Irene Dixe de Oliveira Santo (original) (raw)
Papers by Irene Dixe de Oliveira Santo
Radiographics, Apr 1, 2024
Cancers, Sep 29, 2023
Peritumoral edema can contribute significantly to the development of neurological symptoms in pat... more Peritumoral edema can contribute significantly to the development of neurological symptoms in patients with brain metastases (METS), but the quantification of edema has historically been challenging. PACS-based peritumoral edema volume measurement is feasible, and this study suggests that tracking edema volume may facilitate better prediction of treatment outcome. This need is highlighted in our study as over half of the METS studied do not show congruent changes when comparing peritumoral edema volume changes to changes in diameter measurements of contrast-enhancing lesions in longitudinal assessment. Additionally, our results indicate that changes in peritumoral edema volume can pre-date tumor core size changes and could help with early identification of lesions progressing after treatment. Availability of PACS-integrated segmentation tools will allow the incorporation of edema and tumor core volumetrics into treatment response assessment in clinical practice.
arXiv (Cornell University), Sep 17, 2023
Gliomas with CDKN2A mutations are known to have worse prognosis but imaging features of these gli... more Gliomas with CDKN2A mutations are known to have worse prognosis but imaging features of these gliomas are unknown. Our goal is to identify CDKN2A specific qualitative imaging biomarkers in glioblastomas using a new informatics workflow that enables rapid analysis of qualitative imaging features with Visually AcceSAble Rembrandtr Images (VASARI) for large datasets in PACS. Sixty nine patients undergoing GBM resection with CDKN2A status determined by wholeexome sequencing were included. GBMs on magnetic resonance images were automatically 3D segmented using deep learning algorithms incorporated within PACS. VASARI features were assessed using FHIR forms integrated within PACS. GBMs without CDKN2A alterations were significantly larger (64% vs. 30%, p=0.007) compared to tumors with homozygous deletion (HOMDEL) and heterozygous loss (HETLOSS). Lesions larger than 8 cm were four times more likely to have no CDKN2A alteration (OR: 4.3; 95% CI:1.5-12.1; p<0.001). We developed a novel integrated PACS informatics platform for the assessment of GBM molecular subtypes and show that tumors with HOMDEL are more likely to have radiographic evidence of pial invasion and less likely to have deep white matter invasion or subependymal invasion. These imaging features may allow noninvasive identification of CDKN2A allele status.
European Radiology, Jan 21, 2021
Objective To develop machine learning (ML) models capable of predicting ICU admission and extende... more Objective To develop machine learning (ML) models capable of predicting ICU admission and extended length of stay (LOS) after torso (chest, abdomen, or pelvis) trauma, by using clinical and/or imaging data. Materials and methods This was a retrospective study of 840 adult patients admitted to a level 1 trauma center after injury to the torso over the course of 1 year. Clinical parameters included age, sex, vital signs, clinical scores, and laboratory values. Imaging data consisted of any injury present on CT. The two outcomes of interest were ICU admission and extended LOS, defined as more than the median LOS in the dataset. We developed and tested artificial neural network (ANN) and support vector machine (SVM) models, and predictive performance was evaluated by area under the receiver operating characteristic (ROC) curve (AUC). Results The AUCs of SVM and ANN models to predict ICU admission were up to 0.87 ± 0.03 and 0.78 ± 0.02, respectively. The AUCs of SVM and ANN models to predict extended LOS were up to 0.80 ± 0.04 and 0.81 ± 0.05, respectively. Predictions based on imaging alone or imaging with clinical parameters were consistently more accurate than those based solely on clinical parameters. Conclusions The best performing models incorporated imaging findings and outperformed those with clinical findings alone. ML models have the potential to help predict outcomes in trauma by integrating clinical and imaging findings, although further research may be needed to optimize their performance. Key Points • Artificial neural network and support vector machine–based models were used to predict the intensive care unit admission and extended length of stay after trauma to the torso. • Our input data consisted of clinical parameters and CT imaging findings derived from radiology reports, and we found that combining the two significantly enhanced the prediction of both outcomes with either model. • The highest accuracy (83%) and highest area under the receiver operating characteristic curve (0.87) were obtained for artificial neural networks and support vector machines, respectively, by combining clinical and imaging features in the prediction of intensive care unit admission.
