution which uses patients discharge summaries as a dataset to train and test the machine learning model created. Current systems does not take into consideration one of the very important aspect of solving readmission problem by taking Big data into consideration. This study also takes into consideration Big data aspect of solutions which can be deployed in the field for real world use. We have used HPCC compute platform which provides distributed parallel programming platform to create, run and manage applications which involves large amount of data. We have also proposed some feature engineering and data balancing techniques which have shown to greatly enhance the machine learning model performance. This was achieved by reducing the dimensionality in the data and fixing the imbalance in the dataset.
The system presented in this study provides a real world machine learning based predictive modeling for reducing readmissions which could be templatized for other diseases.

Dialysis patients are operated to have AV Fistula which is a joint junction of an artery and vein in the arm, operated to increase the blood flow through the dialyzer machine. AV- fistula is a type of vascular access which is a path into the body to connect/disconnect devices, but in this case, it is mainly Dialyzer. To reduce the failure rate during maturation period of AV Fistula, doctors recommend squeezing ball exercise as a necessary precaution for AV Fistula failure. Doing Squeezable interaction for about 3-4 times a day is recommended based on patient’s health condition. Hence, the proposed architecture adopts this squeezable exercise by embedding with sensor and measuring the angle at which the sensor is bent. The framework also proposes a new care coordination system having the hardware layer which has key components such as raspberry Pi, sensor which help in recording the pressure values when user presses the ball and software layer which solely focuses on data sync among the applications used by the user. It has been recorded that 53 % of patients having AV-Fistula fail because of negligence and lack of care. The maturation period is so critical and important which made us to build a gamification platform to monitor the exercise and track the activity through android application to keep users motivated and disciplined. In further chapters of the study will focus on different clinical like procedure around AV-Fistula and technical information such as different technologies used and implemented in the proposed system along with sensor circuit. This project goal is to present a way of monitoring patients and to keep track of the compliance whether the patient is active doing exercise daily. This way we are trying to present a care monitoring system for patients to help prevent AV Fistula failure.

With the increasing complexity of the system design, it has become very critical to
enhance system design productivity to meet with the time-to-market demands. Real Time
embedded system designers are facing extreme challenges in underlying architectural
design selection. It involves the selection of a programmable, concurrent, heterogeneous
multiprocessor architecture platform. Such a multiprocessor system on chip (MPSoC)
platform has set new innovative trends for the real-time systems and system on Chip
(SoC) designers. The consequences of this trend imply the shift in concern from
computation and sequential algorithms to modeling concurrency, synchronization and
communication in every aspect of hardware and software co-design and development.
Some of the main problems in the current deep sub-micron technologies characterized by
gate lengths in the range of 60-90 nm arise from non scalable wire delays, errors in signal
integrity and un-synchronized communication. These problems have been addressed by
the use of packet switched Network on Chip (NOC) architecture for future SoCs and
thus, real-time systems. Such a NOC based system should be able to support different levels of quality of service (QoS) to meet the real time systems requirements. It will
further help in enhancing the system productivity by providing a reusable communication
backbone. Thus, it becomes extremely critical to properly design a communication
backbone (CommB) for NOC. Along with offering different levels of QoS, CommB is
responsible directing the flow of data from one node to another node through routers,
allocators, switches, queues and links. In this dissertation I present a reusable component
based, design of CommB, suitable for embedded applications, which supports three types
of QoS (real-time, multi-media and control applications).

Recent federal legislation has incentivized hospitals to focus on quality of patient
care. A primary metric of care quality is patient readmissions. Many methods exist to
statistically identify patients most likely to require hospital readmission. Correct
identification of high-risk patients allows hospitals to intelligently utilize limited resources
in mitigating hospital readmissions. However, these methods have seen little practical
adoption in the clinical setting. This research attempts to identify the many open research
questions that have impeded widespread adoption of predictive hospital readmission
systems.
Current systems often rely on structured data extracted from health records systems.
This data can be expensive and time consuming to extract. Unstructured clinical notes are
agnostic to the underlying records system and would decouple the predictive analytics
system from the underlying records system. However, additional concerns in clinical
natural language processing must be addressed before such a system can be implemented. Current systems often perform poorly using standard statistical measures.
Misclassification cost of patient readmissions has yet to be addressed and there currently
exists a gap between current readmission system evaluation metrics and those most
appropriate in the clinical setting. Additionally, data availability for localized model
creation has yet to be addressed by the research community. Large research hospitals may
have sufficient data to build models, but many others do not. Simply combining data from
many hospitals often results in a model which performs worse than using data from a single
hospital.
Current systems often produce a binary readmission classification. However,
patients are often readmitted for differing reasons than index admission. There exists little
research into predicting primary cause of readmission. Furthermore, co-occurring evidence
discovery of clinical terms with primary diagnosis has seen only simplistic methods
applied.
This research addresses these concerns to increase adoption of predictive hospital
readmission systems.

