Computerized texture analysis of persistent part-solid ground-glass nodules: differentiation of preinvasive lesions from invasive pulmonary adenocarcinomas

Radiogenomics in Interventional Oncology

PURPOSE OF REVIEW

Radiogenomics is a growing field that has garnered immense interest over the past decade, owing to its numerous applications in the field of oncology and its potential value in improving patient outcomes. Current applications have only begun to delve into the potential of radiogenomics, and particularly in interventional oncology, there is room for development and increased value of these applications.

RECENT FINDINGS

The field of interventional oncology (IO) has seen valuable radiogenomic applications, from prediction of response to locoregional therapies in hepatocellular carcinoma to identification of genetic mutations in non-small cell lung cancer. Future directions that can increase the value of radiogenomics include applications that address tumor heterogeneity, predict immune responsiveness of tumors, and differentiate between oligoprogression and early widespread progression, among others. Radiogenomics, whether in terms of methodologies or applications, is still in the early stages of development and far from maturation. Future applications, particularly in the field of interventional oncology, will allow realization of its full potential.

Pretreatment MRI-Derived Radiomics May Evaluate the Response of Different Induction Chemotherapy Regimens in Locally advanced Nasopharyngeal Carcinoma

RATIONALE AND OBJECTIVES

To evaluate and compare the performance of radiomics in predicting induction chemotherapy response treated with two different regimens in patients with advanced nasopharyngeal carcinoma.

MATERIALS AND METHODS

A total of 265 patients with pathologically confirmed locally advanced nasopharyngeal carcinoma (stage II-IV), including 115 treated with gemcitabine plus cisplatin (GP group) and 150 treated with docetaxel plus cisplatin (TP group) were retrospectively enrolled. Radiomics features were extracted from the volume of interest delineated in multi-MR sequences on a 3T scanner. After random stratified grouping (training and validation cohorts) and logistic regression based on selected features, the association between the radiomics signature and the early response to induction chemotherapy were established for GP and TP regiments, respectively.

RESULTS

Clinical factors showed no significant difference between the response and non-response groups for the GP and TP regiments (all p > 0.05). The accuracy of the radiomics signature consisting of selected features from the joint T1, T2, and T1C in the GP group (0.852 in the training cohort vs. 0.853 in the validation cohort) was significantly higher than that in the TP group (0.774 vs 0.727). The overall performance of the GP model was steady, with efficiency to distinguish responders from nonresponders with an AUC reaching 0.907 (95% confidence interval [CI] [0.843-0.970]) in the training cohort and 0.886 (95% CI [0.772-0.998]) in the validation cohort, while leveling at 0.800 (95% CI [0.712-0.888]) in the training cohort and 0.863 (95% CI [0.758-0.967]) in the validation cohort in the TP group.

CONCLUSION

Pretreatment MR radiomics signature can better predict the early response to IC in the GP regimen than the TP regimen, which may be helpful to guide IC management.

Radiomics nomogram: A noninvasive tool for preoperative evaluation of the invasiveness of pulmonary adenocarcinomas manifesting as ground-glass nodules

•

It is vital to distinguish indolent pulmonary adenocarcinomas from invasive pulmonary adenocarcinomas before surgery.

•

Radiomics is a cutting-edge technology that mines quantitative features from CT images.

•

We designed a nomogram, which incorporated clinical and CT morphological characteristics with the radiomics signature.

•

We applied the radiomics nomogram to preoperatively predict the invasiveness of GGNs.

In this study, we aimed to establish a radiomics nomogram that noninvasively evaluates the invasiveness of pulmonary adenocarcinomas manifesting as ground-glass nodules (GGNs). Computed tomography (CT) images of 509 patients manifesting as GGNs were collected: 70% of cases were included in the training cohort and 30% in the validation cohort. The Max-Relevance and Min-Redundancy (mRMR) and the least absolute shrinkage and selection operator (LASSO) algorithm were used to select the radiomics features and construct a radiomics signature. Univariate and multivariate logistic regression were used to select the invasiveness-related clinical and CT morphological predictors. Age, smoking history, long diameter, and average CT value were retained as independent predictors of GGN invasiveness. A radiomics nomogram was established by integrating clinical and CT morphological features with the radiomics signature. The radiomics nomogram showed good predictive ability in the training set (area under the curve [AUC], 0.940; 95% confidence interval [CI], 0.916–0.964) and validation set (AUC, 0.946; 95% CI, 0.907–0.986). This radiomics nomogram may serve as a noninvasive and accurate predictive tool to determine the invasiveness of GGNs prior to surgery and assist clinicians in creating personalized treatment strategies.

The Usefulness of Imaging Quantification in Discriminating Non-Calcified Pulmonary Hamartoma From Adenocarcinoma

Background

Patients with non-calcified hamartoma were more susceptible to surgery or needle biopsy for the tough discrimination from lung adenocarcinoma. Radiomics have the ability to quantify the lesion features and potentially improve disease diagnosis. Thus, this study aimed to discriminate non-calcified hamartoma from adenocarcinoma by employing imaging quantification and machine learning.

Methods

Forty-two patients with non-calcified hamartoma and 49 patients with adenocarcinoma were retrospentation; Manual lesion segmentation, feature quantification (e.g., texture features), and artificial neural network were performed consecutively. Independent t-test was used to conduct the inter-group comparisons of those imaging features. Receiver operating characteristic curve was performed to investigate the discriminating efficacy.

Results

Significantly higher contrast, cluster prominence, cluster shade, dissimilarity, energy, and entropy in non-calcified hamartoma were observed compared with lung adenocarcinoma. Texture-grey-level co-occurrence matrix showed a well discrimination between non-calcified hamartoma and adenocarcinoma as the detection sensitivity, specificity, accuracy, and the area under the curve were 87.22% ± 9.07%, 82.64% ± 8.07%, 85.11% ± 5.40%, and 0.942, respectively.

Conclusion

Quantifying imaging features is a potentially useful tool for clinical diagnosis. This study demonstrated that non-calcified hamartoma has a heterogeneous distribution of attenuations probably resulting from its complex organizations. Based on this property, imaging quantification could improve discrimination of non-calcified hamartoma from adenocarcinoma.

The Growth Trend Predictions in Pulmonary Ground Glass Nodules Based on Radiomic CT Features

Background: The management of ground glass nodules (GGNs) remains a distinctive challenge. This study is aimed at comparing the predictive growth trends of radiomic features against current clinical features for the evaluation of GGNs.

Methods: A total of 110 GGNs in 85 patients were included in this retrospective study, in which follow up occurred over a span ≥2 years. A total of 396 radiomic features were manually segmented by radiologists and quantitatively analyzed using an Analysis Kit software. After feature selection, three models were developed to predict the growth of GGNs. The performance of all three models was evaluated by a receiver operating characteristic (ROC) curve. The best performing model was also assessed by calibration and clinical utility.

Results: After using a stepwise multivariate logistic regression analysis and dimensionality reduction, the diameter and five specific radiomic features were included in the clinical model and the radiomic model. The rad-score [odds ratio (OR) = 5.130; P < 0.01] and diameter (OR = 1.087; P < 0.05) were both considered as predictive indicators for the growth of GGNs. Meanwhile, the area under the ROC curve of the combined model reached 0.801. The high degree of fitting and favorable clinical utility was detected using the calibration curve with the Hosmer-Lemeshow test and the decision curve analysis was utilized for the nomogram.

Conclusions: A combined model using the current clinical features alongside the radiomic features can serve as a powerful tool to assist clinicians in guiding the management of GGNs.

Does a Deep Learning-Based Computer-Assisted Diagnosis System Outperform Conventional Double Reading by Radiologists in Distinguishing Benign and Malignant Lung Nodules?

Background

In differentiating indeterminate pulmonary nodules, multiple studies indicated the superiority of deep learning–based computer-assisted diagnosis system (DL-CADx) over conventional double reading by radiologists, which needs external validation. Therefore, our aim was to externally validate the performance of a commercial DL-CADx in differentiating benign and malignant lung nodules.

Methods

In this retrospective study, 233 patients with 261 pathologically confirmed lung nodules were enrolled. Double reading was used to rate each nodule using a four-scale malignancy score system, including unlikely (0–25%), malignancy cannot be completely excluded (25–50%), highly likely (50–75%), and considered as malignant (75–100%), with any disagreement resolved through discussion. DL-CADx automatically rated each nodule with a malignancy likelihood ranging from 0 to 100%, which was then quadrichotomized accordingly. Intraclass correlation coefficient (ICC) was used to evaluate the agreement in malignancy risk rating between DL-CADx and double reading, with ICC value of <0.5, 0.5 to 0.75, 0.75 to 0.9, and >0.9 indicating poor, moderate, good, and perfect agreement, respectively. With malignancy likelihood >50% as cut-off value for malignancy and pathological results as gold standard, sensitivity, specificity, and accuracy were calculated for double reading and DL-CADx, separately.

Results

Among the 261 nodules, 247 nodules were successfully detected by DL-CADx with detection rate of 94.7%. Regarding malignancy rating, DL-CADx was in moderate agreement with double reading (ICC = 0.555, 95% CI 0.424 to 0.655). DL-CADx misdiagnosed 40 true malignant nodules as benign nodules and 30 true benign nodules as malignant nodules with sensitivity, specificity, and accuracy of 79.2, 45.5, and 71.7%, respectively. In contrast, double reading achieved better performance with 16 true malignant nodules misdiagnosed as benign nodules and 26 true benign nodules as malignant nodules with sensitivity, specificity, and accuracy of 91.7, 52.7, and 83.0%, respectively.

Conclusion

Compared with double reading, DL-CADx we used still shows inferior performance in differentiating malignant and benign nodules.

Diagnostic Performance of MR Imaging-based Features and Texture Analysis in the Differential Diagnosis of Ovarian Thecomas/Fibrothecomas and Uterine Fibroids in the Adnexal Area

RATIONALE AND OBJECTIVES

To investigate the value of MRI-based features and texture analysis (TA) in the differential diagnosis between ovarian thecomas/fibrothecomas (OTCA/f-TCAs) and uterine fibroids in the adnexal area (UF-iaas).