Current Problems in Diagnostic Radiology, Jul 1, 2021
Clinical Imaging, Nov 1, 2020
Radiology: Artificial Intelligence
Journal of the American College of Radiology
Cureus, May 19, 2022
Central venous catheters (CVCs) are often crucial in managing severely ill patients, especially t... more Central venous catheters (CVCs) are often crucial in managing severely ill patients, especially those in the intensive care unit. It is estimated that over 5 million CVCs are inserted per year in the United States. The internal jugular, subclavian, or femoral veins are the most used access sites. The catheter is advanced until its tip lies within the proximal third of the superior vena cava, the right atrium, or the inferior vena cava. Unfortunately, the use of CVCs is not without its drawbacks, and multiple immediate and delayed complications have been described. Herein, we report a case of a 70-year-old female with a past medical history significant for chronic obstructive pulmonary disease, coronavirus disease 2019, pneumonia, type 2 diabetes mellitus, and hypertension, who presented to the emergency department from a skilled nursing facility with a two-day history of dyspnea. She was later diagnosed with an intraperitoneal hematoma, an uncommon complication caused by a CVC placement.
European Radiology, 2021
Objective To develop machine learning (ML) models capable of predicting ICU admission and extende... more Objective To develop machine learning (ML) models capable of predicting ICU admission and extended length of stay (LOS) after torso (chest, abdomen, or pelvis) trauma, by using clinical and/or imaging data. Materials and methods This was a retrospective study of 840 adult patients admitted to a level 1 trauma center after injury to the torso over the course of 1 year. Clinical parameters included age, sex, vital signs, clinical scores, and laboratory values. Imaging data consisted of any injury present on CT. The two outcomes of interest were ICU admission and extended LOS, defined as more than the median LOS in the dataset. We developed and tested artificial neural network (ANN) and support vector machine (SVM) models, and predictive performance was evaluated by area under the receiver operating characteristic (ROC) curve (AUC). Results The AUCs of SVM and ANN models to predict ICU admission were up to 0.87 ± 0.03 and 0.78 ± 0.02, respectively. The AUCs of SVM and ANN models to predict extended LOS were up to 0.80 ± 0.04 and 0.81 ± 0.05, respectively. Predictions based on imaging alone or imaging with clinical parameters were consistently more accurate than those based solely on clinical parameters. Conclusions The best performing models incorporated imaging findings and outperformed those with clinical findings alone. ML models have the potential to help predict outcomes in trauma by integrating clinical and imaging findings, although further research may be needed to optimize their performance. Key Points • Artificial neural network and support vector machine–based models were used to predict the intensive care unit admission and extended length of stay after trauma to the torso. • Our input data consisted of clinical parameters and CT imaging findings derived from radiology reports, and we found that combining the two significantly enhanced the prediction of both outcomes with either model. • The highest accuracy (83%) and highest area under the receiver operating characteristic curve (0.87) were obtained for artificial neural networks and support vector machines, respectively, by combining clinical and imaging features in the prediction of intensive care unit admission.
Current Problems in Diagnostic Radiology, 2021
International Journal of Environmental Research and Public Health, 2020
Background: International research has shown that healthcare professionals (HCPs) and nonhealthca... more Background: International research has shown that healthcare professionals (HCPs) and nonhealthcare professionals (NHCPs) are unaware of the goals and purposes of palliative care. This study evaluates the knowledge of palliative care among a sample of Portuguese adults and correlates their level of knowledge with age, gender, profession, and experience of family member’s palliative care. Method: A cross-sectional online survey was carried out on a sample of 152 HCPs and 440 NHCPs who completed an anonymous questionnaire of sociodemographic, family, and professional data, and an instrument of 26 dichotomous (true or false) questions focusing on palliative care goals and purposes. Results: The 592 participants had a mean age of 31.3 ± 11.1 years, and most were female. Statistically significant differences between statements considered as correct by HCPs and NHCPs were found in 24 statements; HCPs had the highest percentage of correct answers. The terms most frequently associated with ...
Clinical Imaging, 2020
Myofibroma is a benign, soft tissue neoplasm that predominantly affects infants and young childre... more Myofibroma is a benign, soft tissue neoplasm that predominantly affects infants and young children. Most occur in the skin or subcutaneous tissues, with a predilection for the head and neck regions. We describe the magnetic resonance (MR) imaging and histophathologic findings of a rare case of intramuscular myofibroma of the right deltoid in a healthy 30-year-old male. MR imaging revealed a well-circumscribed intramuscular mass, with isointense signal on T1-weighted images, hyperintense signal on T2-weighed images, and a "target-sign" with peripheral rim enhancement after gadolinium administration. The lesion was surgically excised with no complications, and the histopathologic analysis revealed the typical morphologic and histochemical markers of a myofibroma. We conclude that, although rare, myofibroma can be considered in the differential diagnosis of adults with lesions the above signal characteristics.
