Conference PaperPDF Available

Architecture of Smart Health Care System Using Artificial Intelligence

Authors:
ARCHITECTURE OF SMART HEALTH CARE SYSTEM USING ARTIFICIAL
INTELLIGENCE
M.M.Kamruzzaman
Department of Computer and Information Science, Jouf University, Sakaka, Al-Jouf, KSA
mmkamruzzaman@ju.edu.sa
ABSTRACT
Artificial intelligence is becoming increasingly
useful for doctors, nurses, radiologists,
researchers, pharmacists, emergency medical
service, and many other healthcare professionals.
This paper proposes the creation of a smart
healthcare system using artificial intelligence as a
means of efficiently solving challenges in the
healthcare industry and as a tool for optimizing
patient care plans. The proposed AI-assisted
system shows that it can support a patient who is
admitted to the hospital through emergency
medical services, easily process the patient’s data,
and offer early detection of serious diseases. It can
automatically recognize the complicated patterns
which have been obtained from radiologists, can
analyze complete human molecular data and
genetics in the clinic, and can support doctors by
producing AI-generated radiologist reports,
clinical laboratory reports, and many other
decision-support tools. The proposed architecture
can easily handle diverse and complicated
healthcare problems and can be used by any
modern hospital to save time and money. This
work also shows the recent development of AI
applications in healthcare, which could be used in
the proposed architecture.
Index Terms— Artificial intelligence, smart
healthcare, Machine learning, Deep learning.
1. INTRODUCTION
In the healthcare market, Artificial intelligence (AI) was
valued at about $1441 million in 2016, and by 2023, it is
estimated that it will reach approximately $22,700 million.
AI plays a role in gathering information, then it processes
that information and makes a perfect prediction by using
algorithms that have been tested repeatedly to diminish the
margin of error. The important AI technologies in this
context include the natural processing of languages, physical
robotic systems, and machine learning, along with deep
learning and neural networks. The main aim of AI in
healthcare is to analyze the important connections between
patient outcomes and treatment methods for prevention [1-
5].
Different AI algorithms have been created for different
applications. [6] proposes using target recognition and
image interpretation, [7] suggests using a remote sensing
image retrieval algorithm based on an improved Sobel
operator, [8] indicates large-scale, high-dimensional data
processing for images of the human brain in order to
establish a hierarchical model of hidden relational logic,
detecting the criminals of future smart cities using AI is
explained in [9], video stream based on object detection is
proposed in [10], and [11] shows how to extract images of
cultivated land using deep learning. [12] uses deep learning
for diagnosis fault from images. A hash network algorithm
based on deep neural networks is proposed in [13] to
retrieve animal images from a massive network of images.
[14] proposes a moving target tracking algorithm based on
block information.
AI has also been applied in different fields of smart health
services, such as robotic surgery, cardiology, cancer
treatment, and neurology [16]. Drug development has been
addressed, and so has patient monitoring and personalized
medication, as well as helping doctors make perfect
decisions [17], finding related medical data or information
from different textbooks and journals [18], storing patient
data on the Cloud for easy access, and so on.
The purpose of this paper is to provide an
architecture for an AI-based smart healthcare system where
patients will get complete support throughout the course of
978-1-7281-1485-9/20/$31.00 (c) 2020 IEEE
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
their lives. This AI-based support will come from all
departments, including Emergency Medical Service (EMS),
nursing, doctors, radiologists, clinical laboratories,
pharmacy, and so on. Researchers will also be able to use
patients’ data and AI-based decision-making for further
development in the health care system. This proposed
system will also explain how to deploy the recent
developments in AI to automate the whole healthcare
system, which is unprecedented in existing research.
2. ARCHITECTURE OF SMART HEALTH SYSTEM
USING ARTIFICIAL INTELLIGENCE
Figure 1 shows the proposed architecture for AI-
supported smart healthcare systems, where AI supports
doctors, patients, nurses, Emergency Medical Service
(EMS), pharmacists, radiologists, clinical laboratories, and
researchers. The following sections explain how those actors
are receiving that support:
2.1 Patient:
Hospitals that use this architecture for smart
healthcare systems will provide a sensor-based wireless
device to track each of its listed patients. That device,
whether placed at home or in the hospital, will automatically
update the patient’s daily information in a centralized
database system. AI will make decisions based on the
patient’s data, and Emergency Medical Service (EMS) will
be provided if required.
The most important role AI will play for patients is that
it provides all the pertinent patient data in a timely manner
and properly screens the patients’ affected part. AI advises
the patient and guides the physician in giving proper
treatment. These AI-based support systems are sometimes
smarter than a physician and can give a proper diagnosis and
can treat patients with serious conditions more effectively
than a doctor. This AI-based decision support system is
under a process of continuous improvement based on
several factors such as new patient data, rate of cure, et
cetera. The medical research center will always maintain the
system, which will be explained later.
