WEB This is a Web exclusive article. OBJECTIVE. The purposes of this study were to combine a thor... more WEB This is a Web exclusive article. OBJECTIVE. The purposes of this study were to combine a thorough understanding of the technical aspects of the Whipple procedure with advanced rendering techniques by introducing a virtual Whipple procedure and to evaluate the utility of this new rendering technique in prediction of the arterial variants that cross the anticipated surgical resection plane. CONCLUSION. The virtual Whipple is a novel technique that follows the complex surgical steps in a Whipple procedure. Three-dimensional reconstructed angiographic images are used to identify arterial variants for the surgeon as part of the preoperative radiologic assessment of pancreatic and ampullary tumors.
Diffusion-weighted imaging (DWI) can depict random motions of water molecules in biological tissu... more Diffusion-weighted imaging (DWI) can depict random motions of water molecules in biological tissues during magnetic resonance (MR) examinations. Few papers have tested its application to pancreatic neuroendocrine neoplasms (PanNENs). The aim of this paper is to assess the clinical value of DWI regarding identification and characterization of PanNENs and diagnosis of liver metastases. preoperative MR examinations of 30 PanNEN patients were retrospectively reviewed; thirty patients with pathologically proven pancreatic ductal adenocarcinoma (PDAC) were included to compare imaging features. Qualitative and quantitative MR features were compared between histotypes. A blinded-readers comparison of diagnostic confidence for PanNEN and liver metastases was conducted on randomized image sets. All results were compared to pathological data. PanNENs conspicuity was higher on DW images compared to conventional MR sequences. DWI had higher detection rate for PanNENs than conventional sequences (93.3 vs. 71.1%). Sharp margins and absence of main pancreatic duct/common bile duct dilation and chronic pancreatitis were more common among PanNENs as compared to PDACs. Arterial iso- or hyperenhancement and portal hyperenhancement were more frequent within PanNENs as compared to PDACs. No differences between histotypes were found for quantitative features. Arterial phase images had the highest inter-observer agreement for the diagnosis of PanNEN (k = 0.667). DWI provided the highest detection rate for liver metastases and excellent inter-observer agreement for the diagnosis of liver metastases (k = 0.932), with good accuracy (AUC = 0.879-0.869). DWI has a clinical value regarding the identification of PanNENs and the diagnosis of liver metastases, while conventional MR sequences are fundamental for characterization. Acknowledgements: Italian Cancer Genome Project (FIRB RBAP10AHJB); FIMP-Ministero Salute (CUP_J33G13000210001).
Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death... more Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death from cancer worldwide; more- over the liver is the second most commonly involved organ by metastatic disease. A precise analysis of the lesions would help in the staging of the tumor and in the evaluation of the possible applicable therapies. In this paper
PURPOSE To automatically extract data from CT scans through segmentation to build dynamic models ... more PURPOSE To automatically extract data from CT scans through segmentation to build dynamic models of the nodes and their environment for a preoperative characterization and staging. METHOD AND MATERIALS We extract lymphnode shape and position from CT images and analyze the trend of pixel intensities in the various enhancement phases to determine tissue properties: the trend of contrast medium absorption/reflux varies according to tissue nature: this allows to differentiate between malignant and benign masses, though their HU levels are not visibly different. Knowing the trends of pixel intensities of tumoral and normal nodes in all the acquisition phases, and having the stiffness data of sample lymphnodes, we calibrate the images to quantify the elasticity/rigidity of the masses and classify them. We use lymphnode topography and calibrated data to initialize mass spring models that are inserted in a laparoscopy simulator with force feedback. RESULTS We segmented from abdominal MDCT s...
Gastrointestinal Endoscopy Clinics of North America, 2010
Conventional radiologic and endoscopic evaluations of the small bowel are often limited by the le... more Conventional radiologic and endoscopic evaluations of the small bowel are often limited by the length, caliber, and motility of the small bowel loops. The development of new multidetector-row CT scanners, with faster scan times and isotropic spatial resolution, allows high-resolution multiphasic and multiplanar assessment of the bowel, bowel wall, and lumen. CT Enterography (CTE) is a variant of routine abdominal scanning, geared toward more sustained bowel filling with oral contrast material, and the use of multiplanar images, that can enhance gastrointestinal (GI) tract imaging. This article examines the techniques and clinical applications of CTE in comparison with CT enteroclysis, focusing on Crohn disease, obscure GI bleeding, GI tumors, acute abdominal pain, and bowel obstruction.
