Table 1. Summary of Hyperledger fabric and ML integrated literatures.

Ref	Domain adopted	Implemented blockchain	ML models used	Metrics	Dataset	Shortcomings
Hu & Kar (2024)	Health monitoring system	Hyperledger fabric	ModelChain	Throughput, model accuracy	No	Two blockchains increases the complexity and overhead. Scalability challenges.
Sasikumar & Karthikeyan (2023)	Healthcare (cardiovascular disease severity identification)	Hyperledger consortium	Naïve Bayes	Throughput, latency	Yes	The ML component has been unevaluated and it has been incorporated into the Chaincode, possibly elevating system complexity.
Sharma & Rohilla (2023)	Drug discovery chain management	Hyperledger fabric	No details	Throughput, latency	No	Exclusively utilized ML for visualization of data and preliminary processing, without applying any specific algorithms.
Heidari et al. (2023)	Healthcare (Lung cancer detection)	No details	Federated learning	Accuracy, precision, recall, F1 score	Yes	Lack of transparency in blockchain part and limited interpretability,
Bharimalla et al. (2021)	Electronic health record system	Hyperledger fabric	Natural Language Processing (NLP)	Accuracy, throughput, latency	No	The focus is only on the digital transformation of prescriptions and there is no extensive information about the dataset used.
Pandey et al. (2022)	Waste identification and reuse idea recommending system	Hyperledger fabric	ResNet50	Accuracy, latency	Yes	For reuse idea recommendation system, blockchain integration increases complexity