Fig.1 Architecture of smart health system using artificial intelligence
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
2.2 Emergency Medical Service (EMS):
Sensor-based devices that are always with the patients’
hands can be useful for EMS. AI assists EMS in decision-
making as well as providing emergency treatment for
serious patients, like stroke patients. AI-based algorithmic
computers ask for several inputs like temperature, blood
pressure, and so on, to monitor the patients’ status and
provide fast aid for a short time. The AI plays a role in
drawing blood from the patient, and also provides directions
to nurses. Nurses update the patients’ status with the help of
AI. It also helps direct patients to the designated hospital for
a particular disease.
2.3 Nurses
Nurses can easily handle massive amounts of patient
data with the help of smart AI-based devices without having
to enter the data manually. Nurses are always updated from
the AI-based system, which receives the patients’ data from
their smart devices. AI also assists nurses in evaluating
disease status and helps them to anticipate future
interventions.
2.4 Doctors:
AI makes everything easier and faster by monitoring
and screening the patient as well as providing support for
making decisions quickly. AI can easily convert the
unstructured data into the structured form, giving accurate
results, and leading to the perfect diagnosis. Doctors get
help from AI-generated radiologist reports, clinical
laboratory reports, and many other decision-support tools.
Doctors can utilize AI to analyze important discussion
between nurses and patients as well as patient notes. It also
helps doctors to pinpoint the effected parts which reduce
patients’ quality of life and target effect areas before they
develop into chronic ailments.
2.5 Radiologists
Radiologists benefit from AI with regard to detecting
and monitoring diseases. AI algorithms have shown
exceptional progress with tasks related to image recognition.
AI mainly provides assessments of radiological graphs, and
automatically recognizes complicated patterns in the form of
images. Recent AI applications detect the pediatric bone age
known as RSNA, to show the effected part in greater detail
than can be detected by the naked eye. As we know, lung
cancer is the most common and harmful type of tumor
which needs rapid detection and treatment. Screening for
lung cancer is necessary to identify the pulmonary nodules.
AI can automatically identify nodules, which are then
categorized as malignant or benign.
Mammography screening is a big challenge for most
radiologists because of the long process. But AI can assist in
interpreting the results, and easily characterizes calcium
deposits in the breast of the affected person. The use of AI
algorithms in this respect not only reduces the chronic
disease’s progression but also helps radiologists identify
patients with serious conditions and treat them on a priority
basis. AI also suggests which patients need to go to the
radiologist’s department. For example, the person whose leg
is broken and who is unaware of this fact needs to go to
Radiology immediately.
2.6 Clinical laboratories
The use of AI has increased tremendously in the daily
operations and procedures of clinical laboratories.
Digital pathology enables capturing pathology
information such as whole slide images (WSI) and uses
machine learning to spot subtle patterns and provide the
pathologist with detailed information. Use of AI in clinical
microbiology laboratory settings supplements human
ingenuity and is the gold standard in full-laboratory
automation. The algorithms used make it possible to
automatically read and interpret growth on plates, recognize
colony morphology, and serve other essential functions. The
AI-based clinical laboratories of this proposed system will
test the diminutive volume of serum or blood from different
samples in a single day and provide accurate answers to all
the clinical questions which are difficult for human beings.
For the detection of colon, lung, and breast cancer, the CAD
system has been implemented progressively, which is one of
the greatest applications of AI. Thus, CAD has become a
common and popular AI application in clinical practice. AI
has given a new approach for the analysis of complete
human molecular data, as well as genetics. AI has the
potential to fathom and solve numerous challenges related to
clinical trials.
2.7 Pharmacy
In pharma, AI mainly refers to the use of automated
algorithms to perform those tasks or activities which depend
or rely on human intelligence. AI has a great impact on
pharmacies and drug stores’ operational efficiency. The
proposed system will automatically provide notification
about shortages of medicine, oversold or high/low demand
of medicine, et cetera. AI has empowered pharmacists to
move from just filling prescriptions to the management of
the disease and of the patient’s engagement. AI mainly
identifies the relationship between different health problems
and the types of medicines or drugs best suited to treat them.
Hospitals that use the proposed system will provide AI-
based mobile apps to monitor the patients’ use of
medications in real-time. These apps utilize a webcam to
verify whether or not the patients are taking these drugs or
medicines as prescribed. Researchers and scientists who are
developing new medicines or drugs are also using these
pharmaceutical data, which will be explained in the next
section.