European Journal of Public Health, 2019
version, was used at the beginning (7th year) and at the end of the intervention (8th year). Resu... more version, was used at the beginning (7th year) and at the end of the intervention (8th year). Results 460 students were interviewed, with 279 sessions of health education in the classroom context. All the sessions were evaluated, using a questionnaire that integrates questions related to the contents and a grid of opinion about the session, constituted by 5 items, with scores between [5; 25]. In the last edition of the project, 13% of 7th graders said they had already smoked and, in the previous one, 10% of the students said they smoked, the lowest value found. In order to foster complementarity and convergence solutions to generate positive synergies, the intervention occurred in the classroom in the different disciplines. Obtaining health outcomes implies a consistent and continuous intervention that accompanies the students throughout their formative course. With the GYTS it has been possible to evaluate the impact of the developed intervention.
RadioGraphics
Although eating disorders are common, they tend to be underdiagnosed and undertreated because soc... more Although eating disorders are common, they tend to be underdiagnosed and undertreated because social stigma tends to make patients less likely to seek medical attention and less compliant with medical treatment. Diagnosis is crucial because these disorders can affect any organ system and are associated with the highest mortality rate of any psychiatric disorder. Because of this, imaging findings, when recognized, can be vital to the diagnosis and management of eating disorders and their related complications. The authors familiarize the radiologist with the pathophysiology and sequelae of eating disorders and provide an overview of the related imaging findings. Some imaging findings associated with eating disorders are nonspecific, and others are subtle. The presence of these findings should alert the radiologist to correlate them with the patient's medical history and laboratory results and the clinical team's findings at the physical examination. The combination of these findings may suggest a diagnosis that might otherwise be missed. Topics addressed include (a) the pathophysiology of eating disorders, (b) the clinical presentation of patients with eating disorders and their medical complications and sequelae, (c) the imaging features associated with common and uncommon sequelae of eating disorders, (d) an overview of management and treatment of eating disorders, and (e) conditions that can mimic eating disorders (eg, substance abuse, medically induced eating disorders, and malnourishment in patients with cancer).
Radiographics, Apr 1, 2024
Cancers, Sep 29, 2023
Peritumoral edema can contribute significantly to the development of neurological symptoms in pat... more Peritumoral edema can contribute significantly to the development of neurological symptoms in patients with brain metastases (METS), but the quantification of edema has historically been challenging. PACS-based peritumoral edema volume measurement is feasible, and this study suggests that tracking edema volume may facilitate better prediction of treatment outcome. This need is highlighted in our study as over half of the METS studied do not show congruent changes when comparing peritumoral edema volume changes to changes in diameter measurements of contrast-enhancing lesions in longitudinal assessment. Additionally, our results indicate that changes in peritumoral edema volume can pre-date tumor core size changes and could help with early identification of lesions progressing after treatment. Availability of PACS-integrated segmentation tools will allow the incorporation of edema and tumor core volumetrics into treatment response assessment in clinical practice.
arXiv (Cornell University), Sep 17, 2023
Gliomas with CDKN2A mutations are known to have worse prognosis but imaging features of these gli... more Gliomas with CDKN2A mutations are known to have worse prognosis but imaging features of these gliomas are unknown. Our goal is to identify CDKN2A specific qualitative imaging biomarkers in glioblastomas using a new informatics workflow that enables rapid analysis of qualitative imaging features with Visually AcceSAble Rembrandtr Images (VASARI) for large datasets in PACS. Sixty nine patients undergoing GBM resection with CDKN2A status determined by wholeexome sequencing were included. GBMs on magnetic resonance images were automatically 3D segmented using deep learning algorithms incorporated within PACS. VASARI features were assessed using FHIR forms integrated within PACS. GBMs without CDKN2A alterations were significantly larger (64% vs. 30%, p=0.007) compared to tumors with homozygous deletion (HOMDEL) and heterozygous loss (HETLOSS). Lesions larger than 8 cm were four times more likely to have no CDKN2A alteration (OR: 4.3; 95% CI:1.5-12.1; p<0.001). We developed a novel integrated PACS informatics platform for the assessment of GBM molecular subtypes and show that tumors with HOMDEL are more likely to have radiographic evidence of pial invasion and less likely to have deep white matter invasion or subependymal invasion. These imaging features may allow noninvasive identification of CDKN2A allele status.
