Efficacy and prognostic value of delta radiomics on dual-energy computed tomography for gastric cancer with neoadjuvant chemotherapy: a preliminary study.

BACKGROUND: A non-invasive tool for tumor regression grade (TRG) evaluation is urgently needed for gastric cancer (GC) treated with neoadjuvant chemotherapy (NAC). PURPOSE: To develop and validate a radiomics signature (RS) to evaluate TRG for locally advanced GC after NAC and assess its prognostic value. MATERIAL AND METHODS: A total of 103 patients with GC treated with NAC were retrospectively recruited from April 2018 to December 2019 and were randomly allocated into a training cohort (n = 69) and a validation cohort (n = 34). Delineation was performed on both mixed and iodine-uptake images based on dual-energy computed tomography (DECT). A total of 4094 radiomics features were extracted from the pre-NAC, post-NAC, and delta feature sets. Spearman correlation and the least absolute shrinkage and selection operator were used for dimensionality reduction. Multivariable logistic regression was used for TRG evaluation and generated the optimal RS. Kaplan-Meier survival analysis with the log-rank test was implemented in an independent cohort of 40 patients to validate the prognostic value of the optimal RS. RESULTS: Three, five, and six radiomics features were finally selected for the pre-NAC, post-NAC, and delta feature sets. The delta model demonstrated the best performance in assessing TRG in both the training and the validation cohorts (AUCs=0.91 and 0.76, respectively; P>0.1). The optimal RS from the delta model showed a significant capability to predict survival in the independent cohort (P<0.05). CONCLUSION: Delta radiomics based on DECT images serves as a potential biomarker for TRG evaluation and shows prognostic value for patients with GC treated with NAC.

Development and Validation of a DeepSurv Nomogram to Predict Survival Outcomes and Guide Personalized Adjuvant Chemotherapy in Non-Small Cell Lung Cancer.

Objective: To develop and validate a DeepSurv nomogram based on radiomic features extracted from computed tomography images and clinicopathological factors, to predict the overall survival and guide individualized adjuvant chemotherapy in patients with non-small cell lung cancer (NSCLC). Patients and Methods: This retrospective study involved 976 consecutive patients with NSCLC (training cohort, n=683; validation cohort, n=293). DeepSurv was constructed based on 1,227 radiomic features, and the risk score was calculated for each patient as the output. A clinical multivariate Cox regression model was built with clinicopathological factors to determine the independent risk factors. Finally, a DeepSurv nomogram was constructed by integrating the risk score and independent clinicopathological factors. The discrimination capability, calibration, and clinical usefulness of the nomogram performance were assessed using concordance index evaluation, the Greenwood-Nam-D'Agostino test, and decision curve analysis, respectively. The treatment strategy was analyzed using a Kaplan-Meier curve and log-rank test for the high- and low-risk groups. Results: The DeepSurv nomogram yielded a significantly better concordance index (training cohort, 0.821; validation cohort 0.768) with goodness-of-fit (P<0.05). The risk score, age, thyroid transcription factor-1, Ki-67, and disease stage were the independent risk factors for NSCLC.The Greenwood-Nam-D'Agostino test showed good calibration performance (P=0.39). Both high- and low-risk patients did not benefit from adjuvant chemotherapy, and chemotherapy in low-risk groups may lead to a poorer prognosis. Conclusions: The DeepSurv nomogram, which is based on the risk score and independent risk factors, had good predictive performance for survival outcome. Further, it could be used to guide personalized adjuvant chemotherapy in patients with NSCLC.

Development and Validation of a Machine Learning-Based Radiomics Model on Cardiac Computed Tomography of Epicardial Adipose Tissue in Predicting Characteristics and Recurrence of Atrial Fibrillation.

