The primary objectives of this report were, first, to determine the number and incidence of compl... more The primary objectives of this report were, first, to determine the number and incidence of complications of transsphenoidal surgery performed by a cross-section of neurosurgeons in the United States and, second, to ascertain the influence of the surgeon's experience with the procedure on the occurrence of these complications. The secondary objective was to review complications of transsphenoidal surgery from the standpoint of their causation, treatment, and prevention. Questionnaires regarding 14 specific complications of transsphenoidal surgery were mailed to 3172 neurosurgeons. The data reported were analyzed from the 958 respondents (82%) who reported performing the operation. The neurosurgeons surveyed were asked to estimate the number of transsphenoidal operations performed, to identify any complications observed, and to estimate the percentage of operations that had resulted in any of the 14 specific complications. The 958 respondents were placed into three experience groups, based on the number of transsphenoidal operations performed. The data were analyzed by using chi 2 tests and Spearman correlation coefficients. The secondary objectives were met through a detailed review of the literature, in light of our experience. Of the respondents, 87.3% reported having performed < 200 operations and 9.7% reported 200 to 500 previous operations. The remaining 3% reported more than 500 previous operations. More extensive previous experience with transsphenoidal surgery was associated with a greater likelihood of having witnessed each specific complication. The mean operative mortality rate for all three groups was 0.9%. Anterior pituitary insufficiency (19.4%) and diabetes insipidus (17.8%) were complications with the highest incidence of occurrence. The overall incidence of cerebrospinal fluid fistulas was 3.9%. Other significant complications, such as carotid artery injuries, hypothalamic injuries, loss of vision, and meningitis, occurred with incidence rates between 1 and 2%. An inverse relationship was found between the experience group and the likelihood of complications, as indicated by significant negative Spearman correlation coefficients for all but 2 of the 14 complications listed in the survey (P < 0.05). Thus, increased experience with transsphenoidal surgery seems to be associated with a decreased percentage of operations resulting in complications. Some caution should be exercised in interpreting these data, because they are based on the respondents' estimates. Transsphenoidal surgery seems to be a reasonably safe procedure, with a mortality rate of less than 1%. However, a significant number of complications do occur. The incidence of these complications seems to be higher, with statistical significance, in the hands of less experienced surgeons. The learning curve seems to be relatively shallow, because a statistically significantly decreased incidence of morbidity and death could be documented after 200 and even 500 transsphenoidal operations. Better understanding of the indications for transsphenoidal surgery and improved familiarity with the regional anatomy should further lower the incidence of death and morbidity resulting from this procedure in the hands of all neurosurgeons.
Mining from neuroimaging data is becoming increasingly popular in the eld of healthcare and bioin... more Mining from neuroimaging data is becoming increasingly popular in the eld of healthcare and bioinformatics, due to its potential to discover clinically meaningful structure pa erns that could facilitate the understanding and diagnosis of neurological and neuropsychiatric disorders. Most recent research concentrates on applying subgraph mining techniques to discover connected subgraph patterns in the brain network. However, the underlying brain network structure is complicated. As a shallow linear model, subgraph mining cannot capture the highly non-linear structures, resulting in sub-optimal pa erns. erefore, how to learn representations that can capture the highly non-linearity of brain networks and preserve the underlying structures is a critical problem. In this paper, we propose a Structural Deep Brain Network mining method, namely SDBN, to learn highly non-linear and structurepreserving representations of brain networks. Speci cally, we rst introduce a novel graph reordering approach based on module identi cation, which rearranges the order of the nodes to preserve the modular structure of the graph. Next, we perform structural augmentation to further enhance the spatial information of the reordered graph. en we propose a deep feature learning framework for combining supervised learning and unsupervised learning in a small-scale se ing, by augmenting Convolutional Neural Network (CNN) with decoding pathways for reconstruction. With the help of the multiple layers of non-linear mapping, the proposed SDBN approach can capture the highly non-linear structure of brain networks. Further, it has be er generalization capability for high-dimensional brain networks and works well even for small sample learning. Bene t from CNN's task-oriented learning style, the learned hierarchical representation is meaningful for the clinical * Corresponding author. is work was done while the author was at the University of Illinois at Chicago.
Before the introduction of antiretroviral therapy, human immunodeficiency virus (HIV) infection w... more Before the introduction of antiretroviral therapy, human immunodeficiency virus (HIV) infection was often accompanied by central nervous system (CNS) opportunistic infections and HIV encephalopathy marked by profound structural and functional alterations detectable with neuroimaging. Treatment with antiretroviral therapy nearly eliminated CNS opportunistic infections, while neuropsychiatric impairment and peripheral nerve and organ damage have persisted among virally suppressed people with HIV (PWH), suggesting ongoing brain injury. Neuroimaging research must use methods sensitive for detecting subtle HIV-associated brain structural and functional abnormalities, while allowing for adjustments for potential confounders, such as age, sex, substance use, hepatitis C coinfection, cardiovascular risk, and others. Here, we review existing and emerging neuroimaging tools that demonstrated promise in detecting markers of HIV-associated brain pathology and explore strategies to study the impact of potential confounding factors on these brain measures. We emphasize neuroimaging approaches that may be used in parallel to gather complementary information, allowing efficient detection and interpretation of altered brain structure and function associated with suboptimal clinical outcomes among virally suppressed PWH. We examine the advantages of each imaging modality and systematic approaches in study design and analysis. We also consider advantages of combining experimental and statistical control techniques to improve sensitivity and specificity of biotype identification and explore the costs and benefits of aggregating data from multiple studies to achieve larger sample sizes, enabling use of emerging methods for combining and analyzing large, multifaceted data sets. Many of the topics addressed in this article were discussed at the National Institute of Mental Health meeting "Biotypes of CNS Complications in People Living with HIV," held in October 2021, and are part of ongoing research initiatives to define the role of neuroimaging in emerging alternative approaches to identifying biotypes of CNS complications in PWH. An outcome of these considerations may be the development of a common neuroimaging protocol available for researchers to use in future studies examining neurological changes in the brains of PWH.
