Faculty Bibliography

Computer-aided detection (CADe) of pulmonary nodules from computer tomography (CT) scans is critical for assisting radiologists to identify lung lesions at an early stage. In this paper, we propose a novel approach for CADe of lung nodules using a two-stage vector quantization (VQ) scheme. The first-stage VQ aims to extract lung from the chest volume, while the second-stage VQ is designed to extract initial nodule candidates (INCs) within the lung volume. Then rule-based expert filtering is employed to prune obvious FPs from INCs, and the commonly-used support vector machine (SVM) classifier is adopted to further reduce the FPs. The proposed system was validated on 100 CT scans randomly selected from the 262 scans that have at least one juxta-pleural nodule annotation in the publicly available database Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI). The two-stage VQ only missed 2 out of the 207 nodules at agreement level 1, and the INCs detection for each scan took about 30 seconds in average. Expert filtering reduced FPs more than 18 times, while maintaining a sensitivity of 93.24%. As it is trivial to distinguish INCs attached to pleural wall versus not on wall, we investigated the feasibility of training different SVM classifiers to further reduce FPs from these two kinds of INCs. Experiment results indicated that SVM classification over the entire set of INCs was in favor of, where the optimal operating of our CADe system achieved a sensitivity of 89.4% at a specificity of 86.8%.

OBJECTIVES: The aim of this study was to evaluate the prognostic value of pretreatment F-fluorodeoxyglucose PET/computed tomography (CT), particularly in the assessment of metabolic tumor burden markers such as metabolic tumor volume (MTV) and total lesion glycolysis (TLG), with respect to clinical outcomes in stage I non-small-cell lung cancer (NSCLC) patients undergoing stereotactic body radiation therapy (SBRT). METHODS: This retrospective study evaluated 50 patients who underwent SBRT for stage I NSCLC from May 2007 to December 2012. The maximum standardized uptake value (SUVmax), average SUV (SUVavg), MTV, and TLG were measured from the PET/CT scan. The study population was dichotomized at the median into high and low groups. Kaplan-Meier log-rank tests were then used to compare high with low PET/CT parameter groups, and univariate Cox proportional hazards regression analysis was carried out to identify predictors of overall survival. RESULTS: The 2-year local control rate was 93.7%. After a median follow-up of 25.1 months, the 2-year overall survival was 79.3%. Eight patients (16%) had disease recurrence. There were three local failures (6%), three mediastinal failures (6%), and six cases of distant metastases (12%). Both Kaplan-Meier actuarial analysis and Cox proportional hazards regression found no correlation between SUVmax, SUVavg, MTV, and TLG and overall survival. CONCLUSION: Standard PET/CT measures, such as SUVmax, as well as newer measures of metabolic tumor burden, such as MTV and TLG, were not correlated with overall survival in our study population of stage I NSCLC patients undergoing SBRT. Larger studies with longer follow-up periods are needed to confirm these results.

BACKGROUND: CT-guided needle biopsy of lung nodules requires breath holding during needle placement, and is thus inapplicable to patients with difficulty in holding breath. METHODS: A robotic needle biopsy technique is introduced, adapting to the patient respiratory pattern and using a robot manipulator to drive the needle towards a moving lung nodule. Based on the nodule respiratory motion model, needle placement is planned to follow an optimal timing and path, and is triggered based on the respiratory phase tracking. An experimental system has been created to study robotic needle placement. RESULTS: Preliminary phantom tests were conducted based on three representative clinically-collected lung nodule motion paths, using an 18-gauge coaxial needle set. 300 needle paths were implemented. Robotic needle driving was accomplished within 0.4 s (a typical respiratory phase), and resulted in a needle placement accuracy of 0.5 mm with a standard deviation about 0.1 mm over the non-resistance paths. CONCLUSION: The proposed robotic needle placement technique is promising for accurately biopsying lung nodules under respiratory motion and those with very small sizes.