Purpose: This study aimed to evaluate the feasibility of differentiating the atrial fibrillation (AF) subtype and preliminary explore the prognostic value of AF recurrence after ablation using radiomics models based on epicardial adipose tissue around the left atrium (LA-EAT) of cardiac CT images. Method: The cardiac CT images of 314 patients were collected wherein 251 and 63 cases were randomly enrolled in the training and validation cohorts, respectively. Mutual information and the random forest algorithm were used to screen for the radiomic features and construct the radiomics signature. Radiomics models reflecting the features of LA-EAT were built to differentiate the AF subtype, and the multivariable logistic regression model was adopted to integrate the radiomics signature and volume information. The same methodology and algorithm were applied to the radiomic features to explore the ability for predicting AF recurrence. Results: The predictive model constructed by integrating the radiomic features and volume information using a radiomics nomogram showed the best ability in differentiating AF subtype in the training [AUC, 0.915; 95% confidence interval (CI), 0.880-0.951] and validation (AUC, 0.853; 95% CI, 0.755-0.951) cohorts. The radiomic features have shown convincible predictive ability of AF recurrence in both training (AUC, 0.808; 95% CI, 0.750-0.866) and validation (AUC, 0.793; 95% CI, 0.654-0.931) cohorts. Conclusions: The LA-EAT radiomic signatures are a promising tool in the differentiation of AF subtype and prediction of AF recurrence, which may have clinical implications in the early diagnosis of AF subtype and disease management.

CCTA-derived strain analysis in detection of regional myocardial dysfunction in coronary artery disease patients with preserved left ventricular ejection fraction: A feasibility study.

OBJECTIVES: To evaluate the feasibility of using coronary computed tomography angiography (CCTA)-derived strain to detect regional myocardial dysfunction in coronary artery disease (CAD) patients with normal left ventricular ejection fraction (LVEF). METHODS: A total of 1,580 segments from 101 patients who underwent stressed CT myocardial perfusion imaging (CT-MPI) and CCTA were retrospectively enrolled in this study. The CT-derived global and segmental strain values were evaluated using the feature tracking technique. Segments with myocardial blood flow (MBF) < 125 ml/min/100 ml and 95 ml/min/100 ml were categorized as ischemic and infarcted, respectively. RESULTS: Segmental radial strain (SRS) and segmental circumferential strain (SCS) in the abnormal segments (including all segments with MBF < 125 ml/min/100 ml) were significantly lower than those in the normal segments (14.81±8.65% vs 17.17±9.13%, p < 0.001; -10.21±5.79% vs -11.86±4.52%, p < 0.001, respectively). SRS and SCS values in infarcted segments were significantly impaired compared with the ischemic segments (12.43±8.03% vs. 15.32±8.71%, p = 0.038; -7.72±5.91% vs. -10.67±5.66%, p = 0.010, respectively). The AUCs for SRS and SCS in detecting infarcted segments were 0.622 and 0.698, respectively (p < 0.05). CONCLUSIONS: It is feasible for using CCTA-derived strain parameters to detect regional myocardial dysfunction in CAD patients with preserved LVEF. Segmental radial and circumferential strain have the potential ability to distinguish myocardial ischemia from infarction, and normal from ischemic myocardium.

Dual-Energy CT Iodine Concentration to Evaluate Postoperative Pancreatic Fistula after Pancreatoduodenectomy.

Background Pancreatic fibrosis and fatty infiltration are associated with postoperative pancreatic fistula (POPF), but accurate preoperative assessment remains a challenge. Iodine concentration (IC) and fat fraction derived from dual-energy CT (DECT) may reflect the amount of fibrosis and steatosis, potentially enabling the preoperative prediction of POPF. Purpose To identify multiphasic DECT-derived IC and fat fraction that improve the prediction of POPF risks compared with contrast-enhanced CT attenuation values and to evaluate the underlying histopathologic changes. Materials and Methods This retrospective study included patients who underwent pancreatoduodenectomy and DECT (including pancreatic parenchymal, portal venous, and delayed phase scanning) between January 2020 and December 2020. The relationships of the quantitative DECT-derived IC and fat fraction, along with CT attenuation values from enhanced images with POPF risk, were analyzed with logistic regression analysis. The predictive performance of the IC was compared with that of the CT values. The histopathologic underpinnings of IC were evaluated with multivariable linear regression analysis. Results A total of 107 patients (median age, 65 years; interquartile range, 57-70 years; 56 men) were included. Of these, 23 (21%) had POPF. The pancreatic parenchymal-to-portal venous phase IC ratio (adjusted odds ratio [OR], 13; 95% CI: 2, 162; P < .001) was an independent predictor of POPF occurrence. The accuracy of the pancreatic parenchymal-to-portal venous phase IC ratio in predicting POPF was higher than that of the CT value ratio in the same phases (78% vs 65%, P < .001). The pancreatic parenchymal-to-portal venous phase IC ratio was independently associated with pancreatic fibrosis (ß = -1.04; 95% CI: -0.44, -1.64; P = .001). Conclusion A higher pancreatic parenchymal-to-portal venous phase IC ratio was associated with less histologic fibrosis and greater risk of POPF. © RSNA, 2022 Online supplemental material is available for this article. See also the editorial by Lee and Yoon in this issue.