This study aimed to examine cognitive function in acute/early HIV infection over the subsequent 2... more This study aimed to examine cognitive function in acute/early HIV infection over the subsequent 2 years. Fifty-six HIV+ subjects and 21 seronegative participants of the Chicago Early HIV Infection Study were evaluated using a comprehensive neuropsychological assessment at study enrollment and at 2-year follow-up. Cognitive performance measures were compared in the groups using t tests and mixed-effect models. Patterns of relationship with clinical measures were determined between cognitive function and clinical status markers using Spearman's correlations. At the initial timepoint, the HIV group demonstrated significantly weaker performance on measures of verbal memory, visual memory, psychomotor speed, motor speed, and executive function. A similar pattern was found when cognitive function was examined at follow-up and across both timepoints. The HIV subjects had generally weaker performance on psychomotor speed, executive function, motor speed, visual memory, and verbal memory. The rate of decline in cognitive function across the 2-year follow-up period did not differ between groups. Correlations between clinical status markers and cognitive function at both timepoints showed weaker performance associated with increased disease burden. Neurocognitive difficulty in chronic HIV infection may have very early onset and reflect consequences of initial brain viral invasion and neuroinflammation during the intense, uncontrolled viremia of acute HIV infection. Further characterization of the changes occurring in initial stages of infection and the risk and protective factors for cognitive function could inform new strategies for neuroprotection.
Journal of Vascular and Interventional Radiology, Feb 1, 2010
To assess the feasibility and long-term clinical outcomes of percutaneous image-guided, multi-pro... more To assess the feasibility and long-term clinical outcomes of percutaneous image-guided, multi-probe cryotherapy in patients refusing surgery for breast cancer. Materials and Methods: After breast magnetic resonance (MR) imaging and thorough consultation, patient consents were obtained for institutional review board approved breast cryotherapy. In 13 cryotherapy sessions, 25 breast cancer foci were treated in 13 patients, Stages I-IV, using multiple 2.4mm cryoprobes. Using only local anesthesia with mild sedation, ultrasound (US) guidance alone was used in 6 patients and 7 patients required both computed tomography (CT) and US to better define ice margins. Saline injections protected the skin and/or chest wall. MR and/or clinical follow-up were available for up to 65 months post cryotherapy. Results: MR helped define multifocal recurrences and initial tumors as well as careful follow up of avascular ablation zones covering all initial tumors. US produced sufficient ice visualization for small tumors while CT confirmed posterior margins and overall ice extent for larger regions requiring Ͼ3 probes or deep lesions. Pretreatment breast tumor diameter was 1.7 ϩ 1.2 cm (range: 0.5-5.8 cm) and an average of 3.3 cryoprobes produced ice diameters of 5.2 ϩ 2.2 cm (range: 2.0-10.0 cm). No significant complications were noted. All patients reported minimal discomfort and satisfaction with cosmetic results. Biopsies at the margins of the cryotherapy site immediately after the procedure and at the cryotherapy site in follow-up were all negative. No local recurrences have been noted at 18 months average followup. Conclusion: In conjunction with thorough pre and postablation MRI, clinic-based multi-probe cryotherapy of breast cancer for local control and/or potential cure provides safe, effective breast conservation with minimal discomfort.
Magnetization transfer imaging (MT) may have considerable promise for early detection and monitor... more Magnetization transfer imaging (MT) may have considerable promise for early detection and monitoring of subtle brain changes before they are apparent on conventional magnetic resonance images. At 3 Tesla (T), MT affords higher resolution and increased tissue contrast associated with macromolecules. The reliability and reproducibility of a new high-resolution MT strategy were assessed in brain images acquired from 9 healthy subjects. Repeated measures were taken for 12 brain regions of interest (ROIs): genu, splenium, and the left and right hemispheres of the hippocampus, caudate, putamen, thalamus, and cerebral white matter. Spearman's correlation coefficient, coefficient of variation, and intraclass correlation coefficient (ICC) were computed. Multivariate mixed-effects regression models were used to fit the mean ROI values and to test the significance of the effects due to region, subject, observer, time, and manual repetition. A sensitivity analysis of various model specifications and the corresponding ICCs was conducted. Our statistical methods may be generalized to many similar evaluative studies of the reliability and reproducibility of various imaging modalities.
Cerebrovascular dual-venc 4D Flow MRI has utility in evaluating intracranial hemodynamics - benef... more Cerebrovascular dual-venc 4D Flow MRI has utility in evaluating intracranial hemodynamics - benefitting from full circle of Willis (CoW) volumetric coverage, time-resolved, three-directional velocity encoding, and velocity dynamic range. Analysis of hemodynamic measures was completed using a comprehensive, semi-automated vessel identification and segmentation workflow for 10 subjects who underwent two scans within 30 days. Test-retest analysis of cerebrovascular dual-venc 4D Flow MRI and measures of CoW 3D flow dynamics demonstrated excellent repeatability for flow and peak velocity measures and lower reproducibility for pulsatility and resistive indices. Future work may include a larger cohort or more advanced analysis of velocity-derived parameters.
Mechanisms underlying heart-brain hemodynamic coupling and effects on the brain remain unclear du... more Mechanisms underlying heart-brain hemodynamic coupling and effects on the brain remain unclear due to challenges of measuring both heart and brain in a single MRI exam. We have developed a comprehensive MRI protocol that incorporates 1) 4D flow MRI of chest, 2) 4D flow MRI of head and 3) structural neuroimaging into one MRI exam to assess cardiovascular and cerebrovascular flow as well as white matter lesions and brain atrophy. This study demonstrates the utility of heart-brain MRI as viable new tool for in vivo evaluation of hemodynamic coupling along the entire heart-brain pathway.
4D flow MRI provides a comprehensive assessment of hemodynamics through the 3D visualization and ... more 4D flow MRI provides a comprehensive assessment of hemodynamics through the 3D visualization and quantification of blood flow. However, its current clinical usage is hindered by long scan times. Recent developments have shown deep learning to be highly effective in accelerating image reconstruction, but no study has shown the its effectiveness in reconstructing highly undersampled cerebrovascular 4D flow MRI data. As such, we developed a CNN for the reconstruction of kt-accelerated intracranial dual-venc 4D flow MRI (R=5). We found that the CNN showed excellent SSIM values (0.94[0.93-0.95]) and moderate-to-good net flow and peak velocity agreement with the conventionally reconstructed data.