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
2.8 Researchers:
The proposed system helps researchers collect data
more rapidly and obtain predictive analyses of new
medicines or new diseases. AI can automatically write the
stories provided by the doctors and nurses, and researchers
can use formatted data for research purposes, obtaining a
perfect and accurate report. This report helps researchers
find the main cause of disease and gather the best evidence
of its interactions with various biomedical entities, as well
as optimizing the manufacturing process.
AI also permits researchers to verify the blend of
biomarkers and recruit the patient, providing the opportunity
for diagnosis. With the new indications, AI allows
researchers to repurpose different drugs, as well as extricate
the biological knowledge for discovering new ones. AI also
plays an important role in testing different compounds
against cells and identifying different compounds that need
more analysis and time. AI speeds up the screening process
and gives quicker results compared to the analysis done by
human beings.
The proposed system will also support Contract
Research Organizations (CRO) for research. The CRO is a
firm that helps support medical industries, pharmacies, and
biotechnology companies, but on a contract basis. They are
mainly introduced to minimize the costs of those companies
who are developing new drugs or medicines for healthcare
markets. CROs have a wide range of experience working
with scientists, researchers, and clinicians to define
particular questions that are beneficial for all.
3. RECENT DEVELOPMENT IN HEALTH SERVICE
USING ARTIFICIAL INTELLIGENCE
AI has amazed the world with its recent technologies,
and everyone in the medical professions is trying to adopt it.
A few of the recent developments, such as qXR for TB
screening, qER tool for Head CT scans, qScout-EMR,
InMotion ARM, Google AI for breast cancer detection,
which will be used in the proposed system are listed below:
3.1 qXR for TB screening:
qXR is a tool for scanning chest X-rays that detects
signs of pulmonary, hilar, and pleural tuberculosis. qXR
uses an AI algorithm that can detect both classic primary
pulmonary TB and its atypical manifestations [19]. In
addition to its applications with regard to tuberculosis, it has
the ability to screen for conditions such lung malignancies
in high-risk populations, COPD, and some cardiac disorders
simultaneously.
3.2 qER tool for Head CT scans
This is a tool detects, localizes, and determines the
extent a multiple types of brain pathologies, such as all
types of intra-cerebral bleeds, midline shift, mass effect,
infarcts, and cranial fractures, as shown in figure 4. The
accuracy is given in table 1.
Fig. 2. Interface of qXR tool (www.qure.ai)
Fig. 3. Chest X-ray screening by qXR (www.qure.ai)
Fig. 4. Interface of qER (www.qure.ai)
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
3.3 qScout-EMR for contact registration and tracing
This tool can be accessed from any mobile device or
laptop. It is mainly designed for COVID-19, for registration
and contacting. However, we will collaborate with the
company to modify this app for all patients, for monitoring
daily symptoms. Figure 5 shows the interface for this tool.
3.4 InMotion ARM
InMotion robots are used as neurorehabilitation tools in
the United States and over 20 other countries. Extensive
research has shown InMotion robots are effective for
treatment of a number of motor impairments, including
spinal cord injury, cerebral palsy, hemiplegic shoulder pain,
multiple sclerosis, Parkinson’s disease, stroke, and muscle
spasticity.
3. CONCLUSION
The proposed architecture shows that any healthcare
organization can implement this AI-based approach to as a
means of reducing costs and managing all stakeholders
efficiently and accurately. The proposed AI-assisted system
supports a patient starting from the time of their admission
into the hospital using emergency medical services,
processes the patient’s long data, detects serious diseases,
automatically recognizes complicated patterns, and can
analyze the complete human molecular data and patient
genetics in a clinic setting. It minimizes potential human
error on the part of the doctors by generating AI-generated
radiologist reports, clinical laboratory reports, and providing
many other decision-support tools, easing the decision
making process with regard to the early detection and
diagnosis of diseases. It not only processes the patient's raw
data, but also generates reports very quickly, which helps
researchers, doctors, nurses, and other stakeholders get the
support they need in an efficient manner. This work also
shows the recent developments which relate to the role of AI
in healthcare and which could be used in this proposed
architecture, such as detection of cancers in the early stages,
finding the genetic code linkages, developing new drugs, et
cetera. In other words, the proposed system will enhance the
capacity of healthcare professionals to understand their
patients’ basic needs, allowing them to more easily guide
and support their patients effectively, making optimal use of
the staff’s time and reducing costs. AI-based healthcare
offers a wealth of improvements to the healthcare industry,
Fi
g
. 6. InMotion ARM
(
www.bioniklabs.co
m
)
Fig. 5. Interface of qScout (www.qure.ai)
Abnormal
finding
AUC
(Confidence
interval) Specificity Sensitivity
Intraparenchymal
haemorrhage
0.95 0.86 0.9
Extradural
hemorrhage
0.97 0.87 0.95
Intracranial
haemorrhage
0.95 0.89 0.9
Subarachnoid
hemorrhage
0.95 0.89 0.9
Subdural
hemorrhage
0.96 0.89 0.9
Intraventricular
haemorrhage
0.98 0.91 0.95
Cranial Fracture 0.96 0.9 0.9
Infarct 0.94 0.87 0.87
Midline Shift 0.97 0.93 0.93
Mass Effect 0.93 0.88 0.88
Atrophy 0.92 0.84 0.84
Table 1. The accuracy of each algorithm
(www.qure.ai)
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
and healthcare professionals and hospitals around the globe
need to move towards AI-based systems.