MATERIALS AND METHODS

This retrospective study included 16 OTCA/f-TCA and 37 UF-iaa patients who underwent conventional MRI and DWI between August 2014 and September 2018. Three-dimensional TA was performed with T2-weighted MRI. The clinical, MRI-based and texture features were compared between OTCA/f-TCAs and UF-iaas. Multivariate logistic regression analysis was used for filtering the independent discriminative features and constructing the discriminating model. ROCs were generated to analyse MRI-based features, texture features and their combination for discriminating between the two diseases.

RESULTS

Six imaging-based features (ipsilateral ovary detection, arterial period enhancement, lesion components, peripheral cysts, "whorl signs", mean ADCs) and six texture features (Histogram-energy, Histogram-entropy, Histogram-kurtosis, GLCM-energy, GLCM-entropy, and Haralick correlation) were significantly different between OTCA/f-TCAs and UF-iaas (p < 0.05). Multivariate analysis of the MRI-based features revealed that arterial period enhancement (OR = 0.104), peripheral cysts (OR = 16.513), and whorl signs (OR = 0.029) were independent features for discriminating between OTCA/f-TCAs and UF-iaas (p < 0.05). Multivariate analysis of the texture features showed that Histogram-energy and GLCM-energy were independent features for discriminating between OTCA/f-TCAs and UF-iaas (p < 0.05). The area under the curve of imaging-based diagnosis was 0.85, and the combination of imaging-based diagnosis and TA improved the area under the curve to 0.87, with higher accuracy, specificity and sensitivity of 86%, 92%, and 84%, respectively (p < 0.05).

CONCLUSIONS

MRI-based features can be useful in differentiating OTCA/f-TCAs from UF-iaas. Furthermore, combining imaging-based diagnosis and TA can improve diagnostic performance.

Impact of histological components on selecting limited lymphadenectomy for lung adenocarcinoma ≤ 2 cm

OBJECTIVES

There is still some dispute regarding the performance of limited mediastinal lymphadenectomy (LML) even for lung adenocarcinoma ≤ 2 cm. We aimed to recognize the potential candidates who can benefit from LML based on the percentage of histological components (PHC).

METHODS

We analyzed 1160 surgical patients with invasive lung adenocarcinoma ≤ 2 cm from seven institutions between January 2012 and December 2015. All histological subtypes were listed in 5% increments by pathological slices. To test the accuracy of frozen section in judging PHC, frozen section slides from 140 cases were reviewed by three pathologists.

RESULTS

There were 882 patients with systematic mediastinal lymphadenectomy (SML) and 278 with LML. Multivariable analysis indicated that, the total percentage of micropapillary and solid components (PHC_MIP+S) > 5 % was the independent predictor of N2 metastasis (P < 0.001). Overall, recurrence-free survival (RFS) and overall survival (OS) favored SML compared with LML, but the subgroup analysis revealed LML and SML had similar prognosis in the group of PHC_MIP+S ≤ 5 %. Moreover, multivariable Cox analysis showed LML (vs. SML) was independently associated with worse prognosis for patients with PHC_MIP+S > 5 % (RFS, HR = 2.143, P < 0.001; OS, HR=1.963, P < 0.001), but not for those with PHC_MIP+S ≤ 5 % (RFS, P = 0.398; OS, P = 0.298). The sensitivity and specificity of frozen section to intraoperatively identify PHC_MIP+S ≤ 5 % were 97.6 % and 84.2 %, respectively.

CONCLUSIONS

PHC_MIP+S showed the predictive value for N2 metastasis and procedure-specific outcome (LML vs. SML). It may serve as a feasible indicator for identifying proper candidates of LML by using intraoperative frozen section.

Subsolid pulmonary nodules: Controversy and perspective

Ground glass and part-solid nodules, collectively referred to as subsolid nodules, present a challenge in management, with a high risk of malignancy but, when malignant, demonstrating indolent behavior. Emerging data suggest longer follow-up intervals and shorter duration of follow-up is likely appropriate in these nodules. Additionally, definitive therapy is shifting to less aggressive approaches such as sub-lobar resection. Patients may benefit from individualized approaches, incorporating both patient and imaging features to determine whether treatment is necessary.

Pilot Study of CT-Based Radiomics Model for Early Evaluation of Response to Immunotherapy in Patients With Metastatic Melanoma

Objective: Determine the performance of a computed tomography (CT) -based radiomics model in predicting early response to immunotherapy in patients with metastatic melanoma. Methods: This retrospective study examined 50 patients with metastatic melanoma who received immunotherapy treatment in our hospital with an anti-programmed cell death-1 (PD-1) agent or an inhibitor of cytotoxic T lymphocyte antigen-4 (CTLA-4). Thirty-four patients who received an anti-PD-1 agent were in the training sample and 16 patients who received a CTLA-4 inhibitor were in the validation sample. Patients with true progressive disease (PD) were in the poor response group, and those with pseudoprogression, complete response (CR), partial response (PR), or stable disease (SD) were in the good response group. CT images were examined at baseline and after the first and second cycles of treatment, and the imaging data were extracted for radiomics modeling. Results: The radiomics model based on pre-treatment, post-treatment, and delta features provided the best results for predicting response to immunotherapy. Receiver operating characteristic (ROC) analysis for good response indicated an area under the curve (AUC) of 0.882 for the training group and an AUC of 0.857 for the validation group. The sensitivity, specificity, and accuracy of model were 85.70% (6/7), 66.70% (6/9), and 75% (12/16) for predicting a good response. Conclusion: A CT-based radiomics model for metastatic melanoma has the potential to predict early response to immunotherapy and to identify pseudoprogression.

Non-invasive evaluation for benign and malignant subcentimeter pulmonary ground-glass nodules (≤1 cm) based on CT texture analysis

OBJECTIVES

To investigate potential diagnostic model for predicting benign or malignant status of subcentimeter pulmonary ground-glass nodules (SPGGNs) (≤1 cm) based on CT texture analysis.

METHODS

A total of 89 SPGGNs from 89 patients were included; 51 patients were diagnosed with adenocarcinoma, and 38 were diagnosed with inflamed or infected benign SPGGNs. Analysis Kit software was used to manually delineate the volume of interest of lesions and extract a total of 396 quantitative texture parameters. The statistical analysis was performed using R software. The SPGGNs were randomly divided into a training set (n = 59) and a validation set (n = 30). All pre-normalized (Z-score) feature values were subjected to dimension reduction using the LASSO algorithm,and the most useful features in the training set were selected. The selected imaging features were then combined into a Rad-score, which was further assessed by ROC curve analysis in the training and validation sets.

RESULTS

Four characteristic parameters (ClusterShade_AllDirection_offset4_SD, ShortRunEmphasis_angle45_offset1, Maximum3DDiameter, SurfaceVolumeRatio) were further selected by LASSO (p < 0.05). As a cluster of imaging biomarkers, the above four parameters were used to form the Rad-score. The AUC for differentiating between benign and malignant SPGGNs in the training set was 0.792 (95% CI: 0.671, 0.913), and the sensitivity and specificity were 86.10 and 65.20%, respectively. The AUC in the validation set was 72.9% (95% CI: 0.545, 0.913), and the sensitivity and specificity were 86.70 and 60%, respectively.

CONCLUSION

The present diagnostic model based on the cluster of imaging biomarkers can preferably distinguish benign and malignant SPGGNs (≤1 cm).

ADVANCES IN KNOWLEDGE

Texture analysis based on CT images provide a new and credible technique for accurate identification of subcentimeter pulmonary ground-glass nodules.

Value of 18F-FDG PET/CT-based radiomics model to distinguish the growth patterns of early invasive lung adenocarcinoma manifesting as ground-glass opacity nodules

Background

To establish and validate ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT-based radiomics model and use it to predict the intermediate-high risk growth patterns in early invasive adenocarcinoma (IAC).

Methods

Ninety-three ground-glass nodules (GGNs) from 91 patients with stage I who underwent a preoperative ¹⁸F-FDG PET/CT scan and histopathological examination were included in this study. The LIFEx software was used to extract 52 PET and 49 CT radiomic features. The least absolute shrinkage and selection operator (LASSO) algorithm was used to select radiomic features and develop radiomics signatures. We used the receiver operating characteristics curve (ROC) to compare the predictive performance of conventional CT parameters, radiomics signatures, and the combination of these two. Also, a nomogram based on conventional CT indicators and radiomics signature score (rad-score) was developed.

Results

GGNs were divided into lepidic group (n = 18) and acinar-papillary group (n = 75). Four radiomic features (2 for PET and 2 for CT) were selected to calculate the rad-score, and the area under the curve (AUC) of rad-score was 0.790, which was not significantly different as the attenuation value of the ground-glass opacity component on CT (CT_GGO) (0.675). When rad-score was combined with edge (joint model), the AUC increased to 0.804 (95% CI [0.699–0.895]), but which was not significantly higher than CT_GGO (P = 0.109). Furthermore, the decision curve of joint model showed higher clinical value than rad-score and CT_GGO, especially under the purpose of screening for intermediate-high risk growth patterns.

Conclusion

PET/CT-based radiomics model shows good performance in predicting intermediate-high risk growth patterns in early IAC. This model provides a useful method for risk stratification, clinical management, and personalized treatment.

Radiological-pathological correlation of subsolid pulmonary nodules: A single centre retrospective evaluation of the 2011 IASLC adenocarcinoma classification system

INTRODUCTION

The 2011 IASLC classification system proposes guidelines for radiologists and pathologists to classify adenocarcinomas spectrum lesions as preinvasive, minimally invasive adenocarcinoma (MIA), or invasive adenocarcinoma (IA). IA portends the worst clinical prognosis, and the imaging distinction between MIA and IA is controversial.