BACKGROUND To investigate the reliability of a novel magnetization transfer ratio (MTR) postproce... more BACKGROUND To investigate the reliability of a novel magnetization transfer ratio (MTR) postprocessing technique for the hippocampus using histogram analysis, and compare results to more established volumetric measurements. This study is conducted in healthy volunteers as a precursor to future applications regarding progressive neurologic diseases, such as Alzheimer's disease. METHODS Eight healthy subjects were scanned twice with interval of 1 week using quantitative magnetic resonance imaging (MRI). Automated pixel-wise analysis was performed for the hippocampal regions of each patient. Reliability was assessed using intraclass correlation coefficients (ICCs), coefficients of variation (COVs), and instrumental standard deviation (ISD). RESULTS Reliable metrics were 25th percentile, median, 75th percentile, peak location, and mean approach (ranges: ICC = .93-.96, COV = 2.71-3.88%, ISD .78-1.01). Histogram peak height had ICC below .7, and a COV above 10%. Volumetric measurements had (ICC = .95-.97, COV = 1.43-3.39%). CONCLUSION Excellent scan-rescan reproducibility (ICC > .9, COV < 10%) was observed for specific MTR histogram metrics and the mean MTR approach. These results are comparable to the volumetric approach. Future studies can examine the possibility that MTR changes precede morphological changes as this study suggests that both MTR and volumetric measurements of the hippocampus can be used as reliable imaging tools.
A single institution, retrospective study of 28 patients with inflammatory carcinoma of the breas... more A single institution, retrospective study of 28 patients with inflammatory carcinoma of the breast treated from 1984 to 1990 was performed. Patients received two to four cycles of cyclophosphamide, doxorubicin and 5-fluorouracil (CDF) and were then evaluated for mastectomy. Mastectomy was accomplished in 26 patients after CDF. In 21 patients, the breast was resectable after the initial doses of chemotherapy and modified radical mastectomy was done. Radiation therapy was given to 16 of the 21 patients after six to nine cycles of postoperative chemotherapy. The remaining five of 26 patients had a marginal response to CDF and underwent preoperative radiation therapy. Local recurrence occurred in four of five patients receiving preoperative radiation, in three of 16 receiving postoperative radiation and in one of five receiving mastectomy without radiation therapy. The overall observed five year survival rate was 18 percent, with a median of 34 months. Neither dermal lymphatic invasion nor estrogen receptor status were statistically significant variables when analyzing patients for local recurrence or survival. Despite poor long term survival results, the combination of induction CDF, mastectomy and postoperative radiation achieved local control in 81 percent of patients.
Pacing and Clinical Electrophysiology, Jan 9, 2020
Introduction: A recent study reported diffuse left ventricular (LV) fibrosis is a predictor of at... more Introduction: A recent study reported diffuse left ventricular (LV) fibrosis is a predictor of atrial fibrillation (AF) recurrence following catheter ablation, by measuring post-contrast cardiac T 1 (an error prone metric as per the 2017 SCMR consensus statement) using an inversion-recovery pulse sequence (an error prone method in arrhythmia) in AF ablation candidates. The purpose of this study was to verify the prior study, by measuring extracellular volume fraction (ECV) (an accurate metric) using a saturation-recovery pulse sequence (accurate method in arrhythmia). Methods and Results: This study examined 100 AF patients (mean age=62±11 years, 69 males and 31 females, 67 paroxysmal [pAF] and 33 persistent [peAF]) who underwent a pre-ablation cardiovascular magnetic resonance (CMR). LV ECV and LA and LV functional parameters were quantified using standard analysis methods. During an average follow-up period of 457±261 days with 4±3 rhythm checks per patient, 72 patients maintained sinus rhythm. Between those who maintained sinus rhythm (n=72) and those who reverted to AF (n=28), the only clinical characteristic that was significantly different was age (60 ± 12 years vs.66 ± 9 years); for CMR metrics, neither mean LV ECV (25.1 ± 3.3% vs. 24.7±3.7%), native LV T1 (1093.8 ± 73.5 ms vs. 1070.2 ± 115.9 ms), LVEF (54.1 ± 11.2% vs. 55.7±7.1%), or LA EDV/BSA (42.4 ± 14.8 mL/m 2 vs. 43.4±19.6 mL/m 2) was not significantly different (p ≥ 0.23). According to Cox regression tests, none of the clinical and imaging variables predicts AF recurrence.
Objective: To evaluate the comparative safety of transcervical and transabdominal chorionic villu... more Objective: To evaluate the comparative safety of transcervical and transabdominal chorionic villus sampling (CVS). Methods: From May 1988 to January 1992, CVS was performed by two operators at 9-12 weeks' gestation in 1048 singleton pregnancies. The sampling method for each patient, transabdominal or transcervical, was chosen primarily based upon placental location; the transabdominal route was used for anterior or fundal location and the transcervical route for posterior placentation. Perinatal outcome was assessed by post-procedure patient telephone contact, mid-gestation ultrasound evaluation, postpartum questionnaire completed by the referring obstetrician, and telephone interview with each patient after delivery. Results: Complete follow-up was available in 1012 cases (97%). Excluding 39 elective abortions, 35 of 973 euploid pregnancies aborted spontaneously. The difference in fetal loss rate between transcervical and transabdominal CVS approached statistical significance (5.2 versus 2.9%; P = .058). Bleeding before CVS (P = .006) and multiple placental aspirations (P = .022) were associated with fetal loss for the entire study group. An interaction between uterine position and sampling method was also indicated; an increased loss rate was associated with transcervical CVS in the presence of uterine retroversion (P = .0017). Conclusion: Despite choosing the preferred CVS method for each patient, an increased loss rate may be associated with transcervical sampling in the presence of uterine retroversion.