ACKNOWLEDGMENT
The author would like to thank the reviewers for the
suggestions which help to improve the quality of this paper.
In addition, the authors are also very thankful to Jouf
University, sakaka, Al Jouf, KSA for providing resources.
4.
REFERENCES
[1] Yu, Kun-Hsing, Andrew L. Beam, and Isaac S. Kohane.
"Artificial intelligence in healthcare." Nature
biomedical engineering 2, no. 10 (2018): 719-731.
[2] M. S. Hossain et al. “Smart healthcare monitoring: a
voice pathology detection paradigm for smart cities,”
ACM/Springer Multimedia Systems, vol. 25, no. 5, pp.
565-575, October 2019
[3] M. Alhussein et al. “Cognitive IoT-Cloud Integration
for Smart Healthcare: Case Study for Epileptic Seizure
Detection and Monitoring,” Mobile Networks and
Applications, vol. 23, no. 6, pp 1624–1635, December
2018.
[4] M. Masud et al. “Data Interoperability and Multimedia
Content Management in e-health Systems,” IEEE
Trans. Inf. Technol. Biomed. Vol. 16, No. 6, pp. 1015-
1023, Nov. 2012
[5] M. S. Hossain et al. "Cyber–physical cloud-oriented
multi-sensory smart home framework for elderly
people: An energy efficiency perspective," Journal of
Parallel and Distributed Computing, vol. 103, no. 2017,
pp. 11-21, May 2017.
[6] Chen, G., Wang, L. &Kamruzzaman, M.M. , “Spectral
classification of ecological spatial polarization SAR
image based on target decomposition algorithm and
machine learning” Neural Comput&Applic (2019).
https://doi.org/10.1007/s00521-019-04624-9
[7] Guobin Chena, ZhiyongJiang, M.M.Kamruzzaman,
“Radar Remote Sensing Image Retrieval Algorithm
Based on Improved Sobel Operator” Journal of Visual
Communication and Image Representation,
https://doi.org/10.1016/j.jvcir.2019.102720
[8] Guobin Chen, Qiang Pei, M.M. Kamruzzaman, Remote
sensing image quality evaluation based on deep support
value learning networks, Signal Processing: Image
Communication, vol. 83, 2020
[9] Kamruzzaman M.M. (2020) E-crime Management
System for Future Smart City. In: Huang C., Chan
YW., Yen N. (eds) Data Processing Techniques and
Applications for Cyber-Physical Systems (DPTA
2019). Advances in Intelligent Systems and Computing,
vol 1088. Springer, Singapore
[10] Mingju Chen, Xiaofeng Han, Hua Zhang, Guojun Lin,
M.M. Kamruzzaman, Quality-guided key frames
selection from video stream based on object detection,
Journal of Visual Communication and Image
Representation, Volume 65, 2019, 102678, ISSN 1047-
3203
[11] Chen, GB;Sui, X; Kamruzzaman, MM, “Agricultural
Remote Sensing Image Cultivated Land Extraction
Technology Based on Deep Learning” Revista de la
Facultad de Agronomia de la Universidad del Zulia,
Vol. 36, No.6, 2019, pp.2199-2209
[12] Xue Chen, Lanyong Zhang, Tong Liu, M.M.
Kamruzzaman, Research on deep learning in the field
of mechanical equipment fault diagnosis image quality,
Journal of Visual Communication and Image
Representation, Volume 62, 2019, Pages 402-409,
ISSN 1047-3203,
[13] Pei Yin, M.M.Kamruzzaman,” Animal Image Retrieval
Algorithms Based on Deep Neural Network”
RevistaCientífica, FVC-LUZ / Vol. XXIX, N° 2, 188-
199 , 2019.
