The goal of this work is to predict the malignancy of a lesion from the analysis of DW-MRI in a retrospective study. The DW-MRI sequence is used to compute the intravoxel incoherent motion (IVIM) parameters that allow to divide the water movement into diffusion (due to the water present in the tissues) and perfusion (due to the water present in blood flowing in the capillaries). This second movement is not random, but oriented in the direction of the capillaries, but if we recall that capillaries are very short, randomly oriented and with a high density per volume, we can consider the perfusion as a ‘‘pseudo-diffusion’’. Knowing that benign and malign breast tumour have different perfusion characteristics, if we could identify and quantify this feature, we might be able to determine the type of the tumour. In this work, we use state of the art algorithms to compute the IVIM parameters which are then plugged into a learning algorithm, based on retrospective data, that infer the malignancy of the lesion.

Logistic regression to predict malignancy of breast tumors using IVIM parameters

M. Statache;B. M. Maris;R. Menghini;A. Cybulski;M. Barillari;Giulia Zamboni;Paolo Fiorini
2020-01-01

Abstract

The goal of this work is to predict the malignancy of a lesion from the analysis of DW-MRI in a retrospective study. The DW-MRI sequence is used to compute the intravoxel incoherent motion (IVIM) parameters that allow to divide the water movement into diffusion (due to the water present in the tissues) and perfusion (due to the water present in blood flowing in the capillaries). This second movement is not random, but oriented in the direction of the capillaries, but if we recall that capillaries are very short, randomly oriented and with a high density per volume, we can consider the perfusion as a ‘‘pseudo-diffusion’’. Knowing that benign and malign breast tumour have different perfusion characteristics, if we could identify and quantify this feature, we might be able to determine the type of the tumour. In this work, we use state of the art algorithms to compute the IVIM parameters which are then plugged into a learning algorithm, based on retrospective data, that infer the malignancy of the lesion.
2020
IVIM, Machine learning, Breast, DW-MRI
File in questo prodotto:
File Dimensione Formato  
CARS_abstract_IVIM.pdf

accesso aperto

Tipologia: Documento in Pre-print
Licenza: Dominio pubblico
Dimensione 158.85 kB
Formato Adobe PDF
158.85 kB Adobe PDF Visualizza/Apri

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/11562/1038620
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 2
  • ???jsp.display-item.citation.isi??? 0
social impact