Journal of Magnetic Resonance Imaging, Jan 10, 2022
BACKGROUND Evaluation of aortic stiffness by pulse wave velocity (PWV) across the adult lifespan ... more BACKGROUND Evaluation of aortic stiffness by pulse wave velocity (PWV) across the adult lifespan is needed to better understand normal aging in women and men. PURPOSE To characterize PWV in the thoracic aorta using 4D flow MRI in an age- and sex-stratified cohort of healthy adults. STUDY TYPE Retrospective. POPULATION Ninety nine healthy participants (age: 46 ± 15 [19-79] years, 50% female), divided into young adults (<45 years) (N = 48), midlife (45-65 years) (N = 37), and later life (>65 years) (N = 14) groups. FIELD STRENGTH/SEQUENCE 1.5 T or 3 T, 2D cine bSSFP, 4D flow MRI. ASSESSMENT Cardiac functional parameters of end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV) and myocardial mass were assessed by 2D cine bSSFP. PWV and aortic blood flow velocity were assessed by 4D flow MRI. Reproducibility of PWV was evaluated in a subset of nine participants. STATISTICAL TESTS Analysis of variance, Pearson's correlation coefficient (r), linear regression, intraclass correlation coefficient (ICC). A P value < 0.05 was considered statistically significant. RESULTS PWV increased significantly with age (young adults: 5.4 ± 0.9 m/sec, midlife: 7.2 ± 1.1 m/sec, and later life: 9.4 ± 1.8 m/sec) (r = 0.79, slope = 0.09 m/sec/year). PWV did not differ in women and men in entire sample (P = 0.40) or within age groups (young adults: P = 0.83, midlife: P = 0.17, and later life: P = 0.96). PWV was significantly correlated with EDV (r = -0.29), ESV (r = -0.23), SV (r = -0.28), myocardial mass (r = 0.21), and mean aortic blood flow velocity (r = -0.62). In the test-retest subgroup (N = 9), PWV was 6.7 ± 1.5 [4.4-9.3] m/sec and ICC = 0.75. DATA CONCLUSION 4D flow MRI quantified higher aortic PWV with age, by approximately 1 m/sec per decade, and significant differences between young adults, midlife and later life. Reproducibility analysis showed good test-retest agreement. Increased PWV was associated with decline in cardiac function and reduced aortic blood flow velocity. This study demonstrates the utility of 4D flow MRI-derived aortic PWV for studying aging. EVIDENCE LEVEL 2 TECHNICAL EFFICACY: Stage 2.
The analysis of brain imaging data has attracted much attention recently. A popular analysis is t... more The analysis of brain imaging data has attracted much attention recently. A popular analysis is to discover a network representation of brain from the neuroimaging data, where each node denotes a brain region and each edge represents a functional association or structural connection between two brain regions. Motivated by the multi-subject and multi-collection settings in neuroimaging studies, in this paper, we consider brain network discovery under two novel settings: 1) unified setting: Given a collection of subjects, discover a single network that is good for all subjects. 2) contrasting setting: Given two collections of subjects, discover a single network that best discriminates two collections. We show that the existing formulation of graphical Lasso (GLasso) cannot address above problems properly. Two novel models, UGLasso (Unified Graphical Lasso) and CGLasso(Contrasting Graphical Lasso), are proposed to address these two problems respectively. We evaluate our methods on synthetic data and two realworld functional magnetic resonance imaging (fMRI) datasets. Empirical results demonstrate the effectiveness of the proposed methods.
Journal of Magnetic Resonance Imaging, Oct 25, 2021
BackgroundCardiac magnetic resonance imaging (MRI) is becoming an alternative to right heart cath... more BackgroundCardiac magnetic resonance imaging (MRI) is becoming an alternative to right heart catheterization (RHC) for evaluating pulmonary hypertension (PH). A need exists to further evaluate cardiac MRI's ability to characterize PH.PurposeTo evaluate the potential for four‐dimensional (4D) flow MRI‐derived pulmonary artery velocities to characterize PH.Study TypeProspective case–control.PopulationFifty‐four PH patients (56% female); 25 controls (36% female).Field Strength/Sequence1.5 T; gradient recalled echo 4D flow and balanced steady‐state free precession cardiac cine.AssessmentRHC was used to derive patients' pulmonary vascular resistance (PVR). 4D flow measured blood velocities at the main, left, and right pulmonary arteries (MPA, LPA, and RPA); cine measured ejection fraction, end diastolic, and end systolic volumes (EF, EDV, and ESV). EDV and ESV were normalized (indexed) to body surface area (ESVI and EDVI). Parameters were evaluated between, and within, PH subgroups: pulmonary arterial hypertension (PAH); PH due to left heart disease (PH‐LHD)/chronic lung disease (PH‐CLD)/or chronic thrombo‐emboli (CTE‐PH).Statistical TestsAnalysis of variance and Kruskal–Wallis tests compared parameters between subgroups. Pearson's r assessed velocity, PVR, and volume correlations. Significance definition: P &lt; 0.05.ResultsPAH peak and mean velocities were significantly lower than in controls at the LPA (36 ± 12 cm/second and 20 ± 4 cm/second vs. 59 ± 15 cm/second and 32 ± 9 cm/second). At the RPA, mean velocities were significantly lower in PAH vs. controls (27 ± 6 cm/second vs. 40 ± 9 cm/second). Peak velocities significantly correlated with right ventricular EF at the MPA (r = 0.286), RPA (r = 0.400), and LPA (r = 0.401). Peak velocity significantly correlated with right ventricular ESVI at the RPA (r = −0.355) and LPA (r = −0.316). Significant correlations between peak velocities and PVR were moderate at the LPA in PAH (r = −0.641) and in PH‐LHD (r = −0.606) patients, and at the MPA in PH‐CLD (r = −0.728). CTE‐PH showed non‐significant correlations between peak velocity and PVR at all locations.Data ConclusionPreliminary findings suggest 4D flow can identify PAH and track PVR changes.Level of Evidence1Technical EfficacyStage 5