[14] Rui Liang, Hui Zhi, M. M. Kamruzzaman, “Methods of
Moving Target Detection and Behavior Recognition in
Intelligent Vision Monitoring” Acta Microscopica,Vol
28 No 4 (2019)
[15] Neill, Daniel B. "Using artificial intelligence to
improve hospital inpatient care." IEEE Intelligent
Systems 28, no. 2 (2013): 92-95.
[16] De Momi, Elena, and Giancarlo Ferrigno. "Robotic and
artificial intelligence for keyhole neurosurgery: the
ROBOCAST project, a multi-modal autonomous path
planner." Proceedings of the Institution of Mechanical
Engineers, Part H: Journal of Engineering in Medicine
224, no. 5 (2010): 715-727.
[17] Noorbakhsh-Sabet, Nariman, RaminZand, Yanfei
Zhang, and Vida Abedi. "Artificial intelligence
transforms the future of healthcare." The American
journal of medicine (2019).
[18] Reddy, Sandeep, John Fox, and Maulik P. Purohit.
"Artificial intelligence-enabled healthcare delivery."
Journal of the Royal Society of Medicine 112, no. 1
(2019): 22-28.
[19] Preetham Putha, Manoj Tadepalli, Bhargava Reddy,
Tarun Raj, Justy Antony Chiramal, Shalini Govil,
Namita Sinha, Manjunath KS, Sundeep Reddivari,
Pooja Rao, and Prashant Warier, Can artificial
intelligence reliably report chest x-rays?: Radiologist
validation of an algorithm trained on 1.2 million x-rays,
CoRR abs/1807.07455 (2018).
Authorized licensed use limited to: Murdoch University. Downloaded on June 16,2020 at 22:57:41 UTC from IEEE Xplore. Restrictions apply.
... In recent years, artificial intelligence (AI), machine learning (ML), and convolutional neural networks (CNN) have emerged as transformative technologies in the healthcare sector (Masud et al., 2020;Chen et al., 2018;Zhang et al., 2019). Leveraging AI and ML in breast cancer diagnosis holds immense promise, offering unparalleled capabilities in prediction, diagnosis, and cost reduction (Kamruzzaman, 2020;Min et al., 2015). Traditional computer-aided design (CAD) systems, while valuable, often rely on manually engineered features, limiting their overall efficacy (Bharati et al., 2020). ...
Article
Full-text available
Breast cancer poses a significant global health challenge, being the most prevalent cancer in women and a leading cause of cancer-related mortality. With the increasing number of diagnoses, there is a pressing need for innovative approaches to enhance detection and treatment. This study explores the effectiveness of boosting algorithms, namely AdaBoost, XGBoost, CatBoost, and LightGBM, in predicting breast cancer by leveraging artificial intelligence (AI) and machine learning (ML) technologies. This research distinguishes itself from previous studies by focusing specifically on gradient boosting algorithms. The research utilizes datasets sourced from Kaggle, comprising 569 instances categorized into Benign (357) and Malignant (212) classes. Data preprocessing steps, such as feature selection and normalization, were performed to improve the model's learning ability and ensure that all features contribute fairly during training. Additionally, Hyperparameter tuning, including adjusting the learning rate, number of trees, and tree depth, was conducted to optimize the model's performance. Following training, various performance metrics including accuracy, sensitivity, precision, specificity, ROC-AUC, and confusion matrix were assessed. The results showed that AdaBoost achieved an accuracy of 97.37%, precision of 0.9545, recall of 0.9767, specificity of 0.9718, and an AUC-ROC of 0.9967. CatBoost achieved an accuracy of 97.49%, precision of 0.9756, recall of 0.9302, specificity of 0.9859, and an AUC-ROC of 0.9977. LightGBM achieved an accuracy of 96.49%, precision of 0.9535, recall of 0.9535, specificity of 0.9718, and an AUC-ROC of 0.9961. XGBoost achieved an accuracy of 95.61%, precision of 0.9524, recall of 0.9302, specificity of 0.9718, and an AUC-ROC of 0.9954. Among the algorithms tested, CatBoost stands out as the top performer. These results demonstrate the potential of ML techniques in breast cancer diagnosis and underscore the importance of ongoing research to address the evolving challenges in oncology. Furthermore, the integration of these algorithms into clinical decision support systems could enhance diagnostic accuracy and facilitate personalized treatment strategies, ultimately improving patient outcomes.
... In the smart healthcare field, the prevailing view is that AI should collaborate with human intelligence. This collaboration could include making the delivery of accurate results easier and faster, thereby assisting doctors in decision making; aiding nurses in managing vast quantities of patient data; and advising and guiding physicians on administering proper treatment; among other benefits [21]. However, in cases where machines surpass human performance, where errors do not result in severe consequences, or in situations where doctors are unavailable and machines can perform effectively, full AI automation is possible [19]. ...
