Guillaume Chassagnon

Computer-aided diagnosis (CAD) of subsolid nodules: Evaluation of a commercial CAD system

OBJECTIVES To evaluate the performance of a commercially available CAD system for automated detection and measurement of subsolid nodules. MATERIALS AND METHODS The CAD system was tested on 50 pure ground-glass and 50 part-solid nodules (median diameter: 17mm) previously found on standard-dose CT scans in 100 different patients. True nodule detection and the total number of CAD marks were evaluated at different sensitivity settings. The influence of nodule and CT acquisition characteristics was analyzed with logistic regression. Software and manually measured diameters were compared with Spearman and Bland-Altman methods. RESULTS With sensitivity adjusted for 3-mm nodule detection, 50/100 (50%) subsolid nodules were detected, at the average cost of 17 CAD marks per CT. These figures were respectively 26/100 (26%) and 2 at the 5-mm setting. At the highest sensitivity setting (2-mm nodule detection), the average number of CAD marks per CT was 41 but the nodule detection rate only increased to 54%. Part-solid nodules were better detected than pure ground glass nodules: 36/50 (72%) versus 14/50 (28%) at the 3-mm setting (p<0.0001), with no influence of the solid component size. Except for the type (i.e. part solid or pure ground glass), no other nodule characteristic influenced the detection rate. High-quality segmentation was obtained for 79 nodules, which for automated measurements correlated well with manual measurements (rho=0.90[0.84-0.93]). All part-solid nodules had software-measured attenuation values above -671Hounsfield units (HU). CONCLUSION The detection rate of subsolid nodules by this CAD system was insufficient, but high-quality segmentation was obtained in 79% of cases, allowing automated measurement of size and attenuation.

Subsolid Lung Nodule Classification: A CT Criterion for Improving Interobserver Agreement

Purpose To evaluate an objective computed tomographic (CT) criterion for distinguishing between part-solid (PS) and nonsolid (NS) lung nodules. Materials and Methods This study received institutional review board approval, and patients gave informed consent. Preoperative CT studies in all patients who underwent surgery for subsolid nodules between 2008 and 2015 were first reviewed by two senior radiologists, who subjectively classified the nodules as PS or NS. A second reading performed 1 month later used predefined classification criteria and involved a third senior radiologist as well as three junior radiologists. Subsolid nodules were classified as PS if a solid portion was detectable in the mediastinal window setting (nonmeasurable, < 50%, or > 50% of the entire nodule) and were otherwise classified as NS (subclassified as pure or heterogeneous). Interreader agreement was assessed with κ statistics and the intraclass correlation coefficient (ICC). Results A total of 99 nodules measuring a median of 20 mm (range, 5-47 mm) in lung window CT images were analyzed. Senior radiologist agreement on the PS/NS distinction increased from moderate (κ = 0.54; 95% confidence interval [CI]: 0.37, 0.71) to excellent (κ = 0.89; 95% CI: 0.80, 0.98) between the first and second readings. At the second readings, agreement among senior and junior radiologists was excellent for PS/NS distinction (ICC = 0.87; 95% CI: 0.83, 0.90) and for subcategorization (ICC = 0.82; 95% CI: 0.77, 0.87). When a solid portion was measurable in the mediastinal window, the specificity for adenocarcinoma invasiveness ranged from 86% to 96%. Conclusion Detection of a solid portion in the mediastinal window setting allows subsolid nodules to be classified as PS with excellent interreader agreement. If the solid portion is measurable, the specificity for adenocarcinoma invasiveness is high.

Long-term computed tomographic changes in cystic fibrosis patients treated with ivacaftor.