The primary objectives of this report were, first, to determine the number and incidence of compl... more The primary objectives of this report were, first, to determine the number and incidence of complications of transsphenoidal surgery performed by a cross-section of neurosurgeons in the United States and, second, to ascertain the influence of the surgeon&amp;amp;#39;s experience with the procedure on the occurrence of these complications. The secondary objective was to review complications of transsphenoidal surgery from the standpoint of their causation, treatment, and prevention. Questionnaires regarding 14 specific complications of transsphenoidal surgery were mailed to 3172 neurosurgeons. The data reported were analyzed from the 958 respondents (82%) who reported performing the operation. The neurosurgeons surveyed were asked to estimate the number of transsphenoidal operations performed, to identify any complications observed, and to estimate the percentage of operations that had resulted in any of the 14 specific complications. The 958 respondents were placed into three experience groups, based on the number of transsphenoidal operations performed. The data were analyzed by using chi 2 tests and Spearman correlation coefficients. The secondary objectives were met through a detailed review of the literature, in light of our experience. Of the respondents, 87.3% reported having performed &amp;amp;lt; 200 operations and 9.7% reported 200 to 500 previous operations. The remaining 3% reported more than 500 previous operations. More extensive previous experience with transsphenoidal surgery was associated with a greater likelihood of having witnessed each specific complication. The mean operative mortality rate for all three groups was 0.9%. Anterior pituitary insufficiency (19.4%) and diabetes insipidus (17.8%) were complications with the highest incidence of occurrence. The overall incidence of cerebrospinal fluid fistulas was 3.9%. Other significant complications, such as carotid artery injuries, hypothalamic injuries, loss of vision, and meningitis, occurred with incidence rates between 1 and 2%. An inverse relationship was found between the experience group and the likelihood of complications, as indicated by significant negative Spearman correlation coefficients for all but 2 of the 14 complications listed in the survey (P &amp;amp;lt; 0.05). Thus, increased experience with transsphenoidal surgery seems to be associated with a decreased percentage of operations resulting in complications. Some caution should be exercised in interpreting these data, because they are based on the respondents&amp;amp;#39; estimates. Transsphenoidal surgery seems to be a reasonably safe procedure, with a mortality rate of less than 1%. However, a significant number of complications do occur. The incidence of these complications seems to be higher, with statistical significance, in the hands of less experienced surgeons. The learning curve seems to be relatively shallow, because a statistically significantly decreased incidence of morbidity and death could be documented after 200 and even 500 transsphenoidal operations. Better understanding of the indications for transsphenoidal surgery and improved familiarity with the regional anatomy should further lower the incidence of death and morbidity resulting from this procedure in the hands of all neurosurgeons.
Mining from neuroimaging data is becoming increasingly popular in the eld of healthcare and bioin... more Mining from neuroimaging data is becoming increasingly popular in the eld of healthcare and bioinformatics, due to its potential to discover clinically meaningful structure pa erns that could facilitate the understanding and diagnosis of neurological and neuropsychiatric disorders. Most recent research concentrates on applying subgraph mining techniques to discover connected subgraph patterns in the brain network. However, the underlying brain network structure is complicated. As a shallow linear model, subgraph mining cannot capture the highly non-linear structures, resulting in sub-optimal pa erns. erefore, how to learn representations that can capture the highly non-linearity of brain networks and preserve the underlying structures is a critical problem. In this paper, we propose a Structural Deep Brain Network mining method, namely SDBN, to learn highly non-linear and structurepreserving representations of brain networks. Speci cally, we rst introduce a novel graph reordering approach based on module identi cation, which rearranges the order of the nodes to preserve the modular structure of the graph. Next, we perform structural augmentation to further enhance the spatial information of the reordered graph. en we propose a deep feature learning framework for combining supervised learning and unsupervised learning in a small-scale se ing, by augmenting Convolutional Neural Network (CNN) with decoding pathways for reconstruction. With the help of the multiple layers of non-linear mapping, the proposed SDBN approach can capture the highly non-linear structure of brain networks. Further, it has be er generalization capability for high-dimensional brain networks and works well even for small sample learning. Bene t from CNN's task-oriented learning style, the learned hierarchical representation is meaningful for the clinical * Corresponding author. is work was done while the author was at the University of Illinois at Chicago.
Before the introduction of antiretroviral therapy, human immunodeficiency virus (HIV) infection w... more Before the introduction of antiretroviral therapy, human immunodeficiency virus (HIV) infection was often accompanied by central nervous system (CNS) opportunistic infections and HIV encephalopathy marked by profound structural and functional alterations detectable with neuroimaging. Treatment with antiretroviral therapy nearly eliminated CNS opportunistic infections, while neuropsychiatric impairment and peripheral nerve and organ damage have persisted among virally suppressed people with HIV (PWH), suggesting ongoing brain injury. Neuroimaging research must use methods sensitive for detecting subtle HIV-associated brain structural and functional abnormalities, while allowing for adjustments for potential confounders, such as age, sex, substance use, hepatitis C coinfection, cardiovascular risk, and others. Here, we review existing and emerging neuroimaging tools that demonstrated promise in detecting markers of HIV-associated brain pathology and explore strategies to study the impact of potential confounding factors on these brain measures. We emphasize neuroimaging approaches that may be used in parallel to gather complementary information, allowing efficient detection and interpretation of altered brain structure and function associated with suboptimal clinical outcomes among virally suppressed PWH. We examine the advantages of each imaging modality and systematic approaches in study design and analysis. We also consider advantages of combining experimental and statistical control techniques to improve sensitivity and specificity of biotype identification and explore the costs and benefits of aggregating data from multiple studies to achieve larger sample sizes, enabling use of emerging methods for combining and analyzing large, multifaceted data sets. Many of the topics addressed in this article were discussed at the National Institute of Mental Health meeting "Biotypes of CNS Complications in People Living with HIV," held in October 2021, and are part of ongoing research initiatives to define the role of neuroimaging in emerging alternative approaches to identifying biotypes of CNS complications in PWH. An outcome of these considerations may be the development of a common neuroimaging protocol available for researchers to use in future studies examining neurological changes in the brains of PWH.
