1) Review the approaches for detecting and quantifying lung changes in IPF. 2) Understand the predictive value of disease quantitation with respect to survival and outcome. 3) Become familiar with the role of change in quantitative measures at follow up both in the setting of clinical trials and practice.

In idiopathic inflammatory myopathy (IIM), interstitial lung disease (ILD) is a major cause of morbidity and mortality. ILD in IIM may manifest with a variety of pathological and radiographic abnormalities . Most ILD subtypes have characteristic clinical and radiographic features; hence, diagnosis is usually aided by expert radiologist assessment. Radiography and pulmonary function tests (PFT) may provide a qualitative measurement of severity. However, CT evaluation is subject to inter- and intra-observer variability. PFT results can be influenced by patient effort and do not differentiate specific restrictive pulmonary pathologies. We hypothesize that Computer-Aided Lung Informatics for Pathology Evaluation and Rating (CALIPER) software, which characterizes CT parenchyma, can help predict clinical outcomes, objectively quantify extent of ILD in IIM and help in disease monitoring.

CALIPER was utilized to quantify ILD features on CT in 172 subjects with IIM. We retrospectively collected demographic, PFT and medication data at baseline, years 1, 3 and 5. IIM-related mortality was retrospectively assessed.

CALIPER detected diverse parenchymal involvement, with variable quantities of uninvolved parenchyma, ground glass opacities, reticular densities, honeycombing and low attenuation areas. In 95% of patients, CALIPER detected ≥5% parenchymal abnormalities characteristic of ILD. Compared to treated patients, untreated patients had more baseline parenchymal abnormalities. The treated cohort showed improvement in quantity of reticular densities (year 1, 3) and total interstitial abnormalities (year 1), while our untreated subgroup showed worsening interstitial abnormalities (year 3).

CALIPER analysis, including identification and quantification of baseline ILD and detection of change, in parenchymal involvement may prove to be a useful clinical tool in patients with IIM.

Detection and monitoring of ILD progression in patients with IIM can better inform the use of immunomodulatory treatments, both in the clinic and in future research trials.

1) To review quantitative CT techniques for airway and parenchyma assessment in patients with COPD. 2) To discuss the potential and limitations of these techniques. 3) To review how these techniques can impact on the clinical management of patients with COPD.

To determine in a population of cigarette smokers whether distinct subgroups defined by quantitative CT measures of emphysema and gas trapping differ in symptoms, quality of life, or exacerbation frequency.

We studied 8144 current or former cigarette smokers enrolled in the COPDGene® study. All underwent inspiratory and expiratory volumetric CT with automated quantification of % low attenuation areas (LAA) for estimation of emphysema and gas trapping, using thresholds of -950 on inspiratory CT (LAA-950 insp) and -856 on expiratory CT (LAA-856 exp). Normal cutoff values for these parameters, based on 92 normal subjects, were 5.8% for % LAA-950insp, and 24.3% for % LAA-856exp. Cutoff values were adjusted for current smokers. Dyspnea was evaluated by MMRC questionnaire, respiratory symptoms by St George Respiratory Questionnaire, and quality of life by SF-36 questionnaire. We used binary recursive partitioning (tree function in R) to identify subgroup differences in clinical outcomes.

Of the 8144 subjects, 768 (9%) met criteria for emphysema without gas trapping ("emphysema"), 579 (7%) had gas trapping without emphysema ("gas trapping"), 2413 (30%) had mixed gas trapping and emphysema, and 4384 (54%) did not meet criteria for emphysema or gas trapping. Compared with the emphysema group, the gas trapping group was significantly older, had shorter 6 minute walk distance, higher frequency of exacerbations, and had higher scores for dyspnea, respiratory symptoms, and physical component of SF-36. When binary recursive partitioning was used, a cutoff value of approximately 40% for gas trapping identified dichotomous subgroups of severity, assessed by FEV1% predicted, FEV1/FVC ratio, MMRC score, 6 minute walk distance, exacerbation frequency, and St George Respiratory questionnaire.