International Journal of Computer Assisted Radiology and Surgery, 2009
Purpose Accurate staging of lymph nodes relies mainly on surgical exploration and manual palpatio... more Purpose Accurate staging of lymph nodes relies mainly on surgical exploration and manual palpation. We present a new non-invasive diagnostic approach: simulated palpation through virtual laparoscopic instruments. Methods We set up a diagnostic process to extract lymph nodes shape and position from CTs and to analyze the trend of pixels intensities to determine tissue properties in order to feedback the force information.
Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death... more Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death from cancer worldwide; moreover the liver is the second most commonly involved organ by metastatic disease. A precise analysis of the lesions would help in the staging of the tumor and in the evaluation of the possible applicable therapies. In this paper we present the workflow developed for the semi-automatic segmentation of liver tumors in the data sets provided for the MICCAI Liver Tumor Segmentation contest. Since we wanted to develop a system that could be as automatic as possible and to follow the segmentation process in every single step starting from the image loading to the lesion extraction, we decided to subdivide the workflow in two main steps: first we focus on the liver segmentation and once we have extracted the organ structure we segment the lesions applying an adaptive multi-thresholding system.
Journal of vascular and interventional radiology : JVIR, 2015
A technique of percutaneous hepaticojejunostomy (PHJ) was developed to allow creation of a neoana... more A technique of percutaneous hepaticojejunostomy (PHJ) was developed to allow creation of a neoanastomosis in cases of hepaticojejunostomy (HJ) dehiscence when endoscopic intervention is unfeasible as a result of postsurgical anatomy. PHJ involves transhepatic biliary catheterization and transjejunal retrograde enterotomy. A rendezvous establishes the communication between the bile ducts and the jejunum. PHJ was performed in five patients, and neoanastomosis creation without residual biliary leak was achieved in all cases, with no procedure-related complications. Bilirubin levels and white blood cell counts quickly decreased after PHJ (median, 1 d; range, 1-4 d). Median survival after PHJ was 210 days (range, 45-540 d).
To assess the accuracy of ultrasonography (US) in the identification and grading of hepatic fibro... more To assess the accuracy of ultrasonography (US) in the identification and grading of hepatic fibrosis in patients afflicted with chronic viral liver disease, compared to histological examination as a gold standard. We prospectively studied 105 patients (32 F, 73 M) affected by chronic viral liver disease in 36 months. Patients were studied with B-mode US and then underwent US-guided liver biopsy. All the patients were studied with conventional US with a Sequoia 512, 6.0 (Acuson, Mountain View CA, USA). We evaluated the following US parameters: liver margins, parenchymal echotexture, portal vein caliber and spleen diameter. The four B-mode US parameters were used for the US grading (from 0 to 4). Scheuer's grading (from 0 to 4) was used for the histological score. Grades 3 and 4 were considered as positive for fibrosis. Sensitivity, specificity, positive and negative predictive values and accuracy were calculated in the case of absence, positivity of one or all the US parameters. The correlation between US and histological scores was evaluated with Spearman's test. At histology seventy-seven patients (73%) had absent grade 0 (1 patient; 1%), low-moderate grade 1 (35 patients; 33%) or grade 2 (41 patients; 39%) liver fibrosis. Twenty-eight patients (27%) had severe grade 3 (16 patients; 15%) or grade 4 (12 patients; 11%) fibrosis. In the case of absence of US parameters sensitivity was 32%, specificity 32%, positive predictive value 15%, negative predictive value 57% and accuracy 32%. In the case of positivity of at least one of the US parameters the values were 68%, 68%, 43%, 84% and 69%. In the case of presence of all the US signs the results were 25%, 100%, 100%, 79% and 80%. None of the 77 patients with a healthy liver or with low-grade fibrosis was positive for all the US parameters. All the patients positive for all of the ultrasonographic parameters had high-grade fibrosis or cirrhosis at liver biopsy. Correlation between B-mode and histological scores was not statistically significant (Rs=0.45; p=0.0001). US identification of liver fibrosis in chronic liver disease is possible with 25% sensitivity, 100% specificity, 100% positive predictive value and 79% negative predictive value, with an 80% diagnostic accuracy.