MATERIALS AND METHODS

Subsolid pulmonary nodules resected by microcoil localization over a three-year period were retrospectively reviewed by three chest radiologists and a pulmonary pathologist. Nodules were classified radiologically based on preoperative computed tomography (CT), with the solid nodule component measured on mediastinal windows applied to high-frequency lung kernel reconstructions, and pathologically according to 2011 IASLC criteria. Radiology interobserver and radiological-pathological variability of nodule classification, and potential reasons for nodule classification discordance were assessed.

RESULTS

Seventy-one subsolid nodules in 67 patients were included. The average size of invasive disease focus at histopathology was 5 mm (standard deviation 5 mm). Radiology interobserver agreement of nodule classification was good (Cohen's Kappa = 0.604, 95 % CI: 0.447 to 0.761). Agreement between consensus radiological interpretation and pathological category was fair (Cohen's Kappa = 0.236, 95 % CI: 0.054-0.421). Radiological and pathological nodule classification were concordant in 52 % (37 of 71) of nodules. The IASLC proposed CT solid component cut-off of 5 mm to distinguish MIA and IA yielded a sensitivity of 59 % and specificity of 80 %. Common reasons for nodule classification discordance included multiple solid components within a nodule on CT, scar and stromal collapse at pathology, and measurement variability.

CONCLUSION

Solid component(s) within persistent part-solid pulmonary nodules raise suspicion for invasive adenocarcinoma. Preoperative imaging classification is frequently discordant from final pathology, reflecting interpretive and technical challenges in radiological and pathological analysis.

Diagnostic value and imaging features of multi-detector CT in lung adenocarcinoma with ground glass nodule patients

This study investigated the application value and imaging features of multi-detector CT (MDCT) in the treatment of lung adenocarcinoma with ground glass nodules (GGN). The medical data of 168 patients with pulmonary GGN in Shengli Oilfield Central Hospital from January 2013 to June 2015 were analyzed. Patients with microinvasive adenocarcinoma and invasive adenocarcinoma were included in group A (invasive lung adenocarcinoma, n=98), while patients with atypical adenomatous hyperplasia and adenocarcinoma in situ were included in group B (pre-invasive lung adenocarcinoma, n=70). The imaging features of MDCT were compared. ROC curves of the size of nidus and the size of solid component were drawn for the diagnosis of invasive lung adenocarcinoma. Logistic multivariate regression analysis was used to analyze the risk factors that affected invasive lung adenocarcinoma. There were significant differences in nidus, burr, and lobes of the patients between groups A and B. The size of nidus and the size of solid component of the patients in group A were significantly higher than those of the patients in group B. The AUCs of the size of the nidus and the size of the solid component of the invasive lung adenocarcinoma were 0.891 and 0.902, respectively. The AUC of the combined diagnosis was 0.984. Size of the nidus, size of the solid component, nature of the lesion, burr, and lobes were all risk factors for invasive lung adenocarcinoma. In patients with GGN, size of the nidus and size of the solid component can be used as excellent diagnostic parameters for invasive lung adenocarcinoma, and nidus size (≥9.8 mm), size of the solid component (≥0.9 mm), the mixed GGN nature of the nidus, burr and lobes can distinguish invasive lung adenocarcinoma and pre-invasive lesions.

Do we need to see to believe?-radiomics for lung nodule classification and lung cancer risk stratification

Despite multiple recent advances, the diagnosis and management of lung cancer remain challenging and it continues to be the deadliest malignancy. In 2011, the National Lung Screening Trial (NLST) reported 20% reduction in lung cancer related mortality using annual low dose chest computed tomography (CT). These results led to the approval and nationwide establishment of lung cancer CT-based lung cancer screening programs. These findings have been further validated by the recently published Nederlands-Leuvens Longkanker Screenings Onderzoek (NELSON) and Multicentric Italian Lung Detection (MILD) trials, the latter showing benefit of screening even beyond the 5 years. However, the implementation of lung cancer screening has been impeded by several challenges, including the differentiation between benign and malignant nodules, the large number of false positive studies and the detection of indolent, potentially clinically insignificant lung cancers (overdiagnosis). Hence, the development of non-invasive strategies to accurately classify and risk stratify screen-detected pulmonary nodules in order to individualize clinical management remains a high priority area of research. Radiomics is a recently coined term which refers to the process of imaging feature extraction and quantitative analysis of clinical diagnostic images to characterize the nodule phenotype beyond what is possible with conventional radiologist assessment. Even though it is still in early phase, several studies have already demonstrated that radiomics approaches are potentially useful for lung nodule classification, risk stratification, individualized management and prediction of overall prognosis. The goal of this review is to summarize the current literature regarding the radiomics of screen-detected lung nodules, highlight potential challenges and discuss its clinical application along with future goals and challenges.

Pathologic categorization of lung nodules: Radiomic descriptors of CT attenuation distribution patterns of solid and subsolid nodules in low-dose CT

PURPOSE

Develop a quantitative image analysis method to characterize the heterogeneous patterns of nodule components for the classification of pathological categories of nodules.

MATERIALS AND METHODS

With IRB approval and permission of the National Lung Screening Trial (NLST) project, 103 subjects with low dose CT (LDCT) were used in this study. We developed a radiomic quantitative CT attenuation distribution descriptor (qADD) to characterize the heterogeneous patterns of nodule components and a hybrid model (qADD+) that combined qADD with subject demographic data and radiologist-provided nodule descriptors to differentiate aggressive tumors from indolent tumors or benign nodules with pathological categorization as reference standard. The classification performances of qADD and qADD + were evaluated and compared to the Brock and the Mayo Clinic models by analysis of the area under the receiver operating characteristic curve (AUC).

RESULTS

The radiomic features were consistently selected into qADDs to differentiate pathological invasive nodules from (1) preinvasive nodules, (2) benign nodules, and (3) the group of preinvasive and benign nodules, achieving test AUCs of 0.847 ± 0.002, 0.842 ± 0.002 and 0.810 ± 0.001, respectively. The qADD + obtained test AUCs of 0.867 ± 0.002, 0.888 ± 0.001 and 0.852 ± 0.001, respectively, which were higher than both the Brock and the Mayo Clinic models.

CONCLUSION

The pathologic invasiveness of lung tumors could be categorized according to the CT attenuation distribution patterns of the nodule components manifested on LDCT images, and the majority of invasive lung cancers could be identified at baseline LDCT scans.

Radiomics technology for identifying early-stage lung adenocarcinomas suitable for sublobar resection

OBJECTIVE

Early-stage lung adenocarcinomas that are suitable for limited resection to preserve lung function are difficult to identify. Using a radiomics approach, we investigated the efficiency of voxel-based histogram analysis of 3-dimensional computed tomography images for detecting less-invasive lesions suitable for sublobar resection.

METHODS

We retrospectively reviewed the medical records of 197 patients with pathological stage 0 or IA adenocarcinomas who underwent lung resection for primary lung cancer at our institution between January 2014 and June 2018. The lesions were categorized as either less invasive or invasive. We evaluated tumor volumes, solid volume percentages, mean computed tomography values, and variance, kurtosis, skewness, and entropy levels. We analyzed the relationships between these variables and pathologically less-invasive lesions and designed an optimal model for detecting less-invasive adenocarcinomas.

RESULTS

Univariate analysis revealed seven variables that differed significantly between less invasive (n = 71) and invasive (n = 141) lesions. A multivariate analysis revealed odds ratios for tumor volumes (0.64; 95% confidence interval (CI), 0.46-0.89; P = .008), solid volume percentages (0.96; 95% CI, 0.93-0.99; P = .024), skewness (3.45; 95% CI, 1.38-8.65; P = .008), and entropy levels (0.21; 95% CI, 0.07-0.58; P = .003). The area under the receiver operating characteristic curve was 0.90 (95% CI, 0.85-0.94) for the optimal model containing these 4 variables, with 85% sensitivity and 79% specificity.

CONCLUSIONS

Voxel-based histogram analysis of 3-dimensional computed tomography images accurately detected early-stage lung adenocarcinomas suitable for sublobar resection.

An effective inflation treatment for frozen section diagnosis of small-sized lesions of the lung

Background

The accuracy of intraoperative pathological diagnosis of small-sized pulmonary nodules including ground-glass opacity (GGO) is important for the surgeon to choose a suitable surgical procedure. Diagnosis of the small-sized lesions of the lung by frozen section (FS) is very difficult for the pathologist because of limited FS technology. Here we tested an effective inflation treatment for FS to improve the diagnostic accuracy of small-sized lung lesions.

Methods

The lung specimens were derived from 113 patients who underwent the surgery at Shanghai Chest Hospital in 2018–2019. The specimens were randomly divided into two groups—uninflated or inflated with diluted embedding medium (Tissue-Tek OCT; Sakura Finetek-USA, CA). The qualities of the FSs were compared with that of corresponding permanent paraffin sections. The FS diagnoses were compared with the final pathologic diagnoses of corresponding permanent sections.

Results

Our results showed that the quality of FS of lung tissue was excellent after inflation with diluted embedding medium (1:1). The total consistency between diagnosis of inflated FS and final pathological diagnosis was 85.7%. In control group, however, the consistency was only 70.2%. When the lesions were less than 1cm, the consistency between diagnosis of inflated FS and final pathological diagnosis was 90.3%, compared to 64.9% consistency in uninflated group (P=0.014, <0.05). When the lesions’ computed tomography (CT) measurement threshold ≤−350 HU, the consistency between diagnosis of inflated FS and final pathological diagnosis was 88% compared to 73.2% consistency in uninflated group (P=0.071, >0.05). Accuracy, sensitivity and specificity were observed about 90% for adenocarcinoma in situ (AIS), whereas it is drop to more than 80% for minimally invasive adenocarcinoma (MIA) in inflated FS.

Conclusions

Inflation with diluted embedding medium (1:1) could make lung tissue expand well during FS. By using this method, small-sized lesions (especially less than 1 cm) could be correctly diagnosed to enable adequate surgical procedure, and evaluation of which can be easily based on the intraoperative pathological diagnosis. The small lesions especially AIS could be readily identified on FS. Therefore, this method improves the diagnostic accuracy of FSs for small-sized lung lesions, and has important practical consequences for further therapy.