Quantitative CT assessment of emphysema and gas trapping identifies subgroups of subjects with clinically significant differences in disease severity.

Quantitative CT may be used to identify important clinically important subtypes of COPD.

1) Discuss the latest approaches to low dose and ultralow dose thoracic CT. 2) Understand the prioritization of X-ray exposure factors for different CT chest protocols. 3) Appreciate the role of Iterative Reconstruction algorithms in low dose and ultraow dose chest CT. 4) Understand the approach to compressive sensing algorithms in low dose and ultralow dose chest CT.

This refresher course wil provide a comprehensive review of the latest approaches to low dose and ultralow dose chest CT

To assess image quality of chest CT reconstructed with image based (SafeCT), adaptive statistical (ASIR), and model based (MBIR) iterative reconstruction techniques (IRT) at less than 1 mGy CTDIvol.

Our IRB approved prospective study included 23 patients (mean age 63±13 years, 80±18 kg, M:F18:5) who underwent routine chest CT on a 64 channel MDCT (GE Discovery CT750 HD) and gave written informed consent for acquisition of ultra low dose (ULD) chest CT series. Standard chest CT (8±3.4 mGy) was followed by 3 ULD chest image series (0.2, 0.4, and 0.8 mGy) (total additional dose <1 mSv). Images were used to reconstruct SafeCT (CH0, CH1) and sinogram data were used to reconstructed with ASIR (SS70, SS90) and MBIR and standard CT with ASIR (SS40) (n=23*3*5+23=368 series). Board-certified thoracic radiologists performed independent and blinded evaluation for lesion detection, lesion conspicuity, and visibility of small structures from lowest to highest dose of ULD series and subsequently for standard dose CT.

Of 182 lesion, 112 non-calcified lung nodules (LN) and 8 ground glass opacities (GGO). There were 34 missed lesions [24 LN, 4GGO, 2 thyroid nodule (TN), 3 pleural effusions (PL)] at 0.2 mGy, 27 [18 LN, 2GGO, 2TN, 2 PL] at 0.4 mGy, and 11 [3LN,2GGO, 2TN, 2PL] at 0.8 mGy. The size of missed LN was less than 4mm. There were 7 and 4 false positive lesions at 0.2 and 0.4 mGy, respectively but none at 0.8 mGy. The conspicuity of LN was sufficient fo diagnostic performance for 3/19 at 0.2 mGy, 6/19 at 0.4 mGy and 10/17 (SafeCT:10,ASIR:10,MBIR:7) at 0.8 mGy. Visibility of sub-segmental bronchi was suboptimal at 0.2 and 0.4 mGy but sufficient for diagnostic performance at 0.8 mGy. Visibility of major fissure was suboptimal at 0.2and0.4 mGy but sufficient for 11/23 with IRT. Visibility of mediastinal and axillary lymph nodes was suboptimal at 0.2and0.4 mGy but sufficient for 9/23 with SafeCT, 8/23 with ASIR, 14/23 with MBIR at 0.8 mGy. Visibility of other mediastinal structures was limited at 0.8 mGy and suboptimal at 0.2and0.4 mGy.

Most clinically significant lung lesions can be detected at CTDIvol of 0.8 mGy with SafeCT, ASIR, and MBIR. However, mediastinal structures could not be assessed with sufficient diagnostic confidence at 0.2-0.8 mGy with any IRT.

Lung nodules >4mm can be assessed with IRT at CTDIvol as low as 0.2 mGy but those < 4mm can be missed at CTDIvol less than 0.8 mGy regardless of the IRT.

1) To make radiologists familiar with a daily use of spectral imaging. 2) To describe the clinical usefulness of this imaging modality. 3) To discuss the possibility of applying dual energy for routine chest CT.