Article
Full-text available
The rapid development of AI technology in recent years has led to its widespread use in daily life, where it plays an increasingly important role. In healthcare, AI has been integrated into the field to develop the new domain of smart healthcare. In smart healthcare, opportunities and challenges coexist. This article provides a comprehensive overview of past developments and recent progress in this area. First, we summarize the definition and characteristics of smart healthcare. Second, we explore the opportunities that AI technology brings to the smart healthcare field from a macro perspective. Third, we categorize specific AI applications in smart healthcare into ten domains and discuss their technological foundations individually. Finally, we identify ten key challenges these applications face and discuss the existing solutions for each.
... 15 It is designated as a clinical support tool and, when used with original scans, can assist the clinician to improve efficiency, accuracy and TAT in reading head CTs. 17 ...
Article
Full-text available
Introduction Diagnostic imaging is vital in emergency departments (EDs). Accessibility and reporting impacts ED workflow and patient care. With radiology workforce shortages, reporting capacity is limited, leading to image interpretation delays. Turnaround times for image reporting are an ED bottleneck. Artificial intelligence (AI) algorithms can improve productivity, efficiency and accuracy in diagnostic radiology, contingent on their clinical efficacy. This includes positively impacting patient care and improving clinical workflow. The ACCEPT-AI study will evaluate Qure.ai’s qER software in identifying and prioritising patients with critical findings from AI analysis of non-contrast head CT (NCCT) scans. Methods and analysis This is a multicentre trial, spanning four diverse sites, over 13 months. It will include all individuals above the age of 18 years who present to the ED, referred for an NCCT. The project will be divided into three consecutive phases (pre-implementation, implementation and post-implementation of the qER solution) in a stepped-wedge design to control for adoption bias and adjust for time-based changes in the background patient characteristics. Pre-implementation involves baseline data for standard care to support the primary and secondary outcomes. The implementation phase includes staff training and qER solution threshold adjustments in detecting target abnormalities adjusted, if necessary. The post-implementation phase will introduce a notification (prioritised flag) in the radiology information system. The radiologist can choose to agree with the qER findings or ignore it according to their clinical judgement before writing and signing off the report. Non-qER processed scans will be handled as per standard care. Ethics and dissemination The study will be conducted in accordance with the principles of Good Clinical Practice. The protocol was approved by the Research Ethics Committee of East Midlands (Leicester Central), in May 2023 (REC (Research Ethics Committee) 23/EM/0108). Results will be published in peer-reviewed journals and disseminated in scientific findings (ClinicalTrials.gov: NCT06027411 ) Trial registration number NCT06027411 .
... and drug prescriptions. Kamruzzaman (2020) in the article "Architecture smart health care system using artificial intelligence" explains the importance of artificial intelligence for healthcare professionals and the creation of an intelligent healthcare system using artificial intelligence for the purpose of quickly processing patient information. Collecting patient information and available data speeds up diagnosis and time to patient care. ...
Thesis
A MARKET INTELLIGENCE MODEL FOR THE HEALTH CARE BUSINESS SYSTEM Health in Brazil is an old problem that has worsened with the Covid-19 pandemic and resources have become scarce. Studying market behavior and growth trends has become fundamental. This research aims to build a Market Intelligence Model, based on information obtained from official secondary data sources, using the Fuzzy logic methodology. The concepts of Data Analytics and Decision Making applied in the Health Area are addressed. The model was developed in RSdudio software, stored in SQL Database, and presented in Power BI and QGis. The model was validated in private health units in the southeast and northeast regions of Brazil for the treatment of chronic diseases, mainly in the areas of nephrology and oncology. The results were obtained, and it can be concluded that the proposed model has become very effective in estimating the growth of chronic diseases.
Article
Full-text available
Cardiac patients require prompt and effective treatment to prevent heart attacks through accurate prediction of heart disease. The prognosis of heart disease is complex and requires advanced knowledge and expertise. Healthcare systems are increasingly integrated with the internet of things (IoT) to collect data from sensors for diagnosing and predicting diseases. Current methods employ machine learning (ML) for these tasks, but they often fall short in creating an intelligent framework due to difficulties in handling high-dimensional data. A groundbreaking health system leverages IoT and an optimized long short-term memory (LSTM) algorithm, enhanced by the red deer (RD) algorithm, to accurately diagnose cardiac issues. Continuous monitoring of blood pressure and electrocardiograms (ECG) is conducted through heart monitor devices and smartwatches linked to patients. The gathered data is combined using a feature fusion approach, integrating electronic medical records (EMR) and sensor data for the extraction process. The RD-LSTM model classifies cardiac conditions as either normal or abnormal, and its performance is benchmarked against other deep-learning (DL) models. The RD-LSTM model showed better improvement in prediction accuracy over previous models.
