Figure 1:
Figure 2:
Figure 3:
Figure 4:
Figure 5:
Figure 6:
Figure 7:
Figure 8:
Figure 9:
Figure 10:
Comparative study of ML and big data in healthcare
| S.No. | Author(s) | Title | Focus area | Key technologies | Dataset sources | Key findings | Challenges and limitations |
|---|---|---|---|---|---|---|---|
| 1 | Current study | Disease Prediction, Treatment Personalization, and Operational Efficiency in Healthcare using ML & Big Data | Improving disease prognosis, optimizing treatments, and enhancing hospital efficiency | GNNs, RL, FL, XAI | EHRs, Medical Imaging, Wearable Devices | - 90% accuracy in disease prediction | - Data integration complexity |
| - 25% boost in patient adherence | - High computational requirements | ||||||
| - 30% operational efficiency improvement | - Implementation of privacy safeguards at scale | ||||||
| 2 | Ratnaprabha Ravindra Borhade | AI-Enhanced Predictive Analytics for Proactive Healthcare Management | Proactive healthcare management | ML algorithms | Patient health records, operational data | Improved patient care and operational efficiency through predictive analytics | Data privacy concerns, integration with existing systems |
| 3 | John M. Gates, Yulianti Yulianti, Greian April Pangilinan | Big Data Analytics for Predictive Insights in Healthcare | Predictive insights in healthcare | Big data analytics | EHRs, medical imaging data | Enhanced predictive insights leading to better patient outcomes | Data quality issues, computational resource requirements |
| 4 | Solomon Kavuta, Mr Joel Mulepa | Machine Learning in Health Analytics and Patient Monitoring | Health analytics and patient monitoring | ML algorithms | Patient monitoring data, health analytics records | Improved patient monitoring and health analytics through ML | Data heterogeneity, real-time processing challenges |
| 5 | M. M. Uddin, Ashraful Islam, Rina Rani Saha, Debashish Goswami | The Role Of Machine Learning In Transforming Healthcare: A Systematic Review | Transforming healthcare | ML algorithms | Various healthcare datasets | Systematic review highlighting the transformative role of ML in healthcare | Variability in study methodologies, generalizability of findings |
| 6 | Kristijan Cincar, Andrea Amalia Minda, Marija Varga | A Simulation-based Analysis Using Machine Learning Models to Optimize Patient Flow and Treatment Costs | Patient flow and treatment cost optimization | ML models, simulation techniques | Hospital operational data | Optimization of patient flow and reduction in treatment costs through simulation-based analysis | Model complexity, data accuracy |
| 7 | Iman Akour, Said A. Salloum | The Impact of Big Data Analytics on HealthCare: A Systematic Review | Big data analytics in healthcare | Big data analytics | Various healthcare datasets | Systematic review demonstrating the impact of big data analytics on healthcare | Data privacy concerns, integration challenges |
| 8 | Monika Sharma, Dimple Tiwari, Neeta Verma, Anjali Singhal | Revolutionizing Healthcare: The Power of Machine Learning | ML in healthcare | ML algorithms | EHRs, genomic data | Highlighted the revolutionary potential of ML in healthcare | Ethical considerations, data security |
| 9 | Olayanju Adedoyin Zainab, Toochukwu Juliet Mgbole | Utilization of Big Data Analytics to Identify Population Health Trends and Optimize Healthcare Delivery System Efficiency | Population health trends and healthcare delivery efficiency | Big data analytics | Population health data, healthcare delivery records | Identification of health trends and optimization of healthcare delivery through big data analytics | Data integration issues, scalability |
| 10 | Royana Anand | Enhancing Patient Care Pathways through AI-Driven Data Science and Program Management Strategies | Patient care pathways enhancement | AI-driven data science | Patient care data, program management records | Improved patient care pathways through AI-driven strategies | Implementation challenges, data quality |
Meta-analysis
| Key findings | Method used | Advantage | Remarks |
|---|---|---|---|
| Integrated big data sources such as EHRs, medical imaging, and wearable devices for a comprehensive view of patient health | Big data analytics | Improved decision-making and patient outcomes | Improved clinical decision-making and disease diagnosis accuracy |
| Introduced GNNs for disease prediction, leveraging relationships between patient variables | GNNs | Enhanced prediction accuracy and model interpretability | Outperformed traditional ML models in prediction accuracy |
| Demonstrated FL for collaborative training on decentralized data while ensuring data privacy | FL | Preserved data privacy, comparable model accuracy | Ensured compliance with privacy regulations while achieving competitive accuracy |
| Explored ML in optimizing hospital operations, particularly resource allocation and staff scheduling | ML | Enhanced hospital efficiency by 20% | ML models improved hospital operations, leading to a 20% efficiency improvement |
| Expanded on the challenges of integrating and processing fragmented healthcare data using big data technologies | Big data analytics | Real-time analytics improving decision-making | Identified challenges like data quality, but improved decision-making and operational efficiency |
| Applied RL to personalize treatment plans based on patient-specific data | RL | Dynamic treatment optimization, real-time feedback | Demonstrated the effectiveness of RL in personalized treatment regimens |
| Integrated privacy-preserving mechanisms like differential privacy and SMPC with FL | FL with differential privacy and SMPC | Ensured privacy during model training | Safeguarded patient data while ensuring feasible deployment of ML models in healthcare |
| Proposed hybrid system integrating ML with hospital management systems to optimize patient flow and reduce waiting times | Hybrid system (ML + hospital management systems) | Improved operational efficiency by 30% | Increased hospital efficiency through dynamic scheduling, reducing waiting times by 30% |
| ML models applied to large-scale health data to predict patient outcomes and identify disease patterns | ML | Improved diagnostic accuracy and treatment plans | Applied ML to enhance disease prediction and patient outcome identification |
| Combined GNNs and RL for a hybrid model to improve both disease prediction and treatment optimization | Hybrid model (GNN + RL) | Enhanced healthcare delivery, better disease prediction and treatment | Integrated GNN and RL for improved prediction accuracy and treatment outcomes |
| Compared FL and centralized models, highlighting the advantages of FL in data privacy and security | FL | Ensured patient data confidentiality | FL offered better data privacy with minimal trade-offs in accuracy |
| Introduced predictive analytics for staffing needs based on patient admission rates, reducing downtime, and improving care | Predictive analytics for staffing | Reduced downtime, improved care delivery | Enhanced patient care by improving staffing efficiency |
| Suggested integrating XAI for improved scalability and interpretability of ML models in healthcare | XAI | Increased clinician trust, better transparency | Proposed XAI for improved transparency in ML model predictions |
| Explored combining ML with edge computing for real-time patient monitoring to reduce latency and enhance efficiency | Edge computing + ML | Faster decision-making, reduced latency | Integrated edge computing to enable faster, real-time decision-making in patient care |
| Proposed using block chain technology to enhance FL, ensuring secure and immutable records of patient data | Block chain + FL | Improved data security and integrity | Block chain ensured secure data transmission and integrity in FL environments |
| Discussed limitations of ML models in handling incomplete or biased data, calling for robust pre-processing techniques | ML | Addressed limitations in data quality | Focused on improving data handling and pre-processing for better model accuracy |