Comparison and Fusion of Deep Learning and Radiomics Features of Ground-Glass Nodules to Predict the Invasiveness Risk of Stage-I Lung Adenocarcinomas in CT Scan

For stage-I lung adenocarcinoma, the 5-years disease-free survival (DFS) rates of non-invasive adenocarcinoma (non-IA) is different with invasive adenocarcinoma (IA). This study aims to develop CT image based artificial intelligence (AI) schemes to classify between non-IA and IA nodules, and incorporate deep learning (DL) and radiomics features to improve the classification performance. We collect 373 surgical pathological confirmed ground-glass nodules (GGNs) from 323 patients in two centers. It involves 205 non-IA (including 107 adenocarcinoma in situ and 98 minimally invasive adenocarcinoma), and 168 IA. We first propose a recurrent residual convolutional neural network based on U-Net to segment the GGNs. Then, we build two schemes to classify between non-IA and IA namely, DL scheme and radiomics scheme, respectively. Third, to improve the classification performance, we fuse the prediction scores of two schemes by applying an information fusion method. Finally, we conduct an observer study to compare our scheme performance with two radiologists by testing on an independent dataset. Comparing with DL scheme and radiomics scheme (the area under a receiver operating characteristic curve (AUC): 0.83 ± 0.05, 0.87 ± 0.04), our new fusion scheme (AUC: 0.90 ± 0.03) significant improves the risk classification performance (p < 0.05). In a comparison with two radiologists, our new model yields higher accuracy of 80.3%. The kappa value for inter-radiologist agreement is 0.6. It demonstrates that applying AI method is an effective way to improve the invasiveness risk prediction performance of GGNs. In future, fusion of DL and radiomics features may have a potential to handle the classification task with limited dataset in medical imaging.

Value of TSCT Features for Differentiating Preinvasive and Minimally Invasive Adenocarcinoma From Invasive Adenocarcinoma Presenting as Subsolid Nodules Smaller Than 3 cm

BACKGROUND

To distinguish preinvasive (adenocarcinoma in situ/atypical adenomatous hyperplasia) and minimally invasive adenocarcinoma (MIA) from invasive adenocarcinoma (IA) appearing as solitary subsolid nodules (SSNs) less than 3 cm based on thin-section computed tomography (TSCT) features to guide therapeutic approaches.

METHODS

A total of 154 lesions that were histopathologically confirmed to have pre/minimally invasive adenocarcinoma (hereafter pre/MIA) and IA presenting as part-solid nodules (PSNs) or pure ground-glass nodules (pGGNs) were retrospectively reviewed. The TSCT features, including diameter, area, CT value, shape, air bronchogram, margins, and location, were compared and assessed. Receiver operating characteristic analyses were conducted to determine the cut-off values for the qualitative variables and their diagnostic performances.

RESULTS

Of 154 nodules, 89 IA, 53 MIA, eight adenocarcinoma in situ, and four atypical adenomatous hyperplasia lesions were found. Univariate and multivariate logistic regression of the pre/MIA and IA lesions were compared and analyzed among PSNs and pGGNs. Among pGGNs, a significant difference was found in the area (p = 0.004, odds ratio [OR] = 0.124, 95% confidence interval [CI] = 0.300-0.515) between the pre/MIA and IA groups. In PSNs, significant differences were found in the diameter (p = 0.001, OR = 0.171, 95% CI = 0.063-0.467) and CT value (p = 0.001, OR = 0.996, 95% CI = 0.993-0.998) between the pre/MIA and IA groups. According to the corresponding receiver operating characteristic curves, the optimal cut-off tumor area in pGGNs to differentiate pre/MIA from IA was 0.595 cm². A higher CT value of the lesion (≥ -298.500 HU) and a larger diameter (≥1.450 cm) in PSNs were significantly associated with IA.

CONCLUSION

Imaging features from TSCT contribute to distinguishing pre/MIA from IA in solitary subsolid nodules and may contribute to guide the clinical management of these lesions.

CT-based deep learning model to differentiate invasive pulmonary adenocarcinomas appearing as subsolid nodules among surgical candidates: comparison of the diagnostic performance with a size-based logistic model and radiologists

OBJECTIVES

To evaluate the deep learning models for differentiating invasive pulmonary adenocarcinomas (IACs) among subsolid nodules (SSNs) considered for resection in a retrospective diagnostic cohort in comparison with a size-based logistic model and expert radiologists.

METHODS

This study included 525 patients (309 women; median, 62 years) to develop models, and an independent cohort of 101 patients (57 women; median, 66 years) was used for validation. A size-based logistic model and deep learning models using 2.5-dimension (2.5D) and three-dimension (3D) CT images were developed to discriminate IAC from less invasive pathologies. Overall performance, discrimination, and calibration were assessed. Diagnostic performances of the three thoracic radiologists were compared with those of the deep learning model.

RESULTS

The overall performances of the deep learning models (Brier score, 0.122 for the 2.5D DenseNet and 0.121 for the 3D DenseNet) were superior to those of the size-based logistic model (Brier score, 0.198). The area under the receiver operating characteristic curve (AUC) of the 2.5D DenseNet (0.921) was significantly higher than that of the 3D DenseNet (0.835; p = 0.037) and the size-based logistic model (0.836; p = 0.009). At equally high sensitivities of 90%, the 2.5D DenseNet showed significantly higher specificity (88.2%; all p < 0.05) and positive predictive value (97.4%; all p < 0.05) than other models. Model calibration was poor for all models (all p < 0.05). The 2.5D DenseNet had a comparable performance with the radiologists (AUC, 0.848-0.910).

CONCLUSION

The 2.5D DenseNet model could be used as a highly sensitive and specific diagnostic tool to differentiate IACs among SSNs for surgical candidates.

KEY POINTS

• The deep learning model developed using 2.5D DenseNet showed higher overall performance and discrimination than the size-based logistic model for the differentiation of invasive adenocarcinomas among subsolid nodules for surgical candidates. • The 2.5D DenseNet demonstrated a thoracic radiologist-level diagnostic performance and had higher specificity (88.2%) at equal sensitivities (90%) than the size-based logistic model (specificity, 52.9%). • The 2.5D DenseNet could be used to reduce potential overtreatment for the indolent subsolid nodules or to select candidates for sublobar resection instead of the standard lobectomy.

Pulmonary adenocarcinoma appearing as ground-glass opacity nodules identified using non-enhanced and contrast-enhanced CT texture analysis: A retrospective analysis

The present study aimed to investigate the ability of CT-based texture analysis to differentiate invasive adenocarcinoma (IA) from pre-invasive lesions (PIL) or minimally IA (MIA) appearing as ground-glass opacity (GGO) nodules, and to further compare the performance of non-enhanced CT (NECT) images with that of contrast-enhanced CT (CECT) images. A total of 77 patients with GGO nodules and surgically confirmed pulmonary adenocarcinoma were included in the present retrospective study. Each GGO nodule was manually segmented and its texture features were extracted from NECT and CECT images using in-house developed software coded in MATLAB (MathWorks). The independent-samples t-test was used to select the texture features with statistically significant differences between IA and MIA/PIL. Multivariate logistic regression and receiver operating characteristics (ROC) curve analyses were performed to identify predictive features. Of the 77 GGO nodules, 12 were atypical adenomatous hyperplasia or adenocarcinoma in situ (15.6%), 36 were MIA (46.8%) and 29 were IA (37.7%). IA and MIA/PIL exhibited significant differences in most histogram features and gray-level co-occurrence matrix features (P<0.05). Multivariate logistic regression and ROC curve analyses revealed that smaller energy and higher entropy were significant differentiators of IA from MIA and PIL, irrespective of whether NECT images [area under the curve (AUC): 0.839, 0.859] or CECT images (AUC: 0.818, 0.820) are used. Texture analysis of CT images, regardless of whether NECT or CECT is used, has the potential to distinguish IA from PIL or MIA, particularly the parameters of energy and entropy. Furthermore, NECT images were simpler to obtain and no contrast agent was required; thus, analysis with NECT may be a preferred choice.

Better efficacy in differentiating WHO grade II from III oligodendrogliomas with machine-learning than radiologist's reading from conventional T1 contrast-enhanced and fluid attenuated inversion recovery images

Background

The medical imaging to differentiate World Health Organization (WHO) grade II (ODG2) from III (ODG3) oligodendrogliomas still remains a challenge. We investigated whether combination of machine leaning with radiomics from conventional T1 contrast-enhanced (T1 CE) and fluid attenuated inversion recovery (FLAIR) magnetic resonance imaging (MRI) offered superior efficacy.

Methods

Thirty-six patients with histologically confirmed ODGs underwent T1 CE and 33 of them underwent FLAIR MR examination before any intervention from January 2015 to July 2017 were retrospectively recruited in the current study. The volume of interest (VOI) covering the whole tumor enhancement were manually drawn on the T1 CE and FLAIR slice by slice using ITK-SNAP and a total of 1072 features were extracted from the VOI using 3-D slicer software. Random forest (RF) algorithm was applied to differentiate ODG2 from ODG3 and the efficacy was tested with 5-fold cross validation. The diagnostic efficacy of radiomics-based machine learning and radiologist’s assessment were also compared.

Results

Nineteen ODG2 and 17 ODG3 were included in this study and ODG3 tended to present with prominent necrosis and nodular/ring-like enhancement (P < 0.05). The AUC, ACC, sensitivity, and specificity of radiomics were 0.798, 0.735, 0.672, 0.789 for T1 CE, 0.774, 0.689, 0.700, 0.683 for FLAIR, as well as 0.861, 0.781, 0.778, 0.783 for the combination, respectively. The AUCs of radiologists 1, 2 and 3 were 0.700, 0.687, and 0.714, respectively. The efficacy of machine learning based on radiomics was superior to the radiologists’ assessment.

Conclusions

Machine-learning based on radiomics of T1 CE and FLAIR offered superior efficacy to that of radiologists in differentiating ODG2 from ODG3.