This study aimed to examine cognitive function in acute/early HIV infection over the subsequent 2... more This study aimed to examine cognitive function in acute/early HIV infection over the subsequent 2 years. Fifty-six HIV+ subjects and 21 seronegative participants of the Chicago Early HIV Infection Study were evaluated using a comprehensive neuropsychological assessment at study enrollment and at 2-year follow-up. Cognitive performance measures were compared in the groups using t tests and mixed-effect models. Patterns of relationship with clinical measures were determined between cognitive function and clinical status markers using Spearman&amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;amp;#39;s correlations. At the initial timepoint, the HIV group demonstrated significantly weaker performance on measures of verbal memory, visual memory, psychomotor speed, motor speed, and executive function. A similar pattern was found when cognitive function was examined at follow-up and across both timepoints. The HIV subjects had generally weaker performance on psychomotor speed, executive function, motor speed, visual memory, and verbal memory. The rate of decline in cognitive function across the 2-year follow-up period did not differ between groups. Correlations between clinical status markers and cognitive function at both timepoints showed weaker performance associated with increased disease burden. Neurocognitive difficulty in chronic HIV infection may have very early onset and reflect consequences of initial brain viral invasion and neuroinflammation during the intense, uncontrolled viremia of acute HIV infection. Further characterization of the changes occurring in initial stages of infection and the risk and protective factors for cognitive function could inform new strategies for neuroprotection.
Journal of Vascular and Interventional Radiology, Feb 1, 2010
To assess the feasibility and long-term clinical outcomes of percutaneous image-guided, multi-pro... more To assess the feasibility and long-term clinical outcomes of percutaneous image-guided, multi-probe cryotherapy in patients refusing surgery for breast cancer. Materials and Methods: After breast magnetic resonance (MR) imaging and thorough consultation, patient consents were obtained for institutional review board approved breast cryotherapy. In 13 cryotherapy sessions, 25 breast cancer foci were treated in 13 patients, Stages I-IV, using multiple 2.4mm cryoprobes. Using only local anesthesia with mild sedation, ultrasound (US) guidance alone was used in 6 patients and 7 patients required both computed tomography (CT) and US to better define ice margins. Saline injections protected the skin and/or chest wall. MR and/or clinical follow-up were available for up to 65 months post cryotherapy. Results: MR helped define multifocal recurrences and initial tumors as well as careful follow up of avascular ablation zones covering all initial tumors. US produced sufficient ice visualization for small tumors while CT confirmed posterior margins and overall ice extent for larger regions requiring Ͼ3 probes or deep lesions. Pretreatment breast tumor diameter was 1.7 ϩ 1.2 cm (range: 0.5-5.8 cm) and an average of 3.3 cryoprobes produced ice diameters of 5.2 ϩ 2.2 cm (range: 2.0-10.0 cm). No significant complications were noted. All patients reported minimal discomfort and satisfaction with cosmetic results. Biopsies at the margins of the cryotherapy site immediately after the procedure and at the cryotherapy site in follow-up were all negative. No local recurrences have been noted at 18 months average followup. Conclusion: In conjunction with thorough pre and postablation MRI, clinic-based multi-probe cryotherapy of breast cancer for local control and/or potential cure provides safe, effective breast conservation with minimal discomfort.
Magnetization transfer imaging (MT) may have considerable promise for early detection and monitor... more Magnetization transfer imaging (MT) may have considerable promise for early detection and monitoring of subtle brain changes before they are apparent on conventional magnetic resonance images. At 3 Tesla (T), MT affords higher resolution and increased tissue contrast associated with macromolecules. The reliability and reproducibility of a new high-resolution MT strategy were assessed in brain images acquired from 9 healthy subjects. Repeated measures were taken for 12 brain regions of interest (ROIs): genu, splenium, and the left and right hemispheres of the hippocampus, caudate, putamen, thalamus, and cerebral white matter. Spearman's correlation coefficient, coefficient of variation, and intraclass correlation coefficient (ICC) were computed. Multivariate mixed-effects regression models were used to fit the mean ROI values and to test the significance of the effects due to region, subject, observer, time, and manual repetition. A sensitivity analysis of various model specifications and the corresponding ICCs was conducted. Our statistical methods may be generalized to many similar evaluative studies of the reliability and reproducibility of various imaging modalities.
Cerebrovascular dual-venc 4D Flow MRI has utility in evaluating intracranial hemodynamics - benef... more Cerebrovascular dual-venc 4D Flow MRI has utility in evaluating intracranial hemodynamics - benefitting from full circle of Willis (CoW) volumetric coverage, time-resolved, three-directional velocity encoding, and velocity dynamic range. Analysis of hemodynamic measures was completed using a comprehensive, semi-automated vessel identification and segmentation workflow for 10 subjects who underwent two scans within 30 days. Test-retest analysis of cerebrovascular dual-venc 4D Flow MRI and measures of CoW 3D flow dynamics demonstrated excellent repeatability for flow and peak velocity measures and lower reproducibility for pulsatility and resistive indices. Future work may include a larger cohort or more advanced analysis of velocity-derived parameters.
Mechanisms underlying heart-brain hemodynamic coupling and effects on the brain remain unclear du... more Mechanisms underlying heart-brain hemodynamic coupling and effects on the brain remain unclear due to challenges of measuring both heart and brain in a single MRI exam. We have developed a comprehensive MRI protocol that incorporates 1) 4D flow MRI of chest, 2) 4D flow MRI of head and 3) structural neuroimaging into one MRI exam to assess cardiovascular and cerebrovascular flow as well as white matter lesions and brain atrophy. This study demonstrates the utility of heart-brain MRI as viable new tool for in vivo evaluation of hemodynamic coupling along the entire heart-brain pathway.
4D flow MRI provides a comprehensive assessment of hemodynamics through the 3D visualization and ... more 4D flow MRI provides a comprehensive assessment of hemodynamics through the 3D visualization and quantification of blood flow. However, its current clinical usage is hindered by long scan times. Recent developments have shown deep learning to be highly effective in accelerating image reconstruction, but no study has shown the its effectiveness in reconstructing highly undersampled cerebrovascular 4D flow MRI data. As such, we developed a CNN for the reconstruction of kt-accelerated intracranial dual-venc 4D flow MRI (R=5). We found that the CNN showed excellent SSIM values (0.94[0.93-0.95]) and moderate-to-good net flow and peak velocity agreement with the conventionally reconstructed data.