Identifying and tracking individuals affected by this virus in densely
populated areas is a unique and an urgent challenge in the public health sector.
Currently, mapping the spread of the Ebola virus is done manually, however with
the help of social contact networks we can model dynamic graphs and predictive
diffusion models of Ebola virus based on the impact on either a specific person or
a specific community.
With the help of this model, we can make more precise forward
predictions of the disease propagations and to identify possibly infected
individuals which will help perform trace – back analysis to locate the possible
source of infection for a social group. This model will visualize and identify the
families and tightly connected social groups who have had contact with an Ebola
patient and is a proactive approach to reduce the risk of exposure of Ebola
spread within a community or geographic location.

This paper evaluates the effectiveness of a wearable device, developed by the
author, to detect different types of epileptic seizures and monitor epileptic patients. The
device uses GSR, Pulse, EMG, body temperature and 3-axis accelerometer sensors to
detect epilepsy. The device first learns the signal patterns of the epileptic patient in ideal
condition. The signal pattern generated during the epileptic seizure, which are distinct from
other signal patterns, are detected and analyzed by the algorithms developed by the author.
Based on an analysis, the device successfully detected different types of epileptic seizures.
The author conducted an experiment on himself to determine the effectiveness of the device
and the algorithms. Based on the simulation results, the algorithms are 100 percent accurate
in detecting different types of epileptic seizures.

Chronic Diseases are the major cause of mortality around the world, accounting for 7 out of 10 deaths each year in the United States. Because of its adverse effect on the quality of life, it has become a major problem globally. Health care costs involved in managing these diseases are also very high. In this thesis, we will focus on two major chronic diseases Asthma and Diabetes, which are among the leading causes of mortality around the globe. It involves design and development of a predictive analytics based decision support system which uses five supervised machine learning algorithm to predict the occurrence of Asthma and Diabetes. This system helps in controlling the disease well in advance by selecting its best indicators and providing necessary feedback.

Wellness and healthy life are the most common concerns for an individual to lead a happy life. A web-based approach known as Wellness Scoring is being developed taking into people’s concerns for their health issues. In this approach, four different classifiers are being investigated to predict the wellness. In this thesis, we investigated four different classifiers (a probabilistic graphical model, simple probabilistic classifier, probabilistic statistical classification and an artificial neural network) to predict the wellness outcome. An approach to calculate wellness score is also addressed. All these classifiers are trained on real data, hence giving more accurate results. With this solution, there is a better way of keeping track of an individuals’ health. In this thesis, we present the design and development of such a system and evaluate the performance of the classifiers and design considerations to maximize the end user experience with the application. A user experience model capable of predicting the wellness score for a given set of risk factors is developed.

Hospital readmission rates are considered to be an important indicator of quality of care
because they may be a consequence of actions of commission or omission made during
the initial hospitalization of the patient, or as a consequence of poorly managed transition
of the patient back into the community. The negative impact on patient quality of life and
huge burden on healthcare system have made reducing hospital readmissions a central
goal of healthcare delivery and payment reform efforts.
In this project, we will focus on COPD (Chronic Obstructive Pulmonary Disease) which
is one of the leading causes of disability and mortality worldwide. This project will
design and develop a prognostic COPD healthcare management system which is a
sustainable clinical decision-support system to reduce the number of readmissions by
identifying those patients who need preventive interventions to reduce the probability of
being readmitted. Based on patient’s clinical records and discharge summary, our system would be able to determine the readmission risk profile of patients treated for COPD. Suitable
interventions could then be initiated with the objective of providing quality and timely
care that helps prevent avoidable readmission.