Preoperative CT-based radiomics combined with intraoperative frozen section is predictive of invasive adenocarcinoma in pulmonary nodules: a multicenter study

Objectives

Develop a CT-based radiomics model and combine it with frozen section (FS) and clinical data to distinguish invasive adenocarcinomas (IA) from preinvasive lesions/minimally invasive adenocarcinomas (PM).

Methods

This multicenter study cohort of 623 lung adenocarcinomas was split into training (n = 331), testing (n = 143), and external validation dataset (n = 149). Random forest models were built using selected radiomics features, results from FS, lesion volume, clinical and semantic features, and combinations thereof. The area under the receiver operator characteristic curves (AUC) was used to evaluate model performances. The diagnosis accuracy, calibration, and decision curves of models were tested.

Results

The radiomics-based model shows good predictive performance and diagnostic accuracy for distinguishing IA from PM, with AUCs of 0.89, 0.89, and 0.88, in the training, testing, and validation datasets, respectively, and with corresponding accuracies of 0.82, 0.79, and 0.85. Adding lesion volume and FS significantly increases the performance of the model with AUCs of 0.96, 0.97, and 0.96, and with accuracies of 0.91, 0.94, and 0.93 in the three datasets. There is no significant difference in AUC between the FS model enriched with radiomics and volume against an FS model enriched with volume alone, while the former has higher accuracy. The model combining all available information shows minor non-significant improvements in AUC and accuracy compared with an FS model enriched with radiomics and volume.

Conclusions

Radiomics signatures are potential biomarkers for the risk of IA, especially in combination with FS, and could help guide surgical strategy for pulmonary nodules patients.

Key Points

• A CT-based radiomics model may be a valuable tool for preoperative prediction of invasive adenocarcinoma for patients with pulmonary nodules.

• Radiomics combined with frozen sections could help in guiding surgery strategy for patients with pulmonary nodules.

Electronic supplementary material

The online version of this article (10.1007/s00330-019-06597-8) contains supplementary material, which is available to authorized users.

Application of computerized 3D-CT texture analysis of pancreas for the assessment of patients with diabetes

Objective

To evaluate the role of computerized 3D CT texture analysis of the pancreas as quantitative parameters for assessing diabetes.

Methods

Among 2,493 patients with diabetes, 39 with type 2 diabetes (T2D) and 12 with type 1 diabetes (T1D) who underwent CT using two selected CT scanners, were enrolled. We compared these patients with age-, body mass index- (BMI), and CT scanner-matched normal subjects. Computerized texture analysis for entire pancreas was performed by extracting 17 variable features. A multivariate logistic regression analysis was performed to identify the predictive factors for diabetes. A receiver operator characteristic (ROC) curve was constructed to determine the optimal cut off values for statistically significant variables.

Results

In diabetes, mean attenuation, standard deviation, variance, entropy, homogeneity, surface area, sphericity, discrete compactness, gray-level co-occurrence matrix (GLCM) contrast, and GLCM entropy showed significant differences (P < .05). Multivariate analysis revealed that a higher variance (adjusted OR, 1.002; P = .005), sphericity (adjusted OR, 1.649×10⁴; P = .048), GLCM entropy (adjusted OR, 1.057×10⁵; P = .032), and lower GLCM contrast (adjusted OR, 0.997; P < .001) were significant variables. The mean AUCs for each feature were 0.654, 0.689, 0.620, and 0.613, respectively (P < .05). In subgroup analysis, only larger surface area (adjusted OR, 1.000; P = .025) was a significant predictor for T2D.

Conclusions

Computerized 3D CT texture analysis of the pancreas could be helpful for predicting diabetes. A higher variance, sphericity, GLCM entropy, and a lower GLCM contrast were the significant predictors for diabetes.

Glycomic Signatures of Plasma IgG Improve Preoperative Prediction of the Invasiveness of Small Lung Nodules

Preoperative assessment of tumor invasiveness is essential to avoid overtreatment for patients with small-sized ground-glass nodules (GGNs) of 10 mm or less in diameter. However, it is difficult to determine the pathological state by computed tomography (CT) examination alone. Aberrant glycans has emerged as a tool to identify novel potential disease biomarkers. In this study, we used a lectin microarray-based strategy to investigate whether glycosylation changes in plasma immunoglobulin G (IgG) provide additional information about the invasiveness of small GGNs before surgery. Two independent cohorts (discovery set, n = 92; test set, n = 210) of GGN patients were used. Five of 45 lectins (Sambucus nigra agglutinin, SNA; Datura stramonium agglutinin, DSA; Galanthus nivalis agglutinin, GNA; Euonymus europaeus lectin, EEL; and Vicia villosa agglutinin, VVA) were identified as independent factors associated with pathological invasiveness of small GGNs (p < 0.01). Receiver-operating characteristic (ROC) curve analysis indicated the combination of these five lectins could significantly improve the accuracy of CT in diagnosing invasive GGNs, with an area under the curve (AUC) of 0.792 (p < 0.001), a sensitivity of 74.6%, and specificity of 74.4%, which was superior to current clinical biomarkers. These results suggest that the multilectin assay based on plasma IgG glycosylation may be a useful in vitro complementary test to enhance preoperative determination of the invasiveness of GGNs and guide surgeons to select proper clinical management to avoid overtreatment.

Utility of FDG PET/CT for Preoperative Staging of Non-Small Cell Lung Cancers Manifesting as Subsolid Nodules With a Solid Portion of 3 cm or Smaller

OBJECTIVE. The objective of our study was to investigate the utility of FDG PET/CT for the preoperative staging of subsolid non-small cell lung cancers (NSCLCs) with a solid portion size of 3 cm or smaller. MATERIALS AND METHODS. We retrospectively enrolled 855 patients with pathologically proven NSCLCs manifesting as subsolid nodules with a solid portion of 3 cm or smaller on CT. We then compared the diagnostic performances of FDG PET/CT and chest CT for detecting lymph node (LN), intrathoracic, or distant metastases in patients who underwent preoperative chest CT and FDG PET/CT. After propensity score matching, we compared the diagnostic performance of FDG PET/CT in the group who underwent both chest CT and FDG PET/CT with that of chest CT in patients who did not undergo FDG PET/CT. RESULTS. There were LN metastases in 25 of 765 patients (3.3%) who underwent surgical LN dissection or biopsy and intrathoracic or distant metastasis in two of 855 patients (0.2%). For LN staging, FDG PET/CT showed a sensitivity of 44.0%, specificity of 81.5%, positive predictive value of 9.6%, negative predictive value of 97.0%, and accuracy of 79.9%, which were lower than those of chest CT for accuracy (p < 0.0001). FDG PET/CT could not accurately detect any intrathoracic or distant metastasis. After propensity score matching, the diagnostic accuracy for LN staging of FDG PET/CT in the group who underwent both CT and FDG PET/CT was lower than that of chest CT in the group who did not undergo FDG PET/CT (p = 0.002), and the diagnostic accuracy for intrathoracic and distant metastases was not different (p > 0.999). CONCLUSION. FDG PET/CT has limited utility in preoperatively detecting LN or distant metastasis in patients with subsolid NSCLCs with a solid portion size of 3 cm or smaller.

A deep residual learning network for predicting lung adenocarcinoma manifesting as ground-glass nodule on CT images

OBJECTIVE

To develop a deep learning-based artificial intelligence (AI) scheme for predicting the likelihood of the ground-glass nodule (GGN) detected on CT images being invasive adenocarcinoma (IA) and also compare the accuracy of this AI scheme with that of two radiologists.

METHODS

First, we retrospectively collected 828 histopathologically confirmed GGNs of 644 patients from two centers. Among them, 209 GGNs are confirmed IA and 619 are non-IA, including 409 adenocarcinomas in situ and 210 minimally invasive adenocarcinomas. Second, we applied a series of pre-preprocessing techniques, such as image resampling, rescaling and cropping, and data augmentation, to process original CT images and generate new training and testing images. Third, we built an AI scheme based on a deep convolutional neural network by using a residual learning architecture and batch normalization technique. Finally, we conducted an observer study and compared the prediction performance of the AI scheme with that of two radiologists using an independent dataset with 102 GGNs.

RESULTS

The new AI scheme yielded an area under the receiver operating characteristic curve (AUC) of 0.92 ± 0.03 in classifying between IA and non-IA GGNs, which is equivalent to the senior radiologist's performance (AUC 0.92 ± 0.03) and higher than the score of the junior radiologist (AUC 0.90 ± 0.03). The Kappa value of two sets of subjective prediction scores generated by two radiologists is 0.6.

CONCLUSIONS

The study result demonstrates using an AI scheme to improve the performance in predicting IA, which can help improve the development of a more effective personalized cancer treatment paradigm.

KEY POINTS

• The feasibility of using a deep learning method to predict the likelihood of the ground-glass nodule being invasive adenocarcinoma. • Residual learning-based CNN model improves the performance in classifying between IA and non-IA nodules. • Artificial intelligence (AI) scheme yields higher performance than radiologists in predicting invasive adenocarcinoma.

CT-based radiomics signature for the stratification of N2 disease risk in clinical stage I lung adenocarcinoma

Background

Risk stratification of N2 disease is vital for selecting candidates to receive invasive mediastinal staging modalities. In this study, we aimed to stratify the risk of N2 metastasis in clinical stage I lung adenocarcinoma using radiomics analysis.

Methods

Two datasets of patients with clinical stage I lung adenocarcinoma who underwent lung resection were included (training dataset, 880; validation dataset, 322). Using PyRadiomics, 1,078 computed tomography (CT)-based radiomics features were extracted after semi-automated lung nodule segmentation. In order to predict N2 status, a radiomics signature was constructed after selecting the optimal radiomics feature subset by sequentially applying minimum-redundancy-maximum-relevance and least absolute shrinkage and selection operator (LASSO) techniques. Its performance was validated in the validation dataset.