BACKGROUND To investigate the reliability of a novel magnetization transfer ratio (MTR) postproce... more BACKGROUND To investigate the reliability of a novel magnetization transfer ratio (MTR) postprocessing technique for the hippocampus using histogram analysis, and compare results to more established volumetric measurements. This study is conducted in healthy volunteers as a precursor to future applications regarding progressive neurologic diseases, such as Alzheimer's disease. METHODS Eight healthy subjects were scanned twice with interval of 1 week using quantitative magnetic resonance imaging (MRI). Automated pixel-wise analysis was performed for the hippocampal regions of each patient. Reliability was assessed using intraclass correlation coefficients (ICCs), coefficients of variation (COVs), and instrumental standard deviation (ISD). RESULTS Reliable metrics were 25th percentile, median, 75th percentile, peak location, and mean approach (ranges: ICC = .93-.96, COV = 2.71-3.88%, ISD .78-1.01). Histogram peak height had ICC below .7, and a COV above 10%. Volumetric measurements had (ICC = .95-.97, COV = 1.43-3.39%). CONCLUSION Excellent scan-rescan reproducibility (ICC > .9, COV < 10%) was observed for specific MTR histogram metrics and the mean MTR approach. These results are comparable to the volumetric approach. Future studies can examine the possibility that MTR changes precede morphological changes as this study suggests that both MTR and volumetric measurements of the hippocampus can be used as reliable imaging tools.
A single institution, retrospective study of 28 patients with inflammatory carcinoma of the breas... more A single institution, retrospective study of 28 patients with inflammatory carcinoma of the breast treated from 1984 to 1990 was performed. Patients received two to four cycles of cyclophosphamide, doxorubicin and 5-fluorouracil (CDF) and were then evaluated for mastectomy. Mastectomy was accomplished in 26 patients after CDF. In 21 patients, the breast was resectable after the initial doses of chemotherapy and modified radical mastectomy was done. Radiation therapy was given to 16 of the 21 patients after six to nine cycles of postoperative chemotherapy. The remaining five of 26 patients had a marginal response to CDF and underwent preoperative radiation therapy. Local recurrence occurred in four of five patients receiving preoperative radiation, in three of 16 receiving postoperative radiation and in one of five receiving mastectomy without radiation therapy. The overall observed five year survival rate was 18 percent, with a median of 34 months. Neither dermal lymphatic invasion nor estrogen receptor status were statistically significant variables when analyzing patients for local recurrence or survival. Despite poor long term survival results, the combination of induction CDF, mastectomy and postoperative radiation achieved local control in 81 percent of patients.
Pacing and Clinical Electrophysiology, Jan 9, 2020
Introduction: A recent study reported diffuse left ventricular (LV) fibrosis is a predictor of at... more Introduction: A recent study reported diffuse left ventricular (LV) fibrosis is a predictor of atrial fibrillation (AF) recurrence following catheter ablation, by measuring post-contrast cardiac T 1 (an error prone metric as per the 2017 SCMR consensus statement) using an inversion-recovery pulse sequence (an error prone method in arrhythmia) in AF ablation candidates. The purpose of this study was to verify the prior study, by measuring extracellular volume fraction (ECV) (an accurate metric) using a saturation-recovery pulse sequence (accurate method in arrhythmia). Methods and Results: This study examined 100 AF patients (mean age=62±11 years, 69 males and 31 females, 67 paroxysmal [pAF] and 33 persistent [peAF]) who underwent a pre-ablation cardiovascular magnetic resonance (CMR). LV ECV and LA and LV functional parameters were quantified using standard analysis methods. During an average follow-up period of 457±261 days with 4±3 rhythm checks per patient, 72 patients maintained sinus rhythm. Between those who maintained sinus rhythm (n=72) and those who reverted to AF (n=28), the only clinical characteristic that was significantly different was age (60 ± 12 years vs.66 ± 9 years); for CMR metrics, neither mean LV ECV (25.1 ± 3.3% vs. 24.7±3.7%), native LV T1 (1093.8 ± 73.5 ms vs. 1070.2 ± 115.9 ms), LVEF (54.1 ± 11.2% vs. 55.7±7.1%), or LA EDV/BSA (42.4 ± 14.8 mL/m 2 vs. 43.4±19.6 mL/m 2) was not significantly different (p ≥ 0.23). According to Cox regression tests, none of the clinical and imaging variables predicts AF recurrence.
Objective: To evaluate the comparative safety of transcervical and transabdominal chorionic villu... more Objective: To evaluate the comparative safety of transcervical and transabdominal chorionic villus sampling (CVS). Methods: From May 1988 to January 1992, CVS was performed by two operators at 9-12 weeks' gestation in 1048 singleton pregnancies. The sampling method for each patient, transabdominal or transcervical, was chosen primarily based upon placental location; the transabdominal route was used for anterior or fundal location and the transcervical route for posterior placentation. Perinatal outcome was assessed by post-procedure patient telephone contact, mid-gestation ultrasound evaluation, postpartum questionnaire completed by the referring obstetrician, and telephone interview with each patient after delivery. Results: Complete follow-up was available in 1012 cases (97%). Excluding 39 elective abortions, 35 of 973 euploid pregnancies aborted spontaneously. The difference in fetal loss rate between transcervical and transabdominal CVS approached statistical significance (5.2 versus 2.9%; P = .058). Bleeding before CVS (P = .006) and multiple placental aspirations (P = .022) were associated with fetal loss for the entire study group. An interaction between uterine position and sampling method was also indicated; an increased loss rate was associated with transcervical CVS in the presence of uterine retroversion (P = .0017). Conclusion: Despite choosing the preferred CVS method for each patient, an increased loss rate may be associated with transcervical sampling in the presence of uterine retroversion.