1) Understand reasons for the renewed interest in thoracic MR. 2) Be familiar with current and emerging clinical applications of MR imaging in the chest. 3) Become acquainted with recently developed MR approaches to chest imaging. 4) Be aware of investigational MR methods for imaging lung function.

Thoracic MRI, exclusive of cardiovascular imaging, has evolved through stages of initial enthusiasm followed by limited clinical utilization for specific applications. Examples of the latter have included differentiation of thymic hyperplasia vs lymphoma, characterization of mediastinal duplication cysts, neurogenic/thoracic spinal lesions, cardiac/paracardiac masses, evaluation of superior sulcus tumors and the brachial plexus, staging mesothelioma, and evaluating primary chest wall lesions. Ongoing advances in CT in combination with the relative complexities of MR and its suboptimal visualization of the pulmonary parenchyma have continued to restrict the use of MR in the chest. However, there has been a recent resurgence of interest in thoracic MR based upon the development of practical protocols for improved lung imaging with faster proton MR sequences, parallel imaging, non-gadolinium MRA, etc. coupled with increased concern regarding radiation exposure with CT. This presentation will provide an overview of current and emerging clinical applications of nonvascular thoracic MR (including diffusion and whole body MR tumor imaging and the recent introduction of PET-MR), present an update on investigational techniques for imaging lung function including hyperpolarized gas MR, and serve as an introduction to these topics covered in further detail by the refresher course faculty.

1) Identify the basic MR pulse sequences for clinical evaluation of lung structure. 2) Explain the advantages of using 3D radial ultrashort echo time MRI to image the lung. 3) List the critical scan parameters for robust evaluation of lung perfusion using time-resolved contrast-enhanced MRI.

Although many small studies have suggested a useful role for MRI in imaging lung structure and perfusion, it has yet to see widespread use. Because CT is well-established as the primary cross-sectional imaging modality for the lungs, most thoracic radiologists are much more comfortable with CT than they are with MRI. This has hindered the translation of lung MRI protocols into clinical practice. However, MRI offers the potential of greater soft tissue contrast and the ability to assess both lung structure and function without the need for ionizing radiation. The purpose of this presentation is to familiarize the thoracic radiologist with the existing MRI methods for imaging both lung structure and perfusion, to highlight how emerging methods such as 3D radial ultrashort echo time MRI may improve the performance of lung MRI, and to suggest clinical scenarios in which thoracic MRI may be most useful.

Computed diffusion-weighted imaging (cDWI) is the newly proposed method to generate DWI with arbitrary b-values from acquired DWIs (aDWIs) with different b values. The purpose of this study is to directly and prospectively compare capabilities for pulmonary nodule/mass detection and differentiation of malignant from benign lesions among cDWI and aDWIs.

Ninety-seven patients (64 men and 33 women, mean age 69.1 years) with 121 pulmonary nodules/masses (mean diameter; 28.9mm, median; 24mm) underwent DWI with b values at 0, 500 and 1000 s/mm2 by 1.5 T MR system. According to pathological and/or follow up examinations, these pulmonary lesions were divided into malignancy (n=97) and benign (n=24). Then, cDWI with b value at 1,000 s/mm2 (cDWI₁₀₀₀) were computationally generated from aDWIs with b-values at 0 and 500 s/mm2 by our propriety software. To evaluate detection capability of DWI, aDWIs with b values at 500 s/mm2 (aDWI₅₀₀) and 1,000 s/mm2 (aDWI₁₀₀₀) and cDWI1000 were visually assessed by means of 5-points scoring system. For quantitative diagnosis of pulmonary lesion, lesion to spinal cord ratio (LSR) on each DWI was calculated. To evaluate the detection capability, detection rate was compared among aDWI₅₀₀, aDWI₁₀₀₀ and cDWI₁₀₀₀ by McNemar's test. To determine the feasible threshold value for differentiation, ROC-based positive test was performed, and differentiation capability was compared by sensitivities (SE) and accuracies (AC) among aDWI₅₀₀ with and without cDWI₁₀₀₀, aDWI₁₀₀₀, and cDWI₁₀₀₀ by McNemar's test.