Results

The incidences of N2 metastasis were 8.4% and 7.1% in the training and validation datasets, respectively. Unsupervised cluster analysis revealed that radiomics features significantly correlated with lymph node status and pathological subtypes. For N2 disease prediction, five radiomics features were selected to establish the radiomics signature, which showed a significantly better predictive performance than clinical factors (P<0.001). The area under the receiver operating characteristic curve was 0.81 (0.77-0.86) and 0.69 (0.63-0.75) for radiomics signature and clinical factors, respectively, in the training dataset, and 0.82 (0.71-0.92) and 0.64 (0.52-0.75), respectively, in the validation dataset.

Conclusions

The established CT-based radiomics signature could stratify the risk of N2 metastasis in clinical stage I lung adenocarcinoma, thus assisting clinicians in making patient-specific mediastinal staging strategy.

Pancreatic Ductal Adenocarcinoma: Machine Learning-Based Quantitative Computed Tomography Texture Analysis For Prediction Of Histopathological Grade

Purpose

To assess the performance of combining computed tomography (CT) texture analysis with machine learning for discriminating different histopathological grades of pancreatic ductal adenocarcinoma (PDAC).

Methods

From July 2012 to August 2017, this retrospective study comprised 56 patients with confirmed histopathological PDAC (32 men, 24 women, mean age 64.04±7.82 years) who had undergone preoperative contrast-enhanced CT imaging within 1 month before surgery. Two radiologists blinded to the histopathological outcome independently segmented lesions for quantitative texture analysis. Histogram features, co-occurrence, and run-length texture were calculated. A support-vector machine was constructed to predict the pathological grade of PDAC based on preoperative texture features.

Results

Pathological analysis confirmed 37 low-grade PDAC (five well-differentiated/grade I and 32 moderately differentiated/grade II) and 19 high-grade PDAC (19 poorly differentiated/grade III) tumors. There were no significant differences in clinical or biological characteristics between patients with high-grade and low-grade tumors (P>0.05). There were significant differences between low-grade PDAC and high-grade PDAC on nine histogram features, seven run-length features, and two co-occurrence features. Cluster shade was the most important predictor (sensitivity 0.315). Using these texture features, the support-vector machine achieved 86% accuracy, 78% sensitivity, 95% and specificity.

Conclusion

Machine learning-based CT texture analysis accurately predicted histopathological differentiation grade of PDAC based on preoperative texture features, leading to maximization patient survival and achievement of personalized precision treatment.

Usefulness of CT texture analysis in differentiating benign and malignant renal tumours

AIM

To elucidate visually imperceptible differences between benign and malignant renal tumours using computed tomography texture analysis (CTTA) using filtration histogram based parameters.

MATERIALS AND METHODS

A retrospective study was performed by texture analysis of pretreatment contrast-enhanced CT examinations in 354 histopathologically confirmed renal cell carcinomas (RCCs) and 147 benign renal tumours. A region-of-interest was drawn encompassing the largest cross-section of the tumour on venous phase axial CT. CTTA features of entropy, kurtosis, mean positive pixel density, and skewness at different spatial filters were calculated and compared in an attempt to differentiate benign lesions from malignancy.

RESULTS

Entropy with fine spatial filter was significantly higher in RCC than benign renal tumours (p=0.022). Entropy with fine and medium filters was higher in RCC than lipid-poor angiomyolipoma (p=0.050 and 0.052, respectively). Entropy >5.62 had high specificity of 85.7%, but low sensitivity of 31.3%, respectively, for predicting RCC.

CONCLUSIONS

Differences in entropy were helpful in differentiating RCC from lipid-poor angiomyolipoma, and chromophobe RCC from oncocytoma. This technique may be useful to differentiate lesions that appear equivocal on visual assessment or alter management in poor surgical candidates.

Whole-Lesion Computed Tomography-Based Entropy Parameters for the Differentiation of Minimally Invasive and Invasive Adenocarcinomas Appearing as Pulmonary Subsolid Nodules

OBJECTIVE

The aim of this study was to investigate the differentiation of computed tomography (CT)-based entropy parameters between minimally invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IAC) lesions appearing as pulmonary subsolid nodules (SSNs).

METHODS

This study was approved by the institutional review board in our hospital. From July 2015 to November 2018, 186 consecutive patients with solitary peripheral pulmonary SSNs that were pathologically confirmed as pulmonary adenocarcinomas (74 MIA and 112 IAC lesions) were included and subdivided into the training data set and the validation data set. Chest CT scans without contrast enhancement were performed in all patients preoperatively. The subjective CT features of the SSNs were reviewed and compared between the MIA and IAC groups. Each SSN was semisegmented with our in-house software, and entropy-related parameters were quantitatively extracted using another in-house software developed in the MATLAB platform. Logistic regression analysis and receiver operating characteristic analysis were performed to evaluate the diagnostic performances. Three diagnostic models including subjective model, entropy model, and combined model were built and analyzed using area under the curve (AUC) analysis.

RESULTS

There were 119 nonsolid nodules and 67 part-solid nodules. Significant differences were found in the subjective CT features among nodule type, lesion size, lobulated shape, and irregular margin between the MIA and IAC groups. Multivariate analysis revealed that part-solid type and lobulated shape were significant independent factors for IAC (P < 0.0001 and P < 0.0001, respectively). Three entropy parameters including Entropy-0.8, Entropy-2.0-32, and Entropy-2.0-64 were identified as independent risk factors for the differentiation of MIA and IAC lesions. The median entropy model value of the MIA group was 0.266 (range, 0.174-0.590), which was significantly lower than the IAC group with value 0.815 (range, 0.623-0.901) (P < 0.0001). Multivariate analysis revealed that the combined model had an excellent diagnostic performance with sensitivity of 88.2%, specificity of 73.0%, and accuracy of 82.1%. The AUC value of the combined model was significantly higher (AUC, 0.869) than that of the subjective model (AUC, 0.809) or the entropy model alone (AUC, 0.836) (P < 0.0001).

CONCLUSIONS

The CT-based entropy parameters could help assess the aggressiveness of pulmonary adenocarcinoma via quantitative analysis of intratumoral heterogeneity. The MIA can be differentiated from IAC accurately by using entropy-related parameters in peripheral pulmonary SSNs.

Correlation Study of 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography in Pathological Subtypes of Invasive Lung Adenocarcinoma and Prognosis

Purpose: To investigate the correlation between ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG PET/CT) metabolic parameters and clinicopathological factors in pathological subtypes of invasive lung adenocarcinoma and prognosis.

Patients and Methods: Metabolic parameters and clinicopathological factors from 176 consecutive patients with invasive lung adenocarcinoma between August 2008 and August 2016 who underwent ¹⁸F-FDG PET/CT examination were retrospectively analyzed. Invasive lung adenocarcinoma was divided into five pathological subtypes:lepidic predominant adenocarcinoma (LPA), acinar predominant adenocarcinoma (APA), papillary predominant adenocarcinoma (PPA), solid predominant adenocarcinoma (SPA), and micropapillary predominant adenocarcinoma (MPA). The differences in metabolic parameters [maximal standard uptake value (SUVmax), mean standard uptake value (SUVmean), total lesion glycolysis (TLG), and metabolic tumor volume (MTV)] and tumor diameter for different pathological subtypes were analyzed. Patients were divided into two groups according to their prognosis: good prognosis group (LPA, APA, PPA) and poor prognosis group (SPA, MPA). Logistic regression was used to filter predictors and construct a predictive model, and areas under the receiver operating curve (AUC) were calculated. Cox regression analysis was performed on prognostic factors.

Results: 82 (46.6%) females and 94 (53.4%) males of patients with invasive lung adenocarcinoma were enrolled in this study. Metabolic parameters and tumor diameter of different pathological subtype had statistically significant (P < 0.05). The predictive model constructed using independent predictors (Distant metastasis, Ki-67, and SUVmax) had good classification performance for both groups. The AUC for SUVmax was 0.694 and combined with clinicopathological factors were 0.745. Cox regression analysis revealed that Stage, TTF-1, MTV, and pathological subtype were independent risk factors for patient prognosis. The hazard ratio (HR) of the poor prognosis group was 1.948 (95% CI 1.042–3.641) times the good prognosis group. The mean survival times of good and poor prognosis group were 50.2621 (95% CI 47.818–52.706) and 35.8214 (95% CI 27.483–44.159) months, respectively, while the median survival time was 47.00 (95% CI 45.000–50.000) and 31.50 (95% CI 23.000–49.000) months, respectively.

Conclusion: PET/CT metabolic parameters combined with clinicopathological factors had good classification performance for the different pathological subtypes, which may provide a reference for treatment strategies and prognosis evaluation of patients.

A radiomics model for determining the invasiveness of solitary pulmonary nodules that manifest as part-solid nodules

AIM

A nomogram model was developed to predict the histological subtypes of lung invasive adenocarcinomas (IAs) and minimally invasive adenocarcinomas (MIAs) that manifest as part-solid ground-glass nodules (GGNs).

MATERIALS AND METHODS

This retrospective study enrolled 119 patients with histopathologically confirmed part-solid GGNs assigned to the training (n=83) or testing cohorts (n=36). Radiomic features were extracted based on the unenhanced computed tomography (CT) images. R software was applied to process the qualitative and quantitative data. The CT features model, radiomic signature model, and combined prediction model were constructed and compared.

RESULTS

A total of 396 radiomic features were extracted from the preoperative CT images, four features including MaxIntensity, RMS, ZonePercentage, and LongRunEmphasis_angle0_offset7 were indicated to be the best discriminators to establish the radiomic signature model. The performance of the model was satisfactory in both the training and testing set with areas under the curve (AUCs) of 0.854 (95% confidence interval [CI]: 0.774 to 0.934) and 0.813 (95% CI: 0.670 to 0.955), respectively. The CT morphology of the lesion shape and diameter of the solid component were confirmed to be a significant feature for building the CT features model, which had an AUC of 0.755 (95% CI: 0.648 to 0.843). A nomogram that integrated lesion shape and radiomic signature was constructed, which contributed an AUC of 0.888 (95% CI: 0.82 to 0.955).

CONCLUSIONS

The radiomic signature could provide an important reference for differentiating IAs from MIAs, and could be significantly enhanced by the addition of CT morphology. The nomogram may be highly informative for making clinical decisions.

