Internet of medical things (IoMT) is gaining enormous attraction from the healthcare research community. In IoMT, vital health-related information is gathered by medical devices with the help of Internet. Profound supportive data is provided to patients to muddle through their recuperations. However, because of numerous medical equipment, the addresses of devices may be changed by intruders, which is a life threat to serious patients, e.g., the ones having brain tumor-like serious diseases. The mass of irregular cells inside the brain results in brain tumor, which can damage the brain and is life-threatening. Identifying brain tumors in initial stages is quite crucial for its diagnosis, prognosis, and treatment. Conventional methods used for the detection is biopsy and examination of CT scans or magnetic resonance imaging (MRI) by a human, which is a tedious job, impractical for large number of data, and requires the time of radiologist for making inferences. To cope with these challenges, various automated techniques have been proposed. However, still there is a great need for the development of a method that may detect a brain tumor with significant accuracy in less time. Moreover, the selection of features set for making prediction is also very important to achieve remarkable accuracy. In this study, Partial Tree (PART) - an association rule learner, with advanced feature set is adopted to detect brain tumors with respect to its grade, i.e., grade I through grade IV. The proposed model is validated through 10-fold cross validation and is compared with the existing methods, i.e., CART, Random Forest, Naive Bayes, and Random Tree. The obtained results reveal that partial tree with advanced feature set supersedes the aforementioned techniques. Precision, recall, and F-measure are the utilized performance measures for evaluation. (C) 2020 Elsevier B.V. All rights reserved.

IoMT-based computational approach for detecting brain tumor

Guerrieri Antonio;Fortino Giancarlo
2020

Abstract

Internet of medical things (IoMT) is gaining enormous attraction from the healthcare research community. In IoMT, vital health-related information is gathered by medical devices with the help of Internet. Profound supportive data is provided to patients to muddle through their recuperations. However, because of numerous medical equipment, the addresses of devices may be changed by intruders, which is a life threat to serious patients, e.g., the ones having brain tumor-like serious diseases. The mass of irregular cells inside the brain results in brain tumor, which can damage the brain and is life-threatening. Identifying brain tumors in initial stages is quite crucial for its diagnosis, prognosis, and treatment. Conventional methods used for the detection is biopsy and examination of CT scans or magnetic resonance imaging (MRI) by a human, which is a tedious job, impractical for large number of data, and requires the time of radiologist for making inferences. To cope with these challenges, various automated techniques have been proposed. However, still there is a great need for the development of a method that may detect a brain tumor with significant accuracy in less time. Moreover, the selection of features set for making prediction is also very important to achieve remarkable accuracy. In this study, Partial Tree (PART) - an association rule learner, with advanced feature set is adopted to detect brain tumors with respect to its grade, i.e., grade I through grade IV. The proposed model is validated through 10-fold cross validation and is compared with the existing methods, i.e., CART, Random Forest, Naive Bayes, and Random Tree. The obtained results reveal that partial tree with advanced feature set supersedes the aforementioned techniques. Precision, recall, and F-measure are the utilized performance measures for evaluation. (C) 2020 Elsevier B.V. All rights reserved.
2020
Istituto di Calcolo e Reti ad Alte Prestazioni - ICAR
Partial Tree (PART)
Magnetic Resonance Image (MRI)
IoMT security
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.14243/381076
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 40
social impact