European Radiology, Jan 21, 2021
Objective To develop machine learning (ML) models capable of predicting ICU admission and extende... more Objective To develop machine learning (ML) models capable of predicting ICU admission and extended length of stay (LOS) after torso (chest, abdomen, or pelvis) trauma, by using clinical and/or imaging data. Materials and methods This was a retrospective study of 840 adult patients admitted to a level 1 trauma center after injury to the torso over the course of 1 year. Clinical parameters included age, sex, vital signs, clinical scores, and laboratory values. Imaging data consisted of any injury present on CT. The two outcomes of interest were ICU admission and extended LOS, defined as more than the median LOS in the dataset. We developed and tested artificial neural network (ANN) and support vector machine (SVM) models, and predictive performance was evaluated by area under the receiver operating characteristic (ROC) curve (AUC). Results The AUCs of SVM and ANN models to predict ICU admission were up to 0.87 ± 0.03 and 0.78 ± 0.02, respectively. The AUCs of SVM and ANN models to predict extended LOS were up to 0.80 ± 0.04 and 0.81 ± 0.05, respectively. Predictions based on imaging alone or imaging with clinical parameters were consistently more accurate than those based solely on clinical parameters. Conclusions The best performing models incorporated imaging findings and outperformed those with clinical findings alone. ML models have the potential to help predict outcomes in trauma by integrating clinical and imaging findings, although further research may be needed to optimize their performance. Key Points • Artificial neural network and support vector machine–based models were used to predict the intensive care unit admission and extended length of stay after trauma to the torso. • Our input data consisted of clinical parameters and CT imaging findings derived from radiology reports, and we found that combining the two significantly enhanced the prediction of both outcomes with either model. • The highest accuracy (83%) and highest area under the receiver operating characteristic curve (0.87) were obtained for artificial neural networks and support vector machines, respectively, by combining clinical and imaging features in the prediction of intensive care unit admission.
Current Problems in Diagnostic Radiology, Jul 1, 2021
Clinical Imaging, Nov 1, 2020
Radiology: Artificial Intelligence
Journal of the American College of Radiology
Cureus, May 19, 2022
Central venous catheters (CVCs) are often crucial in managing severely ill patients, especially t... more Central venous catheters (CVCs) are often crucial in managing severely ill patients, especially those in the intensive care unit. It is estimated that over 5 million CVCs are inserted per year in the United States. The internal jugular, subclavian, or femoral veins are the most used access sites. The catheter is advanced until its tip lies within the proximal third of the superior vena cava, the right atrium, or the inferior vena cava. Unfortunately, the use of CVCs is not without its drawbacks, and multiple immediate and delayed complications have been described. Herein, we report a case of a 70-year-old female with a past medical history significant for chronic obstructive pulmonary disease, coronavirus disease 2019, pneumonia, type 2 diabetes mellitus, and hypertension, who presented to the emergency department from a skilled nursing facility with a two-day history of dyspnea. She was later diagnosed with an intraperitoneal hematoma, an uncommon complication caused by a CVC placement.
European Radiology, 2021
Objective To develop machine learning (ML) models capable of predicting ICU admission and extende... more Objective To develop machine learning (ML) models capable of predicting ICU admission and extended length of stay (LOS) after torso (chest, abdomen, or pelvis) trauma, by using clinical and/or imaging data. Materials and methods This was a retrospective study of 840 adult patients admitted to a level 1 trauma center after injury to the torso over the course of 1 year. Clinical parameters included age, sex, vital signs, clinical scores, and laboratory values. Imaging data consisted of any injury present on CT. The two outcomes of interest were ICU admission and extended LOS, defined as more than the median LOS in the dataset. We developed and tested artificial neural network (ANN) and support vector machine (SVM) models, and predictive performance was evaluated by area under the receiver operating characteristic (ROC) curve (AUC). Results The AUCs of SVM and ANN models to predict ICU admission were up to 0.87 ± 0.03 and 0.78 ± 0.02, respectively. The AUCs of SVM and ANN models to predict extended LOS were up to 0.80 ± 0.04 and 0.81 ± 0.05, respectively. Predictions based on imaging alone or imaging with clinical parameters were consistently more accurate than those based solely on clinical parameters. Conclusions The best performing models incorporated imaging findings and outperformed those with clinical findings alone. ML models have the potential to help predict outcomes in trauma by integrating clinical and imaging findings, although further research may be needed to optimize their performance. Key Points • Artificial neural network and support vector machine–based models were used to predict the intensive care unit admission and extended length of stay after trauma to the torso. • Our input data consisted of clinical parameters and CT imaging findings derived from radiology reports, and we found that combining the two significantly enhanced the prediction of both outcomes with either model. • The highest accuracy (83%) and highest area under the receiver operating characteristic curve (0.87) were obtained for artificial neural networks and support vector machines, respectively, by combining clinical and imaging features in the prediction of intensive care unit admission.