Radiomics from Primary Tumor on Dual-Energy CT Derived Iodine Maps can Predict Cervical Lymph Node Metastasis in Papillary Thyroid Cancer.

RATIONALE AND OBJECTIVES: To develop and validate 2 iodine maps based radiomics nomograms for preoperatively predicting cervical lymph node metastasis (LNM) and central lymph node metastasis (CLNM) in papillary thyroid cancer (PTC). MATERIALS AND METHODS: A total of 346 patients with PTC were enrolled and allocated to training (242) and validation (104) sets. Radiomics features were extracted from arterial and venous phase iodine maps, respectively. Aggregated machine-learning strategy was applied for features selection and construction of 2 radiomics scores (LN rad-score; CLN rad-score). Logistic regression model was employed to establish two radiomics nomograms (nomogram 1: predicting LNM; nomogram 2: predicting CLNM) after incorporating LN or CLN rad-score with clinical predictors. Nomograms performance was determined by discrimination, calibration and clinical usefulness. RESULTS: Nomogram 1 incorporated LN rad-score, age (categorized by 55) and CT reported LN status; Nomogram 2 incorporated CLN rad-score, capsule contact >25% and CT reported CLN status. 2 nomograms both showed good discrimination and calibration in the training (AUC = 0.847; AUC = 0.837) and validation cohorts (AUC = 0.807; AUC = 0.795). Significant improved AUC, net reclassification index (NRI) and integrated discriminatory improvement (IDI) confirmed additional great predictive value of 2 rad-scores, compared with clinical models without radiomics. Decision curve analysis indicated clinical utility of nomograms. 2 nomograms both demonstrated favorable predictive efficacy in CT reported LN or CLN negative subgroup (AUC = 0.766; AUC = 0.744). CONCLUSION: The presented 2 radiomics nomograms are useful tools for preoperative prediction of LNM and CLNM in PTC.

Feasibility study of using virtual non-contrast images derived from dual-energy CT to replace true non-contrast images in patients diagnosed with papillary thyroid carcinoma.

OBJECTIVE: To assess the feasibility of using virtual non-contrast (VNC) images derived from dual-energy computed tomography (DECT) to replace true non-contrast (TNC) images of papillary thyroid carcinoma (PTC) patients. METHODS: Images of 96 PTC patients were retrospectively analyzed. TNC images were acquired under the single-energy mode of DECT after the plain scanning. The arterial and venous phase VNC (VNC-a and VNC-v) images were generated by the post-processing algorithm from the arterial phase and venous phase of contrast-enhanced CT images, respectively. Mean attenuation values, image noise, number and length of calcification were measured. Radiation dose was also calculated. Last, subjective score of image quality was evaluated by a 5-point scale. RESULTS: Signal-to-noise ratio (SNR) of each tissue in TNC images is significantly higher than that of VNC images (p<0.050). Contrast-to-noise ratio (CNR) of fat, muscle, thyroid nodules and internal carotid artery in TNC images is significantly higher than that of VNC images, while CNR in TNC images is lower for cervical vertebra (p<0.001). Calcification is detected on TNC images of 44 patients, while it is omitted on VNC images of 14 patients (31.8%). The subjective score of TNC images is higher than VNC images (p<0.001). The effective dose reduction is 47.6% by avoiding plain scanning. CONCLUSIONS: Considering the different attenuation value, SNR, CNR and especially reduced detection rate of calcification, we deem that VNC images cannot be directly used to replace TNC images in PTC patients, despite the reduced radiation dose.

A radiomic approach to predict myocardial fibrosis on coronary CT angiography in hypertrophic cardiomyopathy.

