Axillary lymph nodes (ALNs) status is the main predictive factor for distant metastasis and recurrence rate in breast cancer patients. Typical pre-operative assessment of axillary lymph nodes is through Sentinel Lymph node Biopsy (SLNB), but it is associated with complication. Thus, a non-invasive pre-operative alternative approach is needed in order to reduce the number of unnecessary biopsies. Among all the imaging modality, Ultrasound is the primary method for evaluating the axillary nodes because of its real time capabilities, non-ionizing properties, and low cost. Angiogenesis is recognized to be a determinant in the prognosis of cancerous tumors, consequently imaging of submillimeter vessel could be useful in assessing metastatic lesions. Here, is proposed a new ultrasound-based quantitative technique, called high-definition microvasculature imaging (HDMI), that provides a non-invasive and quantitative measure of tumor microvasculature morphological features. Quantitative HDMI include a series of enhancement filter used for the image formation, followed by vessel morphological filtering and vessel segmentation, which allow extraction of morphological features including number of vessel segments (NV), number of branch points (NBP) , vessel density (VD), vessel diameter (D), vessel tortuosity (t), micro-vessel fractal dimension (mvFD), Murray’s deviation (MD), bifurcation angle (BA) and spatial vascularity pattern (SVP) calculated by vessel density ratio (VDR). Additionally, the morphological features are statistical analyzed individually and all combined together in order to create a model able to classify metastatic and reactive ALNs. Our findings revealed that most of the HDMI biomarkers (t, D, NV, NBP, VD, mvFD, MD, BA and VDR) demonstrated meaningful differences between metastatic compared to reactive (p<0.05). Combining all the quantitative biomarkers showed an AUC, sensitivity, and specificity of 87% 87% and 80% respectively. The classification was further improved by adding clinical data, showing an AUC, sensitivity, and specificity of 95%,100% and 80%, respectively. These results provide the basis for the clinical implementation of quantitative HDMI for prediction of axillary lymph node metastasis and staging.

Prediction of Metastatic Axillary Lymph nodes using Quantitative High-Definition Ultrasound Microvessel Imaging

FERRONI, GIULIA
2021/2022

Abstract

Axillary lymph nodes (ALNs) status is the main predictive factor for distant metastasis and recurrence rate in breast cancer patients. Typical pre-operative assessment of axillary lymph nodes is through Sentinel Lymph node Biopsy (SLNB), but it is associated with complication. Thus, a non-invasive pre-operative alternative approach is needed in order to reduce the number of unnecessary biopsies. Among all the imaging modality, Ultrasound is the primary method for evaluating the axillary nodes because of its real time capabilities, non-ionizing properties, and low cost. Angiogenesis is recognized to be a determinant in the prognosis of cancerous tumors, consequently imaging of submillimeter vessel could be useful in assessing metastatic lesions. Here, is proposed a new ultrasound-based quantitative technique, called high-definition microvasculature imaging (HDMI), that provides a non-invasive and quantitative measure of tumor microvasculature morphological features. Quantitative HDMI include a series of enhancement filter used for the image formation, followed by vessel morphological filtering and vessel segmentation, which allow extraction of morphological features including number of vessel segments (NV), number of branch points (NBP) , vessel density (VD), vessel diameter (D), vessel tortuosity (t), micro-vessel fractal dimension (mvFD), Murray’s deviation (MD), bifurcation angle (BA) and spatial vascularity pattern (SVP) calculated by vessel density ratio (VDR). Additionally, the morphological features are statistical analyzed individually and all combined together in order to create a model able to classify metastatic and reactive ALNs. Our findings revealed that most of the HDMI biomarkers (t, D, NV, NBP, VD, mvFD, MD, BA and VDR) demonstrated meaningful differences between metastatic compared to reactive (p<0.05). Combining all the quantitative biomarkers showed an AUC, sensitivity, and specificity of 87% 87% and 80% respectively. The classification was further improved by adding clinical data, showing an AUC, sensitivity, and specificity of 95%,100% and 80%, respectively. These results provide the basis for the clinical implementation of quantitative HDMI for prediction of axillary lymph node metastasis and staging.
2021
2023-02-20
Prediction of Metastatic Axillary Lymph nodes using Quantitative High-Definition Ultrasound Microvessel Imaging
File in questo prodotto:
File Dimensione Formato  
Thesis_giulia_ferroni.pdf

non disponibili

Descrizione: Thesis_Ferroni_Giulia
Dimensione 32.83 MB
Formato Adobe PDF
32.83 MB Adobe PDF

I documenti in UNITESI 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/20.500.12075/12166