The detection rate of aDWI₅₀₀ (99.2%) was significantly higher than that of aDWI₁₀₀₀ (92.6%, p<0.05), however no significant difference with that of cDWI₁₀₀₀ (96.7%, p>0.05). There was no significant difference among aDWI₅₀₀ without cDWI₁₀₀₀ (SE; 72.6%, and AC; 70.3%), aDWI₁₀₀₀ (SE; 73.2%, and AC; 71.9%) and cDWI₁₀₀₀ (SE; 78.5%, and AC; 75.2%). However, the SE and AC of aDWI₅₀₀ with cDWI₁₀₀₀ (SE; 80.4%, and AC; 76.9%) were significantly higher than those of aDWI₅₀₀ without cDWI₁₀₀₀ and aDWI₁₀₀₀ (p<0.05).

Computed DWI was useful technique, and the combination of aDWI₅₀₀ with cDWI₁₀₀₀ would be better to choose in clinical practice for the evaluation of pulmonary nodules/masses.

Computed DWI with high b value added to really acquired DWI with a relatively low b value improves the diagnostic capabilities for the evaluation of pulmonary nodule/mass.

1) Learn why nonvascular thoracic MRI has been underutilized despite proven advantages in tissue characterization (e.g. mediastinal masses) and its absence of ionizing radiation. 2) Learn what should be done to increase its utilization to an appropriate level for good patient care. 3) Learn the various components required to build and maintain a successful nonvascular thoracic MRI practice.

Nonvascular thoracic MRI has been underutilized despite proven advantages in tissue characterization and its absence of ionizing radiation because of insufficient nonvascular thoracic MR training during residency and fellowship, a resultant lack of recognition of its value to patient care, and a resultant discomfort in recommending, protocoling, and interpreting thoracic MR studies. Improved education of trainees, technologists, and radiologists is needed to increase its utilization to an appropriate level for good patient care. Nonvascular thoracic MRI can be cost-effective when considered in the context of the full care cycle of the patient. The various components required to build a successful nonvascular thoracic MRI practice, include: 1) continuous development and maintenance of updated MR protocols, 2) continuous sharing of these updated protocols with one's radiology group, 3) MRI technologist training with regard to thoracic anatomy, cardiac gating, and successful breath-hold imaging, 4) education of referring physicians about the value of nonvascular thoracic MRI to their practice, 5) facilitation of ordering of these MR examinations via computer order entry, 6) creation of structured reporting voice recognition Macros to facilitate reporting by trainees and staff, 7) sharing of interesting and instructive MRI cases at weekly conferences, 8) a quality assurance initiative.

Inhaled hyperpolarized 129Xe diffuses across the alveolar-capillary membrane and dissolves into two compartments: interstitium (barrier) and red blood cells (RBC). This results in an almost 200 ppm frequency shift in 129Xe resonance. The aim of this study is to quantify global and regional pulmonary gas-transfer using hyperpolarized (HP) 129Xe gas transfer MR spectroscopy and MRI, in healthy volunteers and subjects with IPF.

This IRB-approved and HIPAA compliant study was performed on a 1.5T GE clinical scanner. Gas transfer spectra were acquired in 11 healthy volunteers (HV) and 6 IPF subjects using 200-mL of HP 129Xe. Global gas-transfer was quantified using the ratio of the areas under the curves of the RBC and barrier resonance spectra. This RBC:Barrier ratio was correlated with DLCO. In two IPF subjects, 3D images of gas transfer to RBCs were reconstructed using a 1-point Dixon acquisition. Regional gas-transfer defects on RBC images were visually scored by dividing each lung into 16 regions (32 per subject). Presence or absence of 129Xe RBC signal in each region was correlated with the extent of fibrosis in the same region on CT (scored as no, mild, or severe fibrosis).