BACKGROUND: Late gadolinium enhancement (LGE) derived from cardiac magnetic resonance (CMR) represents myocardial fibrosis (MF) and is associated with prognosis in hypertrophic cardiomyopathy (HCM). However, it cannot be evaluated when CMR is unavailable. Hence, we aimed to investigate the ability of radiomic features derived from coronary computed tomography angiography (CCTA) to detect the presence and extent of MF in HCM, with LGE as references. METHODS: 161 patients with HCM who underwent CCTA and CMR were retrospectively enrolled and randomly divided into training (107 patients, 1712 segments) and testing cohorts (54 patients, 864 segments). Segments were obtained according to AHA 17-segment method. Radiomic features were extracted from per-segment and entire myocardium regions, and multiple machine-learning algorithms were used for radiomic signatures (Rad-sig) generation and model building. Four models were established by multivariable logistic regression using Rad-sig (R-model), clinical characteristic (C-model), echocardiography parameters (E-model), and all features integrated (Integ-model) to identify LGE/left ventricular mass ≥ 15%. RESULTS: The model achieved good diagnostic accuracy in both training (area under the curve [AUC]:0.81, 95% confidence interval [CI]: 0.78-0.83) and testing cohort (AUC: 0.78, 95%CI: 0.75-0.81) on a per-segment basis for the presence of MF. The Integ-model owned the highest discriminative ability for patients with LGE/left ventricular mass ≥ 15% in both training and testing cohorts with AUC of 0.94 (95%CI: 0.89-0.98) and 0.92 (95%CI: 0.85-0.99), respectively. CONCLUSIONS: Our radiomic models were considered as useful and complementary biomarkers for the evaluation of the presence and extent of MF on CCTA, facilitating clinical decision-making and risk stratification in HCM patients.

Whole-tumour evaluation with MRI and radiomics features to predict the efficacy of S-1 for adjuvant chemotherapy in postoperative pancreatic cancer patients: a pilot study.

BACKGROUND: Multiple guidelines for pancreatic ductal adenocarcinoma (PDAC) suggest that all stages of patients need to receive postoperative adjuvant chemotherapy. S-1 is a recently emerged oral antitumour agent recommended by the guidelines. However, which population would benefit from S-1 needs to be determined, and predictors of chemotherapy response are needed for personalized precision medicine. This pilot study aimed to initially identify whether whole-tumour evaluation with MRI and radiomics features could be used for predicting the efficacy of S-1 and to find potential predictors of the efficacy of S-1 as evidence to assist personalized precision treatment. METHODS: Forty-six patients with PDAC (31 in the primary cohort and 15 in the validation cohort) who underwent curative resection and subsequently adjuvant chemotherapy with S-1 were included. Pre-operative abdominal contrast-enhanced MRI was performed, and radiomics features of the whole PDAC were extracted from the primary cohort. After univariable analysis and radiomics features selection, a multivariable Cox regression model for survival analysis was subsequently used to select statistically significant factors associated with postoperative disease-free survival (DFS). Predictive capacities of the factors were tested on the validation cohort by using Kaplan-Meier method. RESULTS: Multivariable Cox regression analysis identified the probability of T1WI_NGTDM_Strength and tumour location as independent predictors of the efficacy of S-1 for adjuvant chemotherapy of PDAC (p = 0.005 and 0.013) in the primary cohort, with hazard ratios (HRs) of 0.289 and 0.293, respectively. Further survival analysis showed that patients in the low-T1WI_NGTDM_Strength group had shorter DFS (median = 5.1 m) than those in the high-T1WI_NGTDM_Strength group (median = 13.0 m) (p = 0.006), and patients with PDAC on the pancreatic head exhibited shorter DFS (median = 7.0 m) than patients with tumours in other locations (median = 20.0 m) (p = 0.016). In the validation cohort, the difference in DFS between patients with low-T1WI_NGTDM_Strength and high-T1WI_NGTDM_Strength and the difference between patients with PDAC on the pancreatic head and that in other locations were approved, with marginally significant (p = 0.073 and 0.050), respectively. CONCLUSIONS: Whole-tumour radiomics feature of T1WI_NGTDM_Strength and tumour location were potential predictors of the efficacy of S-1 and for the precision selection of S-1 as adjuvant chemotherapy regimen for PDAC.