Journal of Magnetic Resonance Imaging, Jan 10, 2022
BACKGROUND Evaluation of aortic stiffness by pulse wave velocity (PWV) across the adult lifespan ... more BACKGROUND Evaluation of aortic stiffness by pulse wave velocity (PWV) across the adult lifespan is needed to better understand normal aging in women and men. PURPOSE To characterize PWV in the thoracic aorta using 4D flow MRI in an age- and sex-stratified cohort of healthy adults. STUDY TYPE Retrospective. POPULATION Ninety nine healthy participants (age: 46 ± 15 [19-79] years, 50% female), divided into young adults (<45 years) (N = 48), midlife (45-65 years) (N = 37), and later life (>65 years) (N = 14) groups. FIELD STRENGTH/SEQUENCE 1.5 T or 3 T, 2D cine bSSFP, 4D flow MRI. ASSESSMENT Cardiac functional parameters of end-diastolic volume (EDV), end-systolic volume (ESV), stroke volume (SV) and myocardial mass were assessed by 2D cine bSSFP. PWV and aortic blood flow velocity were assessed by 4D flow MRI. Reproducibility of PWV was evaluated in a subset of nine participants. STATISTICAL TESTS Analysis of variance, Pearson's correlation coefficient (r), linear regression, intraclass correlation coefficient (ICC). A P value < 0.05 was considered statistically significant. RESULTS PWV increased significantly with age (young adults: 5.4 ± 0.9 m/sec, midlife: 7.2 ± 1.1 m/sec, and later life: 9.4 ± 1.8 m/sec) (r = 0.79, slope = 0.09 m/sec/year). PWV did not differ in women and men in entire sample (P = 0.40) or within age groups (young adults: P = 0.83, midlife: P = 0.17, and later life: P = 0.96). PWV was significantly correlated with EDV (r = -0.29), ESV (r = -0.23), SV (r = -0.28), myocardial mass (r = 0.21), and mean aortic blood flow velocity (r = -0.62). In the test-retest subgroup (N = 9), PWV was 6.7 ± 1.5 [4.4-9.3] m/sec and ICC = 0.75. DATA CONCLUSION 4D flow MRI quantified higher aortic PWV with age, by approximately 1 m/sec per decade, and significant differences between young adults, midlife and later life. Reproducibility analysis showed good test-retest agreement. Increased PWV was associated with decline in cardiac function and reduced aortic blood flow velocity. This study demonstrates the utility of 4D flow MRI-derived aortic PWV for studying aging. EVIDENCE LEVEL 2 TECHNICAL EFFICACY: Stage 2.
The analysis of brain imaging data has attracted much attention recently. A popular analysis is t... more The analysis of brain imaging data has attracted much attention recently. A popular analysis is to discover a network representation of brain from the neuroimaging data, where each node denotes a brain region and each edge represents a functional association or structural connection between two brain regions. Motivated by the multi-subject and multi-collection settings in neuroimaging studies, in this paper, we consider brain network discovery under two novel settings: 1) unified setting: Given a collection of subjects, discover a single network that is good for all subjects. 2) contrasting setting: Given two collections of subjects, discover a single network that best discriminates two collections. We show that the existing formulation of graphical Lasso (GLasso) cannot address above problems properly. Two novel models, UGLasso (Unified Graphical Lasso) and CGLasso(Contrasting Graphical Lasso), are proposed to address these two problems respectively. We evaluate our methods on synthetic data and two realworld functional magnetic resonance imaging (fMRI) datasets. Empirical results demonstrate the effectiveness of the proposed methods.
Journal of Magnetic Resonance Imaging, Oct 25, 2021
BackgroundCardiac magnetic resonance imaging (MRI) is becoming an alternative to right heart cath... more BackgroundCardiac magnetic resonance imaging (MRI) is becoming an alternative to right heart catheterization (RHC) for evaluating pulmonary hypertension (PH). A need exists to further evaluate cardiac MRI's ability to characterize PH.PurposeTo evaluate the potential for four‐dimensional (4D) flow MRI‐derived pulmonary artery velocities to characterize PH.Study TypeProspective case–control.PopulationFifty‐four PH patients (56% female); 25 controls (36% female).Field Strength/Sequence1.5 T; gradient recalled echo 4D flow and balanced steady‐state free precession cardiac cine.AssessmentRHC was used to derive patients' pulmonary vascular resistance (PVR). 4D flow measured blood velocities at the main, left, and right pulmonary arteries (MPA, LPA, and RPA); cine measured ejection fraction, end diastolic, and end systolic volumes (EF, EDV, and ESV). EDV and ESV were normalized (indexed) to body surface area (ESVI and EDVI). Parameters were evaluated between, and within, PH subgroups: pulmonary arterial hypertension (PAH); PH due to left heart disease (PH‐LHD)/chronic lung disease (PH‐CLD)/or chronic thrombo‐emboli (CTE‐PH).Statistical TestsAnalysis of variance and Kruskal–Wallis tests compared parameters between subgroups. Pearson's r assessed velocity, PVR, and volume correlations. Significance definition: P &lt; 0.05.ResultsPAH peak and mean velocities were significantly lower than in controls at the LPA (36 ± 12 cm/second and 20 ± 4 cm/second vs. 59 ± 15 cm/second and 32 ± 9 cm/second). At the RPA, mean velocities were significantly lower in PAH vs. controls (27 ± 6 cm/second vs. 40 ± 9 cm/second). Peak velocities significantly correlated with right ventricular EF at the MPA (r = 0.286), RPA (r = 0.400), and LPA (r = 0.401). Peak velocity significantly correlated with right ventricular ESVI at the RPA (r = −0.355) and LPA (r = −0.316). Significant correlations between peak velocities and PVR were moderate at the LPA in PAH (r = −0.641) and in PH‐LHD (r = −0.606) patients, and at the MPA in PH‐CLD (r = −0.728). CTE‐PH showed non‐significant correlations between peak velocity and PVR at all locations.Data ConclusionPreliminary findings suggest 4D flow can identify PAH and track PVR changes.Level of Evidence1Technical EfficacyStage 5
Uploads
Papers by Ann Ragin