Increased public and regulatory scrutiny of imaging-related radiation exposure requires familiarity with current dose-monitoring techniques and best practices. CT-related ionizing radiation exposure has been cited as the largest and fastest growing source of population-wide iatrogenic ionizing radiation exposure. Upcoming federal regulations require imaging centers to familiarize themselves with available dose-monitoring techniques and implement comprehensive strategies to track patient dose, with particular emphasis on CT. Because of institution-specific and vendor-specific technologies, there are significant barriers to adoption and implementation. In this article, the authors outline the core components of a universal dose-monitoring strategy and detail a few of the many available commercial platforms. In addition, the authors introduce a cloud-based hybrid model dose-tracking system with the goal of rapid implementation, multicenter scalability, real-time dose feedback for technologists, cumulative dose monitoring, and optional dose communication to patients and into the record; doing so results in improved patient loyalty, referring physician satisfaction, and opportunity for repeat business.

BACKGROUND: Non-small cell lung cancer (NSCLC) has a predilection to occur in emphysematous lungs. The relation between the regional severity of emphysema and the location of NSCLC as well as long-term survival has been poorly studied. METHODS: Computed tomography (CT) scans of 153 patients with biopsy-proven stage I NSCLC diagnosed between 2001 and 2006 were assigned an emphysema severity score in four regions of the lung. The location of the cancer was compared with the severity of emphysema in that region. Survival was also analyzed. RESULTS: Thirty-nine patients had no emphysema documented on CT scan and 114 did. The most common location of cancer was the right upper quadrant with 37% of cancers, followed by the left upper quadrant with 23% of cancers. Twenty-two percent of the cancers occurred in the right lower quadrant, and only 12% were in the left lower quadrant. There is a strong association for cancer being located in the area with the highest degree of emphysema (P < 0.001). Emphysema severity score was also associated with long-term survival (log-rank P = 0.03). CONCLUSIONS: The regional severity of emphysema assessed via a visual scale using CT appears to be associated with the location of lung cancer and is an independent predictor of long-term survival.

PURPOSE: This study aimed to assess the imaging findings in patients with pathologically proven carcinoid tumors and determine if SUV can help to differentiate typical from atypical (more aggressive) pulmonary carcinoid tumors. PATIENTS AND METHODS: A retrospective review of patients with a biopsy-proven diagnosis of a pulmonary carcinoid tumor at our institution from 2002 to 2010 that had a preoperative PET scan was performed after institutional review board approval was obtained. PET results, including SUV uptake and location, were recorded as well as all data from pathology reports. Carcinoids were considered to be more aggressive if they showed pathological diagnosis consistent with atypical carcinoid, lymph node invasion, poor histological grade (poorly differentiated), or evidence of systemic metastases. Atypical carcinoid pathology consisted of focal necrosis or a higher mitotic index (2-10 per square millimeter) with features of nests, trabeculae, pleomorphic cells, or dense hyperchromasia. SUV uptake was then evaluated and compared between the typical and atypical carcinoid groups using nonparametric statistical methods. RESULTS: We identified 29 patients from 2002 to 2010 at our institution with a pathological diagnosis of pulmonary carcinoid. Twenty-three were histopathologically typical, and the other 6 showed atypia. Mean (SD) nodule size was 2.4 (1.3) cm in the typical group versus 5.0 (3.2) cm in the atypical group (P = 0.065). Mean (SD) SUV uptake in the typical carcinoid group was 2.7 (1.6) and in the atypical group the SUV was 8.1 (4.1) (P < 0.01). A cutoff SUV of 6 or greater is predictive of malignancy (odds ratio, 23.6; P < 0.01), as well as a nodule size of 3.5 cm or greater (odds ratio, 5.1; P = 0.024). CONCLUSIONS: Preoperative PET imaging result is frequently positive in carcinoid tumors, and the biological behavior correlates well with SUV; however, size is not as strong of a predictor of malignancy. Size of 3.5 cm or greater and SUV of 6 or greater have a predictive value of greater than 95% for malignant histology.