Conference Paper
Designing a smart environment is crucial stakes. It starts with analysing the environment goals, functional and information requirements, and the process end with decision to employ the smart objects and enabling technologies. One issue is that the risk involved in such investment in these technologies which necessitates ensuring the justification, efficiency, and effectiveness of these employed technologies. Subsequently, there is a need to test these technologies prior to the investment decision, which is difficult to achieve in reality. One feasible solution is to use a simulation model which is capable to test the impact of these technologies. This research presents a new and pioneering methodology for simulating the implementation and testing smart enabling technologies.
Article
Full-text available
Nowadays, with the development of science and technology, image data has become an important part of network data, especially animal images. It is of great significance to effectively retrieve animal images from massive image data of network data for discovering animals or protecting animals. The traditional image retrieval algorithm has the insufficient ability in feature extraction, which is not suitable for effective network image retrieval. The application of depth neural network in image retrieval brings a breakthrough for solving this problem. Aiming at the problem of network animal image data retrieval, a hash network algorithm based on deep neural network is proposed. This method firstly extracts the high-level semantic features of animal images by using depth neural network, and optimizes the depth neural network. Then, the obtained image semantic features are input into the sparse self-coding hash network to obtain short hash codes. Finally, an image retrieval algorithm based on the weight of hash codes is proposed. The simulation results show that the improved depth neural network is used to extract the high-level semantic information of the image, and then the sparse self-coding hash network is used for further training. The obtained short hash codes are retrieved by using the image retrieval algorithm based on the weight of hash codes, which can not only realize animal image retrieval, but also improve the accuracy of image retrieval.
Article
Full-text available
With the rapid development of society, people have higher and higher requirements for the monitoring system and expect personalized and intelligent services. Although the unmanned monitoring regions such as Banks, crossroads, parking lot, has been widely installed surveillance cameras, monitoring to the image is not clear, however, especially in the case of multiple moving targets at the same time appears failure to recognize the detail of a moving target, and often need to monitor the researchers obtained data monitor screen or video for video information, which makes the corresponding judgment, it is to monitor the real-time and judge the accuracy of the great difficulties. Intelligent visual surveillance technology differs from the main characteristics of the previous monitoring technology, is using the method of active vision, in almost without human intervention, by automatically adjusting the rotation of the camera haeundae to locate the region of interest and choice, reach the purpose of tracking and identification of moving targets, and on the basis of further analysis, judgment target behavior, do it can not only in the event of unusual circumstances timely alarm and complete daily management work. Intelligent monitoring system can monitor the moving target and its local details in real time. It is a new generation of highly intelligent monitoring technology. In this paper, from the perspective of theory and practical application, some new explorations are made on the motion behavior recognition with video as input, and effective solutions are sought on the basis of the research on a variety of common tracking and matching algorithms, so as to realize the motion target tracking and matching algorithm with high precision, high robustness and high real-time. In this paper, people or vehicles are taken as the detection objects. Under the condition of fixed cameras, moving targets in the complex background of visual monitoring system are detected and tracked, and moving targets are classified to achieve the purpose of intelligence.
Article
Full-text available
Land resources are the material basis for human survival and development. The characteristics of China's land resources are "one more than three less", that is, the total amount is large, the amount of cultivated land per capita is small, the quality of cultivated land is small, and the reserve resources that can be developed are few. The quantity and quality of cultivated land are related to China's food security. Cultivated land protection has always been the core of land resource management. The acquisition of the quantity of cultivated land and its distribution information is the premise to achieve this goal. Remote sensing technology can objectively obtain the information of cultivated land from the spatial scale of wide area and local area. Remote sensing images have been widely used in large-scale land use surveys and remote sensing monitoring services. However, due to the details of their own ambiguity, large amount of data and large differences within the class, the difficulty of automatic interpretation of images has increased. This makes the actual business work still mainly based on manual visual interpretation, and lacks a highly automated and streamlined working style. Deep learning is a research field that has gradually emerged in recent years. At present, deep learning has become a mainstream research tool in the fields of speech recognition, image recognition, image classification, and target detection. Extracting the information of cultivated land from remote sensing images is a problem of image recognition and classification. Therefore, using deep learning to extract cultivated land from agricultural remote sensing images is a very feasible research program. Based on the deep learning theory, this paper sorts out the research results in the field of remote sensing image classification and deep learning at home and abroad, then preprocesses and labels the remote sensing image to make a training sample set, and finally uses convolutional neural network (CNN). For the classifier, the agricultural remote sensing image was extracted from cultivated land.
