Table 1 Summary of related works.

Arslan et al.¹⁹

ML Healthcare monitoring

k-NN, SVM

Patient health monitoring

Increased healthcare monitoring accuracy

Limited to basic ML models; lacks deep learning integration

Shi et al.²⁰

DSFNet for action recognition

CNN

Activity recognition

Enhanced precision through feature extraction

Lacks temporal modeling and multi-sensor fusion

Gong et al.²¹

Wearable data management system

Intelligent analysis techniques

Healthcare data management

Secured data sharing, improved communication

No real-time dynamic data handling

Alruwaili et al.²²

PTL-DTM Model

Probabilistic transfer learning

Wireless transmission monitoring

Improved accuracy and security in data transmission

Model not validated across diverse environments

Imran et al.²³

CNN-BiGRU activity recognition

CNN, BiGRU, Data Mining

Human activity recognition

High activity recognition accuracy

Limited personalization and adaptive learning

Zha et al.²⁴

SCNN-LSTM for SNR prediction

SCNN, LSTM

Wireless sensor SNR prediction

Increased prediction precision, low latency

Lacks robustness to noise or signal anomalies

Tatli et al.²⁵

Prediabetes detection method

ML, Signal Processing

Early disease detection

Cost-effective, accurate early prediction

Focused only on prediabetes, not generalized

Nihar et al.²⁶

Heat stress prediction model

Data-driven ML

Heat stress detection

Accurate indoor heat stress prediction

Limited scalability and contextual adaptation

Matsumura et al.²⁷

Real-time Healthcare analysis

Edge computing

Vital sign monitoring

Generalized, real-time diagnosis enhancement

Lacks integration with cloud-based long-term analytics

Pini et al.²⁸

Infant development monitoring

Sensor-based monitoring

Child development

Enhanced brain development tracking

Limited to physical activity; lacks cognitive/emotional metrics

Kim et al.²⁹

Channel attention for stress detection

Efficient Channel Attention, DNN

Stress detection

High accuracy, reduced computational latency

No adaptive mechanism for individual variability

Matsumura et al.³⁰

Step length estimation for the elderly

ML with inertial sensor

Mobility estimation

Reduced step-length estimation error

Needs broader age group validation

Mahato et al.³¹

HM-HMS (hybrid multimodal health monitoring)

Hybrid multimodal WS

Health diagnostics

Improved diagnostic precision

Not benchmarked against standard clinical datasets

Gashi et al.³²

MS classification

Behavioral Markers, ML

Multiple sclerosis classification

Higher system effectiveness and prediction quality

Disease-specific model lacks generalizability

Li et al.³³

Athlete health monitoring

Naïve Bayes, Convolutional Vector Network

Athlete health tracking

Maximized prediction of physical condition

Needs inclusion of mental/emotional health data

Zhang et al.³⁴

Gait analysis for Parkinson’s

Quantitative gait modeling

Disease diagnosis

Accurate stage classification, reduced complexity

Not integrated with real-time feedback systems