S100A4 overexpression in pancreatic ductal adenocarcinoma: imaging biomarkers from whole-tumor evaluation with MRI and texture analysis.

OBJECTIVE: To investigate the relationship between imaging findings and S100A4 overexpression in pancreatic ductal adenocarcinoma (PDAC) and to determine imaging biomarkers of S100A4 overexpression from whole-tumor evaluation with MRI and texture analysis. METHODS: A total of 60 patients with pathologically confirmed PDAC were included in the study. All patients underwent preoperative abdominal contrast-enhanced MRI examination with Magnetom Aera (Siemens Healthcare, Germany, 1.5 T) at our institute. Whole-tumor evaluation including texture analysis was performed. Sections of specimens were reviewed, and the S100A4 expression status was quantitatively evaluated. Univariate and multivariate logistic regression analyses were conducted to find imaging biomarkers that could predict S100A4 overexpression. RESULTS: Twenty-four tumors (40.0%) had negative results for S100A4 overexpression, and 36 tumors (60.0%) exhibited overexpression. After univariate and multivariate analysis, distal pancreatic duct dilatation, T1WI_10th percentile and the enhancement rate difference between delayed phase (DP) and portal venous phase (PVP) were identified to predict S100A4 overexpression in PDAC independently (p = 0.009, 0.012 and 0.044), with odds ratios (ORs) of 0.102, 0.139 and 4.645, respectively. The area under the ROC curve (AUC) values were 0.715, 0.707 and 0.691. The AUC value of the proposed model was 0.877 with a sensitivity of 80.6% and specificity of 75.0%. CONCLUSION: A model including distal pancreatic duct dilatation, T1WI_10th percentile and the enhancement rate difference between the DP and PVP could predict S100A4 overexpression in PDAC as imaging biomarkers.

Accuracy, agreement, and reliability of DECT-derived vBMD measurements: an initial ex vivo study.

OBJECTIVES: To assess the agreement and reliability of DECT (dual-energy CT)-derived vBMD (volumetric bone mineral density) measurements from excised human femoral heads and to compare DECT-derived BMD with that measured by DXA (dual-energy X-ray absorptiometry) and QCT (quantitative CT) to determine its accuracy. METHODS: Twenty patients that underwent total hip arthroplasty were enrolled to this study. Femoral heads were excised to rectangles without cortical bones for scanning. A dual-source DECT scanner generated images under 80/Sn140 kVp and 100/Sn140 kVp scanning conditions. Specimens were subsequently scanned by QCT and DXA to produce QCT-derived vBMD (mg/cm3) and DXA-derived BMM (bone mineral mass, g). DECT images were loaded to a post-processing workstation to calculate DECT-derived vBMD and BMM. RESULTS: Higher DECT-derived vBMD and BMM were found under 80/Sn140 and 100/Sn140 kVp compared with those for QCT and DXA (p = 0.005). DECT-derived vBMD was highly correlated with QCT-derived vBMD (r = 0.961 ~ 0.993, p < 0.05). Similarly, DECT-derived BMM was strongly correlated with DXA-derived BMM (r = 0.927 ~ 0.943, p < 0.05). Agreement of the inter- and intra-observation of DECT-derived vBMD was excellent. Linear regression was carried out to calibrate DECT-derived vBMD of 80/Sn140 kVp (14 + 0.7 × DECT-derived vBMD) and 100/Sn140 kVp (74 + 0.4 × DECT-derived vBMD) with the reference of QCT-derived vBMD. After calibration, excellent agreement was found for vBMD and BMM within various imaging modalities. CONCLUSIONS: Our study showed that DECT-derived vBMD exhibited high agreement and reliability features, and after calibration, it also displayed a high degree of accuracy. However, in vivo studies are needed to extend its potential utility in clinical settings. KEY POINTS: • Measurements of DECT-derived vBMD had high intra- and inter-observer agreement and reliability. • Measurements of DECT-derived vBMD and BMM had a high correlation with those derived from QCT and DXA. • DECT-derived vBMD and BMM were accurate after calibration compared with QCT and DXA.

Assessment of liver metastases radiomic feature reproducibility with deep-learning-based semi-automatic segmentation software.