Article
Full-text available
With the development of science and technology, the classification of polarimetric SAR images has become an important part of the research of target recognition and image interpretation. However, for the research method is relatively simple and the accuracy is low, this paper carries out the work from the two aspects of feature extraction and feature classification of the ground object, and analyzes and studies the application and value of the polarimetric SAR system. The basic algorithm of polarization SAR image classification is proposed. A polarimetric SAR image feature classification method based on polarization target decomposition and support vector machine is proposed. Four kinds of scattering features and Freeman decomposition are obtained by Cloude decomposition. The simulation results show that the accuracy of using combined features is about 6.5% higher than that of single features. A polarization classification model based on polarization target decomposition and limit learning method is proposed. The simulation experiment shows ELM learning. The algorithm is indeed much faster than SVM learning. In this paper, a polarimetric SAR image classification method based on improved scattering mechanism coefficients is proposed, and the effectiveness of the polarimetric SAR image classification method based on improved scattering mechanism coefficients is verified. Experimental results show that after feature selection, the method of combining Freeman decomposition and Wishart classifier can get better classification results.
Article
Aiming at the problem that the remote sensing image quality evaluation models with manually extracted features lack robustness and generality, this paper proposes a 3D CNN-based architecture and nuclear power plant for accurate remote sensing image quality assessment. The model incorporates two sub-networks. The DSVL-based sub-network is employed to extract multi-scale, multi-direction and high-level features by layer-wise training. Afterwards, the extracted feature maps are fused as flowed as input data of the second sub-network, which is designed with 3D CNN architecture and nuclear power plant for remote sensing image quality assessment. Experimental results on remote sensing image quality database from the GeoEye-1 and WorldView-2 satellites show that the proposed model can optimally discover the essential features of the image and effectively extract the high-frequency information of each level of image, and has better overall quality assessment performance than the other state-of-the-art methods.
Article
Aiming at the time-consuming problem caused by large computational load of radar image retrieval, based on blocking histogram, Sobel edge detection operator and gray level co-occurrence matrix(GLCCM), new radar remote sensing image retrieval algorithm based on improved Sobel operator is proposed. Firstly, the Sobel edge detection algorithm is used to process the image, the edge image is acquired, the radar remote sensing image is analyzed from different angles, and then the different radar remote sensing images are transformed. Then, based on the above processing, Radar Remote Sensing Image Retrieval Algorithm is acquired; finally, the plurality of statistic of the matrix is recorded as a feature vector describing the radar image, and the image is retrieved according to the feature vector of the radar image. Through a large number of experiments, Radar Remote Sensing Image Retrieval algorithm can greatly reduce the retrieval time, and it also has a good retrieval effect for images with rich texture.
Article
Object detection technique is widely applied in modern intelligent systems, such as pedestrian tracking, video surveillance. Key frames selection aims to select more informative frames and reduce amount of redundant information frames. Traditional methods leveraged SIFT feature, which have high key frame selection error rate. In this paper, we propose a novel key frames selection method based on object detection and image quality. Specifically, we first leverage object detector to detect object, such as pedestrian, vehicles. Then, each training frame will be assigned with a quality score, where frames contain objects have high quality score. Afterwards, we leverage CNN based AlexNet architecture for deep feature representation extraction. Our algorithm combines mutual information entropy and SURF image local features to extract key frames. Comprehensive experiments verify the feasibility of practicing the key frame extractor based on convolutional neural network by training the model, and conduct a quality assessment model study.
Article
Image quality assessment (IQA) is an indispensable technique in computer vision, which is widely applied in image classification, image clustering. With the development of deep learning, deep neural network (DNN)-based methods have shown impressive performance. Thus, in this paper, we propose a novel method for mechanical equipment fault diagnosis based on IQA. More specifically, we first conduct data acquisition base on our practice. Afterwards, we leverage image processing method for removing noise. Subsequently, we leverage CNN-based method for image classification. Finally, different mechanical equipment images will be grouped into different categories and fault detection can be achieved. Extensive experiments demonstrate the effectiveness and robustness of our method.
Article
Life sciences researchers using artificial intelligence (AI) are under pressure to innovate faster than ever. Large, multilevel, and integrated data sets offer the promise of unlocking novel insights and accelerating breakthroughs. Although more data are available than ever, only a fraction is being curated, integrated, understood, and analyzed. AI focuses on how computers learn from data and mimic human thought processes. AI increases learning capacity and provides decision support system at scales that are transforming the future of health care. This article is a review of applications for machine learning in health care with a focus on clinical, translational, and public health applications with an overview of the important role of privacy, data sharing, and genetic information.