Radiomic signature: a non-invasive biomarker for discriminating invasive and non-invasive cases of lung adenocarcinoma

Purpose

We aimed to assess the classification performance of a computed tomography (CT)-based radiomic signature for discriminating invasive and non-invasive lung adenocarcinoma.

Patients and Methods

A total of 192 patients (training cohort, n=116; validation cohort, n=76) with pathologically confirmed lung adenocarcinoma were retrospectively enrolled in the present study. Radiomic features were extracted from preoperative unenhanced chest CT images to build a radiomic signature. Predictive performance of the radiomic signature were evaluated using an intra-cross validation cohort. Diagnostic performance of the radiomic signature was assessed via receiver operating characteristic (ROC) analysis.

Results

The radiomic signature consisted of 14 selected features and demonstrated good discrimination performance between invasive and non-invasive adenocarcinoma. The area under the ROC curve (AUC) for the training cohort was 0.83 (sensitivity, 0.84 ; specificity, 0.78; accuracy, 0.82), while that for the validation cohort was 0.77 (sensitivity, 0.94; specificity, 0.52 ; accuracy, 0.82).

Conclusion

The CT-based radiomic signature exhibited good classification performance for discriminating invasive and non-invasive lung adenocarcinoma, and may represent a valuable biomarker for determining therapeutic strategies in this patient population.

Computer-aided diagnosis of ground-glass opacity pulmonary nodules using radiomic features analysis

This study aims to develop a CT-based radiomic features analysis approach for diagnosis of ground-glass opacity (GGO) pulmonary nodules, and also assess whether computer-aided diagnosis (CADx) performance changes in classifying between benign and malignant nodules associated with histopathological subtypes namely, adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), and invasive adenocarcinoma (IAC), respectively. The study involves 182 histopathology-confirmed GGO nodules collected from two cancer centers. Among them, 59 are benign, 50 are AIS, 32 are MIA, and 41 are IAC nodules. Four training/testing data sets-(1) all nodules, (2) benign and AIS nodules, (3) benign and MIA nodules, (4) benign and IAC nodules-are assembled based on their histopathological subtypes. We first segment pulmonary nodules depicted in CT images by using a 3D region growing and geodesic active contour level set algorithm. Then, we computed and extracted 1117 quantitative imaging features based on the 3D segmented nodules. After conducting radiomic features normalization process, we apply a leave-one-out cross-validation (LOOCV) method to build models by embedding with a Relief feature selection, synthetic minority oversampling technique (SMOTE) and three machine-learning classifiers namely, support vector machine classifier, logistic regression classifier and Gaussian Naïve Bayes classifier. When separately using four data sets to train and test three classifiers, the average areas under receiver operating characteristic curves (AUC) are 0.75, 0.55, 0.77 and 0.93, respectively. When testing on an independent data set, our scheme yields higher accuracy than two radiologists (61.3% versus radiologist 1: 53.1% and radiologist 2: 56.3%). This study demonstrates that: (1) the feasibility of using CT-based radiomic features analysis approach to distinguish between benign and malignant GGO nodules, (2) higher performance of CADx scheme in diagnosing GGO nodules comparing with radiologist, and (3) a consistently positive trend between classification performance and invasive grade of GGO nodules. Thus, to improve the CADx performance in diagnosing of GGO nodules, one should assemble an optimal training data set dominated with more nodules associated with non-invasive lung adenocarcinoma (i.e. AIS and MIA).

Translation of Quantitative Imaging Biomarkers into Clinical Chest CT

Quantitative imaging has been proposed as the next frontier in radiology as part of an effort to improve patient care through precision medicine. In 2007, the Radiological Society of North America launched the Quantitative Imaging Biomarkers Alliance (QIBA), an initiative aimed at improving the value and practicality of quantitative imaging biomarkers by reducing variability across devices, sites, patients, and time. Chest CT occupies a strategic position in this initiative because it is one of the most frequently used imaging modalities, anatomically encompassing the leading causes of mortality worldwide. To date, QIBA has worked on profiles focused on the accurate, reproducible, and meaningful use of volumetric measurements of lung lesions in chest CT. However, other quantitative methods are on the verge of translation from research grounds into clinical practice, including (a) assessment of parenchymal and airway changes in patients with chronic obstructive pulmonary disease, (b) analysis of perfusion with dual-energy CT biomarkers, and (c) opportunistic screening for coronary atherosclerosis and low bone mass by using chest CT examinations performed for other indications. The rationale for and the key facts related to the application of these quantitative imaging biomarkers in cardiothoracic chest CT are presented. ^©RSNA, 2019 See discussion on this article by Buckler (pp 977-980).

Discrimination of Pancreatic Serous Cystadenomas From Mucinous Cystadenomas With CT Textural Features: Based on Machine Learning

Objectives: This study was designed to estimate the performance of textural features derived from contrast-enhanced CT in the differential diagnosis of pancreatic serous cystadenomas and pancreatic mucinous cystadenomas.

Methods: Fifty-three patients with pancreatic serous cystadenoma and 25 patients with pancreatic mucinous cystadenoma were included. Textural parameters of the pancreatic neoplasms were extracted using the LIFEx software, and were analyzed using random forest and Least Absolute Shrinkage and Selection Operator (LASSO) methods. Patients were randomly divided into training and validation sets with a ratio of 4:1; random forest method was adopted to constructed a diagnostic prediction model. Scoring metrics included sensitivity, specificity, accuracy, and AUC.

Results: Radiomics features extracted from contrast-enhanced CT were able to discriminate pancreatic mucinous cystadenomas from serous cystadenomas in both the training group (slice thickness of 2 mm, AUC 0.77, sensitivity 0.95, specificity 0.83, accuracy 0.85; slice thickness of 5 mm, AUC 0.72, sensitivity 0.90, specificity 0.84, accuracy 0.86) and the validation group (slice thickness of 2 mm, AUC 0.66, sensitivity 0.86, specificity 0.71, accuracy 0.74; slice thickness of 5 mm, AUC 0.75, sensitivity 0.85, specificity 0.83, accuracy 0.83).

Conclusions: In conclusion, our study provided preliminary evidence that textural features derived from CT images were useful in differential diagnosis of pancreatic mucinous cystadenomas and serous cystadenomas, which may provide a non-invasive approach to determine whether surgery is needed in clinical practice. However, multicentre studies with larger sample size are needed to confirm these results.

Can texture features improve the differentiation of infiltrative lung adenocarcinoma appearing as ground glass nodules in contrast-enhanced CT?

OBJECTIVES

To investigate the validity and efficacy of comparing texture features from contrast-enhanced images with non-enhanced images in identifying infiltrative lung adenocarcinoma represented as ground glass nodules (GGN).

MATERIALS AND METHODS

A retrospective cohort study was conducted with patients presenting with lung adenocarcinoma and treated at a single centre between January 2015 to December 2017. All patients underwent standard and contrast-enhanced thoracic CT scans with 0.5 mm collimation and 1 mm slice reconstruction thickness before surgery. A total of 34 lung adenocarcinoma patients (representing 34 lesions) were analysed; including 21 instances of invasive adenocarcinoma (IAC) lesions, 4 instances of adenocarcinoma in situ (AIS) lesions, and 9 minimally invasive adenocarcinoma (MIA) lesions. After radiologists manually segmented the lesions, texture features were quantitatively extracted using Artificial Intelligence Kit (AK) software. Then, multivariate logistic regression analysis based on standard and contrast-enhanced CT texture features was employed to analyse the invasiveness of lung adenocarcinoma lesions appearing as GGNs. A receiver operating characteristic (ROC) curve analysis was used to evaluate the performance of those models.

RESULTS

A total of 21 quantitative texture features were extracted using the AK software. After dimensionality reduction, 5 and 3 features extracted from thin-section unenhanced and contrast-enhanced CT, respectively, were used to establish the model. The area under the ROC curve (AUC) values for unenhanced CT and enhanced CT features were 0.890 and 0.868, respectively. There was no significant difference (P = 0.190) in the AUC between models based on non-enhanced and contrast-enhanced CT texture features.

CONCLUSION

Compared with unenhanced CT, texture features extracted from contrast-enhanced CT provided no benefit in improving the differential diagnosis of infiltrative lung adenocarcinoma from non-infiltrative malignancies appearing as GGNs.

Predicting malignant potential of subsolid nodules: can radiomics preempt longitudinal follow up CT?

Background

To assess if radiomics can differentiate benign and malignant subsolid lung nodules (SSNs) on baseline or follow up chest CT examinations. If radiomics can differentiate between benign and malignant subsolid lung nodules, the clinical implications are shorter follow up CT imaging and early recognition of lung adenocarcinoma on imaging.

Materials and methods

The IRB approved retrospective study included 36 patients (mean age 69 ± 8 years; 5 males, 31 females) with 108 SSNs (31benign, 77 malignant) who underwent follow up chest CT for evaluation of indeterminate SSN. All SSNs were identified on both baseline and follow up chest CT. DICOM CT images were deidentified and exported into the open access 3D Slicer software (version 4.7) to obtain radiomic features. Logistic regression analyses and receiver operating characteristic (ROC) curves for various quantitative parameters were generated with SPSS statistical software.

Results

Only 2/92 radiomic features (cluster shade and surface volume ratio) enabled differentiation between malignant and benign SSN on baseline chest CT (P = 0.01 and 0.03) with moderate accuracy [AUC 0.624 (0.505–0.743)]. On follow-up CT, 52/92 radiomic features were significantly different between benign and malignant SSN (P: 0.04 - < 0.0001) with improved accuracy [AUC: 0.708 (0.605–0.811), P = 0.04 - < 0.0001]. Radiomics of benign SSN were stable over time, whereas 63/92 radiomic features of malignant SSNs changed significantly between the baseline and follow up chest CT (P: 0.04 - < 0.0001).

Conclusions

Temporal changes in radiomic features of subsolid lung nodules favor malignant etiology over benign. The change in radiomics features of subsolid lung nodules can allow shorter follow up CT imaging and early recognition of lung adenocarcinoma on imaging. Radiomic features have limited application in differentiating benign and early malignant SSN on baseline chest CT.