The RBC:Barrier ratio in IPF subjects was significantly reduced (0.16±0.03) when compared to healthy volunteers (0.55±0.13, p<0.05). Compared to healthy volunteers, IPF patients had significantly greater 129Xe signal in the barrier and less 129Xe signal in RBCs. The RBC:Barrier ratio correlated significantly with DLCO (r=0.89, p<0.05). RBC gas transfer defects within a total of 64 regions were in 28% (23/64) in regions with no fibrosis, in 39% (25/64) in regions with mild fibrosis and 33% (21/64) in regions with severe fibrosis by CT.

Gas-transfer MR spectroscopy and imaging using HP 129Xe can detect global and regional diffusion impairment in IPF patients and may correlate with extent of pulmonary fibrosis depicted by CT. 129Xe MRI can provide a radiation-free method for sensitive assessment of regional gas transfer and may be a useful biomarker to assess response to therapy.

Inhaled hyperpolarized 129Xe diffuses across the alveolar-capillary membrane and dissolves in the pulmonary red blood cells (RBC). This closely mimics the diffusion of O2 and hence imaging of 129Xe in RBCs can depict pulmonary gas exchange.

1) Understand the limitations of proton lung MRI and the strengths and weaknesses of hyperpolarized gas MRI of the lung. 2) Learn about potential research and clinical applications of hyperpolarized gas lung MRI in lung diseases such as CF, asthma, and COPD.

19Fluorine Gas MRI provides a dynamic assessment of pulmonary ventilatory function. The purpose of this work is the demonstrate extraction and generation of image based biomarkers of pulmonary ventilation for utilization in clinical trial and clinical settings.

Imaging [45 Normals (28 Non smokers, 9 exsmokers, 8 smokers), 7 COPD] was performed on a Siemens TIM Trio 3T MRI scanner and consisted of conventional localizing scout and inspiratory/expiratory breath-held scans (1H) and 3D GRE-VIBE functional scans using Perfluoropropane/Oxygen gas mixtures (19F, TR/TE ,15/1.62 ms, NEX=2, Matrix=64x64, slice=15mm, pixel size=6.25x6.25 mm, flip angle= 70°). All acquisitions were performed at total lung capacity to facilitate anatomical correlation utilizing an in house developed gas delivery and subject monitoring apparatus. A total of at least 7 sequential breath holds were performed, interleaved with 3-4 breaths of the O2/PFP mixture (wash-in), or room air (wash-out). Using an in house developed python based script, all 3D masked [masked using Slicer (www.slicer.org)] lung volumetric image datasets were reduced to a single table representing the x,y and z coordinates and pixel value then concatenated to a 4D x,y,z,t,value table. Data analysis was accomplished using standard features of JMP (SAS Institute).

Image reduction facilitated the use of established statistical algorithms and functions to evaluate biomarkers. Each imaging session provides an array of ventilation assessments throughout the wash-in and wash-out times (seconds) of PFP gas including static and dynamic ventilation distribution, gas trapping, ventilation heterogeneity, ventilation defect persistence and clearance and regional efficiency of ventilation.

Dynamic evaluation of the pulmonary airspaces using PFP enhanced MRI provides a straight-forward and relatively inexpensive means for evaluating ventilatory heterogeneity and providing a spatio-temporal descriptor of 'slow to fast filling compartments' in pulmonary disorders. Simplification of data reduction presents many avenues for generation of pulmonary ventilation based biomarkers to evaluate the integrity and functional status of the pulmonary airspaces.

19F-Enhanced MRI of Pulmonary function using PFP gas facilitates dynamic quantitative and qualitative assessments of pulmonary ventilation and the generation of clinically viable imaging biomarkers.

1) Review practical approach to pediatric patient preparation for thoracic MR imaging. 2) Discuss currently available MRI techniques for evaluating thoracic disorders in children. 3) Learn characteristic MRI findings to narrow the differential diagnoses of various thoracic childhood diseases.