WEB This is a Web exclusive article. OBJECTIVE. The purposes of this study were to combine a thor... more WEB This is a Web exclusive article. OBJECTIVE. The purposes of this study were to combine a thorough understanding of the technical aspects of the Whipple procedure with advanced rendering techniques by introducing a virtual Whipple procedure and to evaluate the utility of this new rendering technique in prediction of the arterial variants that cross the anticipated surgical resection plane. CONCLUSION. The virtual Whipple is a novel technique that follows the complex surgical steps in a Whipple procedure. Three-dimensional reconstructed angiographic images are used to identify arterial variants for the surgeon as part of the preoperative radiologic assessment of pancreatic and ampullary tumors.
Diffusion-weighted imaging (DWI) can depict random motions of water molecules in biological tissu... more Diffusion-weighted imaging (DWI) can depict random motions of water molecules in biological tissues during magnetic resonance (MR) examinations. Few papers have tested its application to pancreatic neuroendocrine neoplasms (PanNENs). The aim of this paper is to assess the clinical value of DWI regarding identification and characterization of PanNENs and diagnosis of liver metastases. preoperative MR examinations of 30 PanNEN patients were retrospectively reviewed; thirty patients with pathologically proven pancreatic ductal adenocarcinoma (PDAC) were included to compare imaging features. Qualitative and quantitative MR features were compared between histotypes. A blinded-readers comparison of diagnostic confidence for PanNEN and liver metastases was conducted on randomized image sets. All results were compared to pathological data. PanNENs conspicuity was higher on DW images compared to conventional MR sequences. DWI had higher detection rate for PanNENs than conventional sequences (93.3 vs. 71.1%). Sharp margins and absence of main pancreatic duct/common bile duct dilation and chronic pancreatitis were more common among PanNENs as compared to PDACs. Arterial iso- or hyperenhancement and portal hyperenhancement were more frequent within PanNENs as compared to PDACs. No differences between histotypes were found for quantitative features. Arterial phase images had the highest inter-observer agreement for the diagnosis of PanNEN (k = 0.667). DWI provided the highest detection rate for liver metastases and excellent inter-observer agreement for the diagnosis of liver metastases (k = 0.932), with good accuracy (AUC = 0.879-0.869). DWI has a clinical value regarding the identification of PanNENs and the diagnosis of liver metastases, while conventional MR sequences are fundamental for characterization. Acknowledgements: Italian Cancer Genome Project (FIRB RBAP10AHJB); FIMP-Ministero Salute (CUP_J33G13000210001).
Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death... more Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death from cancer worldwide; more- over the liver is the second most commonly involved organ by metastatic disease. A precise analysis of the lesions would help in the staging of the tumor and in the evaluation of the possible applicable therapies. In this paper
PURPOSE To automatically extract data from CT scans through segmentation to build dynamic models ... more PURPOSE To automatically extract data from CT scans through segmentation to build dynamic models of the nodes and their environment for a preoperative characterization and staging. METHOD AND MATERIALS We extract lymphnode shape and position from CT images and analyze the trend of pixel intensities in the various enhancement phases to determine tissue properties: the trend of contrast medium absorption/reflux varies according to tissue nature: this allows to differentiate between malignant and benign masses, though their HU levels are not visibly different. Knowing the trends of pixel intensities of tumoral and normal nodes in all the acquisition phases, and having the stiffness data of sample lymphnodes, we calibrate the images to quantify the elasticity/rigidity of the masses and classify them. We use lymphnode topography and calibrated data to initialize mass spring models that are inserted in a laparoscopy simulator with force feedback. RESULTS We segmented from abdominal MDCT s...
Gastrointestinal Endoscopy Clinics of North America, 2010
Conventional radiologic and endoscopic evaluations of the small bowel are often limited by the le... more Conventional radiologic and endoscopic evaluations of the small bowel are often limited by the length, caliber, and motility of the small bowel loops. The development of new multidetector-row CT scanners, with faster scan times and isotropic spatial resolution, allows high-resolution multiphasic and multiplanar assessment of the bowel, bowel wall, and lumen. CT Enterography (CTE) is a variant of routine abdominal scanning, geared toward more sustained bowel filling with oral contrast material, and the use of multiplanar images, that can enhance gastrointestinal (GI) tract imaging. This article examines the techniques and clinical applications of CTE in comparison with CT enteroclysis, focusing on Crohn disease, obscure GI bleeding, GI tumors, acute abdominal pain, and bowel obstruction.