Imaging phenotype using radiomics to predict dry pleural dissemination in non-small cell lung cancer

Background

Dry pleural dissemination (DPD) in non-small cell lung cancer (NSCLC) is defined as having solid pleural metastases without malignant pleural effusion. We aim to identify DPD by applying radiomics, a novel approach to decode the tumor phenotype.

Methods

Preoperative chest computed tomographic images and basic clinical feature were retrospectively evaluated in patients with surgically resected NSCLC between January 1, 2015 and December 31, 2016. Propensity score was applied to match the DPD and non-DPD groups. One thousand and eighty radiomics features were quantitatively extracted by the 3D slicer software and "pyradiomics" package. Least absolute shrinkage and selection operator (LASSO) binary model was applied for feature selection and developing the radiomics signature. The discrimination was evaluated using area under the curve (AUC) and Youden index.

Results

Sixty-four DPD patients and paired 192 non-DPD patients were enrolled. Using the LASSO model, this study developed a radiomics signature including 10 radiomic features. The mean ± standard deviation values of the radiomics signature with DPD status (-2.129±1.444) was significantly higher compared to those with non-DPD disease (0.071±0.829, P<0.001). The ten-feature based signature showed good discrimination between DPD and non-DPD, with an AUC of 0.93 (95% confidence-interval, 0.891-0.958) respectively. The sensitivity and specificity of the radiomics signature was 85.94% and 85.94%, with the optimal cut-off value of -0.696 and Youden index of 0.71.

Conclusions

The signature based on radiomics features can provide potential predictive value to identify DPD in patients with NSCLC.

HIV-infected patients with opportunistic pulmonary infections misdiagnosed as lung cancers: the clinicoradiologic features and initial application of CT radiomics

Background

To characterize clinicoradiologic and radiomic features for identifying opportunistic pulmonary infections (OPIs) misdiagnosed as lung cancers in patients with human immunodeficiency virus (HIV).

Methods

Twenty-four HIV-infected patients who were misdiagnosed with lung cancers on CT images and had OPIs confirmed by pathological examination or integration of clinical and laboratory findings and 49 HIV-infected patients with lung cancers confirmed pathologically were included. Semiautomated segmentation of the lesion was implemented with an in-house software. The lesion boundary was adjusted manually by radiologists. A total of 99 nonenhanced-CT-based radiomic features were then extracted with PyRadiomics. The clinicoradiologic and radiomic features were compared between the OPI and cancer groups.

Results

In the OPI group, 19 patients (79.2%) had tuberculosis (TB) infections, 2 (8.3%) had nontuberculosis mycobacterium (NTM) infections, 2 (8.3%) had cryptococcus infections and 1 (4.2%) had a mixed infection of TB and NTM. There were significant differences in age, proportion of smokers, smoking index, highly active antiretroviral therapy (HAART) duration, CD4⁺ counts and CD4⁺/CD8⁺ ratio between the two groups (P=0.000, 0.012, 0.007, 0.002, 0.000, and 0.000, respectively). In peripheral-type lesions, the presence of pleural indentation was less common, and the presence of satellite lesions was more common in the OPI group (P=0.016 and 0.020, respectively). Four radiomic parameters of central-type lesions were significantly different, including large dependence high gray level emphasis (LDHGLE), skewness, inverse difference normalized (IDN) and kurtosis (P=0.008, 0.017, 0.017, and 0.017, respectively). However, neither CT features of central-type lesions nor radiomic parameters of peripheral-type lesions were significantly different between the two groups.

Conclusions

Clinicoradiologic features together with radiomics may help identify OPIs mimicking lung cancers in HIV-infected patients.

Can Quantitative Volumetric Analysis Predict Tumor Recurrence in the Patients with Mucinous Adenocarcinoma of the Lung After Surgical Resection?

RATIONALE AND OBJECTIVES

Mucinous adenocarcinoma (MAC) is a distinct histologic variant subtype of lung adenocarcinomas. However, detailed radiologic findings and prognostic factors are still poorly understood. Thus, this study aimed to investigate the prognostic value of quantitative volumetric analysis of the computed tomography images of patients with MAC after. surgical resection.

MATERIALS AND METHODS

Semiautomatic segmentation from computed tomography images of 60 patients with pathologically confirmed MAC was performed and retrospectively reviewed. The main cutoff value in Hounsfield Units (HU) to predict tumor recurrence was defined by receiver-operating curve analysis. Solid volume of mass (SVM) was defined as the volume of HU greater than this cutoff, and solid ratio (Sratio) was defined as SVM divided by total volume. Each parameter was compared to clinicopathologic characteristics and maximum standardized uptake value. Disease-free survival (DFS) was assessed and was compared among patients. Univariate and multivariate Cox regression was performed to predict DFS of MAC.

RESULTS

The cutoff value of HU as determined by ROC analysis was 20 HU. SVM and Sratio were positively correlated with the maximum standardized uptake and pathologic invasion size, respectively (p < 0.001). SVM and Sratio were significantly higher in the recurrence group than in the no-recurrence group (p < 0.001). Multivariate Cox proportional hazards regression analysis revealed that the SVM (Hazard Ratio 1.016; 95% Confidence Interval 1.000-1.032; p = 0.048) and Sratio (Hazard Ratio 29.136; 95% Confidence Interval 1.419-598.191; p = 0.029) were independent significant predictors of DFS.

CONCLUSION

Quantitative volumetric parameters can predict the prognosis of patients with MAC after surgical resection.

Textural analysis and lung function study: Predicting lung fitness for radiotherapy from a CT scan

OBJECTIVE

This study tested the hypothesis that shows advanced image analysis can differentiate fit and unfit patients for radical radiotherapy from standard radiotherapy planning imaging, when compared to formal lung function tests, FEV1 (forced expiratory volume in 1 s) and TLCO (transfer factor of carbon monoxide).

METHODS

An apical region of interest (ROI) of lung parenchyma was extracted from a standard radiotherapy planning CT scan. Software using a grey level co-occurrence matrix (GLCM) assigned an entropy score to each voxel, based on its similarity to the voxels around it.

RESULTS

Density and entropy scores were compared between a cohort of 29 fit patients (defined as FEV1 and TLCO above 50 % predicted value) and 32 unfit patients (FEV1 or TLCO below 50% predicted). Mean and median density and median entropy were significantly different between fit and unfit patients (p = 0.005, 0.0008 and 0.0418 respectively; two-sided Mann-Whitney test).

CONCLUSION

Density and entropy assessment can differentiate between fit and unfit patients for radical radiotherapy, using standard CT imaging.

ADVANCES IN KNOWLEDGE

This study shows that a novel assessment can generate further data from standard CT imaging. These data could be combined with existing studies to form a multiorgan patient fitness assessment from a single CT scan.

Histogram-based models on non-thin section chest CT predict invasiveness of primary lung adenocarcinoma subsolid nodules

109 pathologically proven subsolid nodules (SSN) were segmented by 2 readers on non-thin section chest CT with a lung nodule analysis software followed by extraction of CT attenuation histogram and geometric features. Functional data analysis of histograms provided data driven features (FPC1,2,3) used in further model building. Nodules were classified as pre-invasive (P1, atypical adenomatous hyperplasia and adenocarcinoma in situ), minimally invasive (P2) and invasive adenocarcinomas (P3). P1 and P2 were grouped together (T1) versus P3 (T2). Various combinations of features were compared in predictive models for binary nodule classification (T1/T2), using multiple logistic regression and non-linear classifiers. Area under ROC curve (AUC) was used as diagnostic performance criteria. Inter-reader variability was assessed using Cohen’s Kappa and intra-class coefficient (ICC). Three models predicting invasiveness of SSN were selected based on AUC. First model included 87.5 percentile of CT lesion attenuation (Q.875), interquartile range (IQR), volume and maximum/minimum diameter ratio (AUC:0.89, 95%CI:[0.75 1]). Second model included FPC1, volume and diameter ratio (AUC:0.91, 95%CI:[0.77 1]). Third model included FPC1, FPC2 and volume (AUC:0.89, 95%CI:[0.73 1]). Inter-reader variability was excellent (Kappa:0.95, ICC:0.98). Parsimonious models using histogram and geometric features differentiated invasive from minimally invasive/pre-invasive SSN with good predictive performance in non-thin section CT.

Evaluation of the Impact of Iterative Reconstruction Algorithms on Computed Tomography Texture Features of the Liver Parenchyma Using the Filtration-Histogram Method

OBJECTIVE

To evaluate whether computed tomography (CT) reconstruction algorithms affect the CT texture features of the liver parenchyma.

MATERIALS AND METHODS

This retrospective study comprised 58 patients (normal liver, n = 34; chronic liver disease [CLD], n = 24) who underwent liver CT scans using a single CT scanner. All CT images were reconstructed using filtered back projection (FBP), hybrid iterative reconstruction (IR) (iDOSE⁴), and model-based IR (IMR). On arterial phase (AP) and portal venous phase (PVP) CT imaging, quantitative texture analysis of the liver parenchyma using a single-slice region of interest was performed at the level of the hepatic hilum using a filtration-histogram statistic-based method with different filter values. Texture features were compared among the three reconstruction methods and between normal livers and those from CLD patients. Additionally, we evaluated the inter- and intra-observer reliability of the CT texture analysis by calculating intraclass correlation coefficients (ICCs).

RESULTS

IR techniques affect various CT texture features of the liver parenchyma. In particular, model-based IR frequently showed significant differences compared to FBP or hybrid IR on both AP and PVP CT imaging. Significant variation in entropy was observed between the three reconstruction algorithms on PVP imaging (p < 0.05). Comparison between normal livers and those from CLD patients revealed that AP images depend more strongly on the reconstruction method used than PVP images. For both inter- and intra-observer reliability, ICCs were acceptable (> 0.75) for CT imaging without filtration.

CONCLUSION

CT texture features of the liver parenchyma evaluated using the filtration-histogram method were significantly affected by the CT reconstruction algorithm used.