Current Problems in Diagnostic Radiology, 2021
International Journal of Environmental Research and Public Health, 2020
Background: International research has shown that healthcare professionals (HCPs) and nonhealthca... more Background: International research has shown that healthcare professionals (HCPs) and nonhealthcare professionals (NHCPs) are unaware of the goals and purposes of palliative care. This study evaluates the knowledge of palliative care among a sample of Portuguese adults and correlates their level of knowledge with age, gender, profession, and experience of family member’s palliative care. Method: A cross-sectional online survey was carried out on a sample of 152 HCPs and 440 NHCPs who completed an anonymous questionnaire of sociodemographic, family, and professional data, and an instrument of 26 dichotomous (true or false) questions focusing on palliative care goals and purposes. Results: The 592 participants had a mean age of 31.3 ± 11.1 years, and most were female. Statistically significant differences between statements considered as correct by HCPs and NHCPs were found in 24 statements; HCPs had the highest percentage of correct answers. The terms most frequently associated with ...
Clinical Imaging, 2020
Myofibroma is a benign, soft tissue neoplasm that predominantly affects infants and young childre... more Myofibroma is a benign, soft tissue neoplasm that predominantly affects infants and young children. Most occur in the skin or subcutaneous tissues, with a predilection for the head and neck regions. We describe the magnetic resonance (MR) imaging and histophathologic findings of a rare case of intramuscular myofibroma of the right deltoid in a healthy 30-year-old male. MR imaging revealed a well-circumscribed intramuscular mass, with isointense signal on T1-weighted images, hyperintense signal on T2-weighed images, and a "target-sign" with peripheral rim enhancement after gadolinium administration. The lesion was surgically excised with no complications, and the histopathologic analysis revealed the typical morphologic and histochemical markers of a myofibroma. We conclude that, although rare, myofibroma can be considered in the differential diagnosis of adults with lesions the above signal characteristics.
European Journal of Public Health, 2019
version, was used at the beginning (7th year) and at the end of the intervention (8th year). Resu... more version, was used at the beginning (7th year) and at the end of the intervention (8th year). Results 460 students were interviewed, with 279 sessions of health education in the classroom context. All the sessions were evaluated, using a questionnaire that integrates questions related to the contents and a grid of opinion about the session, constituted by 5 items, with scores between [5; 25]. In the last edition of the project, 13% of 7th graders said they had already smoked and, in the previous one, 10% of the students said they smoked, the lowest value found. In order to foster complementarity and convergence solutions to generate positive synergies, the intervention occurred in the classroom in the different disciplines. Obtaining health outcomes implies a consistent and continuous intervention that accompanies the students throughout their formative course. With the GYTS it has been possible to evaluate the impact of the developed intervention.
RadioGraphics
Although eating disorders are common, they tend to be underdiagnosed and undertreated because soc... more Although eating disorders are common, they tend to be underdiagnosed and undertreated because social stigma tends to make patients less likely to seek medical attention and less compliant with medical treatment. Diagnosis is crucial because these disorders can affect any organ system and are associated with the highest mortality rate of any psychiatric disorder. Because of this, imaging findings, when recognized, can be vital to the diagnosis and management of eating disorders and their related complications. The authors familiarize the radiologist with the pathophysiology and sequelae of eating disorders and provide an overview of the related imaging findings. Some imaging findings associated with eating disorders are nonspecific, and others are subtle. The presence of these findings should alert the radiologist to correlate them with the patient's medical history and laboratory results and the clinical team's findings at the physical examination. The combination of these findings may suggest a diagnosis that might otherwise be missed. Topics addressed include (a) the pathophysiology of eating disorders, (b) the clinical presentation of patients with eating disorders and their medical complications and sequelae, (c) the imaging features associated with common and uncommon sequelae of eating disorders, (d) an overview of management and treatment of eating disorders, and (e) conditions that can mimic eating disorders (eg, substance abuse, medically induced eating disorders, and malnourishment in patients with cancer).