International Journal of Computer Assisted Radiology and Surgery, 2009
Purpose Accurate staging of lymph nodes relies mainly on surgical exploration and manual palpatio... more Purpose Accurate staging of lymph nodes relies mainly on surgical exploration and manual palpation. We present a new non-invasive diagnostic approach: simulated palpation through virtual laparoscopic instruments. Methods We set up a diagnostic process to extract lymph nodes shape and position from CTs and to analyze the trend of pixels intensities to determine tissue properties in order to feedback the force information.
Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death... more Liver cancer is the fifth most commonly diagnosed cancer and the third most common cause of death from cancer worldwide; moreover the liver is the second most commonly involved organ by metastatic disease. A precise analysis of the lesions would help in the staging of the tumor and in the evaluation of the possible applicable therapies. In this paper we present the workflow developed for the semi-automatic segmentation of liver tumors in the data sets provided for the MICCAI Liver Tumor Segmentation contest. Since we wanted to develop a system that could be as automatic as possible and to follow the segmentation process in every single step starting from the image loading to the lesion extraction, we decided to subdivide the workflow in two main steps: first we focus on the liver segmentation and once we have extracted the organ structure we segment the lesions applying an adaptive multi-thresholding system.
Journal of vascular and interventional radiology : JVIR, 2015
A technique of percutaneous hepaticojejunostomy (PHJ) was developed to allow creation of a neoana... more A technique of percutaneous hepaticojejunostomy (PHJ) was developed to allow creation of a neoanastomosis in cases of hepaticojejunostomy (HJ) dehiscence when endoscopic intervention is unfeasible as a result of postsurgical anatomy. PHJ involves transhepatic biliary catheterization and transjejunal retrograde enterotomy. A rendezvous establishes the communication between the bile ducts and the jejunum. PHJ was performed in five patients, and neoanastomosis creation without residual biliary leak was achieved in all cases, with no procedure-related complications. Bilirubin levels and white blood cell counts quickly decreased after PHJ (median, 1 d; range, 1-4 d). Median survival after PHJ was 210 days (range, 45-540 d).
To assess the accuracy of ultrasonography (US) in the identification and grading of hepatic fibro... more To assess the accuracy of ultrasonography (US) in the identification and grading of hepatic fibrosis in patients afflicted with chronic viral liver disease, compared to histological examination as a gold standard. We prospectively studied 105 patients (32 F, 73 M) affected by chronic viral liver disease in 36 months. Patients were studied with B-mode US and then underwent US-guided liver biopsy. All the patients were studied with conventional US with a Sequoia 512, 6.0 (Acuson, Mountain View CA, USA). We evaluated the following US parameters: liver margins, parenchymal echotexture, portal vein caliber and spleen diameter. The four B-mode US parameters were used for the US grading (from 0 to 4). Scheuer's grading (from 0 to 4) was used for the histological score. Grades 3 and 4 were considered as positive for fibrosis. Sensitivity, specificity, positive and negative predictive values and accuracy were calculated in the case of absence, positivity of one or all the US parameters. The correlation between US and histological scores was evaluated with Spearman's test. At histology seventy-seven patients (73%) had absent grade 0 (1 patient; 1%), low-moderate grade 1 (35 patients; 33%) or grade 2 (41 patients; 39%) liver fibrosis. Twenty-eight patients (27%) had severe grade 3 (16 patients; 15%) or grade 4 (12 patients; 11%) fibrosis. In the case of absence of US parameters sensitivity was 32%, specificity 32%, positive predictive value 15%, negative predictive value 57% and accuracy 32%. In the case of positivity of at least one of the US parameters the values were 68%, 68%, 43%, 84% and 69%. In the case of presence of all the US signs the results were 25%, 100%, 100%, 79% and 80%. None of the 77 patients with a healthy liver or with low-grade fibrosis was positive for all the US parameters. All the patients positive for all of the ultrasonographic parameters had high-grade fibrosis or cirrhosis at liver biopsy. Correlation between B-mode and histological scores was not statistically significant (Rs=0.45; p=0.0001). US identification of liver fibrosis in chronic liver disease is possible with 25% sensitivity, 100% specificity, 100% positive predictive value and 79% negative predictive value, with an 80% diagnostic accuracy.
Uploads
Papers by Giulia Zamboni