Cystic fibrosis (CF) patients with gating mutations in the gene coding for the CF transmembrane conductance regulator (CFTR) protein have been the first to benefit from CFTR-targeted therapies since the approval of ivacaftor in 2012 by the Food and Drug Administration and the European Medicines Agency. Ivacaftor is a potentiator that increases chloride transport through the defective CFTR protein [1, 2]. In phase III clinical trials, ivacaftor significantly improved lung function and clinical status in patients with a G551D mutation aged ≥6 years [3, 4]. In the two studies conducted in adults and children, the forced expiratory volume in 1 s (FEV1) showed a 10.4% predicted and 12.5% pred point increase from baseline after 24 weeks of treatment, as compared to a 0.1 and 0.2% pred point decrease in the control group. Clinical efficacy of ivacaftor was also shown in patients with other gating mutations of the CFTR gene, with a 7.5% pred point increase in FEV1 after 8 weeks of treatment [5]. In all these phase III studies, the primary endpoint was the change in FEV1 and secondary endpoints were other clinical and functional parameters, such as change in weight, respiratory exacerbation rate, health status, and sweat-chloride concentration. Imaging was never listed as an endpoint. Data on computed tomography (CT) changes upon ivacaftor treatment are scarce [6–8] and long-term changes remain unknown. The aim of this study was to assess short- and long-term CT changes in adult CF patients treated with ivacaftor. In CF patients treated with ivacaftor, improvement of CT abnormalities remains stable on long-term follow-up http://ow.ly/Zxu2B

AtlasNet: Multi-atlas Non-linear Deep Networks for Medical Image Segmentation

Deep learning methods have gained increasing attention in addressing segmentation problems for medical images analysis despite the challenges inherited from the medical domain, such as limited data availability, lack of consistent textural or salient patterns, and high di-mensionality of the data. In this paper, we introduce a novel multi-network architecture that exploits domain knowledge to address those challenges. The proposed architecture consists of multiple deep neural networks that are trained after co-aligning multiple anatomies through multi-metric deformable registration. This multi-network architecture can be trained with fewer examples and leads to better performance, robustness and generalization through consensus. Comparable to human accuracy, highly promising results on the challenging task of interstitial lung disease segmentation demonstrate the potential of our approach.

Nouvelle classification TNM des cancers du poumon non à petites cellules

Initial staging is a key part of the initial evaluation of non-small cell lung cancer. It relies on the 7th edition of the TNM classification. Proposals have been recently developed for the 8th edition of the classification, which is due to be enacted in early 2017. Among these proposals, the weight of tumor size has been increased and new N descriptors have been introduced to further describe N category depending on the number station involved. Regarding M descriptors, oligometastatic disease is distinguished from multiple distant extrathoracic metastases.

Risk factors for hemoptysis complicating 17-18 gauge CT-guided transthoracic needle core biopsy: multivariate analysis of 249 procedures

PURPOSE We aimed to identify modifiable and nonmodifiable risk factors for hemoptysis complicating computed tomography (CT)-guided transthoracic needle biopsy. METHODS All procedures performed in our institution from November 2013 to May 2015 were reviewed. Hemoptysis was classified as mild if limited to hemoptoic sputum and abundant otherwise. Presence of intra-alveolar hemorrhage on postbiopsy CT images was also evaluated. Patient- and lesion-related variables were considered nonmodifiable, while procedure-related variables were considered modifiable. RESULTS A total of 249 procedures were evaluated. Hemoptysis and alveolar hemorrhage occurred in 18% and 58% of procedures, respectively, and were abundant or significant in 8% and 17% of procedures, respectively. Concordance between the occurrence of significant alveolar hemorrhage (grade ≥2) and hemoptysis was poor (κ=0.28; 95% CI [0.16-0.40]). In multivariate analysis, female gender (P = 0.008), a longer transpulmonary needle path (P = 0.014), and smaller lesion size (P = 0.044) were independent risk factors for hemoptysis. Transpulmonary needle-path length was the only risk factor for abundant hemoptysis with borderline statistical significance (P = 0.049). CONCLUSION The transpulmonary needle path should be as short as possible to reduce the risk of abundant hemoptysis during CT-guided transthoracic needle biopsy.