Lung tumors represent a major health impact globally. Pulmonary cryoablation treatment as a palliative measure for patients with non-operable pulmonary lesions has gained popularity over the last decade. With increasing case load and patients medical status becoming more complex, preparation for pulmonary cryoablation requires the implementation of an enhanced perioperative anesthetic plan. Current literature as well as our institutional experience shows that this patient population presents with multiple comorbidities raising the challenge of providing anesthetic care. These procedures are done under challenging conditions with limited resources and in remote locations in the hospital. A team approach by the anesthesiologist, thoracic surgeon, and interventional radiologist is critical to the success of this treatment. The present review examines our institution’s anesthetic management of percutaneous cryoablation treatment (PCT) in the treatment of non-operable lung cancer and metastases. The objective of this article is to review the current literature guidelines and to discuss our retrospective institutional experience in anesthesia management of PCT procedures

(18)F-FDG PET/CT was compared with non-contrast chest CT in monitoring for recurrence 1-year after lobectomy of stage 1 non-small-cell lung cancer (NSCLC). For surveillance after treatment with curative intent, current (April 2012) National Comprehensive Cancer network guidelines recommend chest CT with or without contrast every 6-12 months for 2 years, then non-contrast chest CT annually. PET/CT is not currently indicated for routine follow-up. One hundred patients receiving surveillance PET/CT 1-year after lobectomy for the treatment of stage 1a or 1b NSCLC were included in the study. Exclusion criteria included the presence or interval diagnosis of a second malignancy, or surgical treatment more radical than single lobectomy. The non-contrast CT obtained from the 1-year PET/CT was interpreted by an experienced chest radiologist blinded to the PET/CT for evidence of recurrence using the following findings: pulmonary nodule, pleural effusion, pleural mass, adenopathy, and extrathoracic mass. The ecision about recurrence was made solely from the non-contrast CT without PET/CT findings. This was compared with the determination made with PET/CT. The reference standard for determination of recurrence was the multi-disciplinary tumor board who had access to all imaging and clinical data. Recurrence at 1 year was documented in 16 of 90 patients. All 16 recurrences were documented with PET/CT and 9 were found with non-contrast CT. Five of the 7 recurrences missed with non-contrast CT were extrathoracic metastases. Sensitivity of CT and PET/CT for recurrence was 56.3% and 100%, respectively (p = 0.015). Specificity of CT and PET/CT for recurrence was 95.9% and 93.2%, respectively (p = 0.62).

To distinguish malignant pulmonary nodules from benign ones is of much importance in computer-aided diagnosis of lung diseases. Compared to many previous methods which are based on shape or growth assessing of nodules, this proposed three-dimensional (3D) texture feature based approach extracted fifty kinds of 3D textural features from gray level, gradient and curvature co-occurrence matrix, and more derivatives of the volume data of the nodules. To evaluate the presented approach, the Lung Image Database Consortium public database was downloaded. Each case of the database contains an annotation file, which indicates the diagnosis results from up to four radiologists. In order to relieve partial-volume effect, interpolation process was carried out to those volume data with image slice thickness more than 1mm, and thus we had categorized the downloaded datasets to five groups to validate the proposed approach, one group of thickness less than 1mm, two types of thickness range from 1mm to 1.25mm and greater than 1.25mm (each type contains two groups, one with interpolation and the other without). Since support vector machine is based on statistical learning theory and aims to learn for predicting future data, so it was chosen as the classifier to perform the differentiation task. The measure on the performance was based on the area under the curve (AUC) of Receiver Operating Characteristics. From 284 nodules (122 malignant and 162 benign ones), the validation experiments reported a mean of 0.9051 and standard deviation of 0.0397 for the AUC value on average over 100 randomizations

The intravenous, intramuscular or intraperitoneal administration of water solubilized graphene nanoparticles for biomedical applications will result in their interaction with the hematological components and vasculature. Herein, we have investigated the effects of dextran functionalized graphene nanoplatelets (GNP-Dex) on histamine release, platelet activation, immune activation, blood cell hemolysis in vitro, and vasoactivity in vivo. The results indicate that GNP-Dex formulations prevented histamine release from activated RBL-2H3 rat mast cells, and at concentrations >/= 7 mg/ml, showed a 12-20% increase in levels of complement proteins. Cytokine (TNF-Alpha and IL-10) levels remained within normal range. GNP-Dex formulations did not cause platelet activation or blood cell hemolysis. Using the hamster cheek pouch in vivo model, the initial vasoactivity of GNP-Dex at concentrations (1-50 mg/ml) equivalent to the first pass of a bolus injection was a brief concentration-dependent dilation in arcade and terminal arterioles. However, they did not induce a pro-inflammatory endothelial dysfunction effect.