BACKGROUND: Good feature reproducibility enhances model reliability. The manual segmentation of gastric cancer with liver metastasis (GCLM) can be time-consuming and unstable. PURPOSE: To assess the value of a semi-automatic segmentation tool in improving the reproducibility of the radiomic features of GCLM. MATERIAL AND METHODS: Patients who underwent dual-source computed tomography were retrospectively reviewed. As an intra-observer analysis, one radiologist segmented metastatic liver lesions manually and semi-automatically twice. Another radiologist re-segmented the lesions once as an inter-observer analysis. A total of 1691 features were extracted. Spearman rank correlation was used for feature reproducibility analysis. The times for manual and semi-automatic segmentation were recorded and analyzed. RESULTS: Seventy-two patients with 168 lesions were included. Most of the GCLM radiomic features became more reliable with the tool than the manual method. For the intra-observer feature reproducibility analysis of manual and semi-automatic segmentation, the rates of features with good reliability were 45.5% and 62.3% (P < 0.02), respectively; for the inter-observer analysis, the rates were 29.3% and 46.0% (P < 0.05), respectively. For feature types, the semi-automatic method increased reliability in 6/7 types in the intra-observer analysis and 5/7 types in the inter-observer analysis. For image types, the reliability of the square and exponential types was significantly increased. The mean time of semi-automatic segmentation was significantly shorter than that of the manual method (P < 0.05). CONCLUSION: The application of semi-automated software increased feature reliability in the intra- and inter-observer analyses. The semi-automatic process took less time than the manual process.

3D radiomics predicts EGFR mutation, exon-19 deletion and exon-21 L858R mutation in lung adenocarcinoma.

BACKGROUND: To establish a radiomic approach to identify epidermal growth factor receptor (EGFR) mutation status in lung adenocarcinoma patients based on CT images, and to distinguish exon-19 deletion and exon-21 L858R mutation. METHODS: Two hundred sixty-three patients who underwent pre-surgical contrast-enhanced CT and molecular testing were included, and randomly divided into the training (80%) and test (20%) cohort. Tumor images were three-dimensionally segmented to extract 1,672 radiomic features. Clinical features (age, gender, and smoking history) were added to build classification models together with radiomic features. Subsequently, the top-10 most relevant features were used to establish classifiers. For the classifying tasks including EGFR mutation, exon-19 deletion, and exon-21 L858R mutation, four logistic regression models were established for each task. RESULTS: The training and test cohort consisted of 210 and 53 patients, respectively. Among the established models, the highest accuracy and sensitivity among the four models were 75.5% (61.7-86.2%) and 92.9% (76.5-99.1%) to classify EGFR mutation, respectively. The highest specificity values were 86.7% (69.3-96.2%) and 70.4% (49.8-86.3%) to classify exon-19 deletion and exon-21 L858R mutation, respectively. CONCLUSIONS: CT radiomics can sensitively identify the presence of EGFR mutation, and increase the certainty of distinguishing exon-19 deletion and exon-21 L858R mutation in lung adenocarcinoma patients. CT radiomics may become a helpful non-invasive biomarker to select EGFR mutation patients for invasive sampling.

18F-fluorodeoxyglucose positron emission tomography/computed tomography-based radiomic features for prediction of epidermal growth factor receptor mutation status and prognosis in patients with lung adenocarcinoma.

BACKGROUND: To investigate whether radiomic features from (18F)-fluorodeoxyglucose positron emission tomography/computed tomography [(18F)-FDG PET/CT] can predict epidermal growth factor receptor (EGFR) mutation status and prognosis in patients with lung adenocarcinoma. METHODS: One hundred and seventy-four consecutive patients with lung adenocarcinoma underwent (18F)-FDG PET/CT and EGFR gene testing were retrospectively analyzed. Radiomic features combined with clinicopathological factors to construct a random forest (RF) model to identify EGFR mutation status. The mutant/wild-type model was trained on a training group (n=139) and validated in an independent validation group (n=35). The second RF classifier predicting the 19/21 mutation site was also built and evaluated in an EGFR mutation subset (training group, n=80; validation group, n=25). Radiomic score and 5 clinicopathological factors were integrated into a multivariate Cox proportional hazard (CPH) model for predicting overall survival (OS). AUC (the area under the receiver characteristic curve) and C-index were calculated to evaluate the model's performance. RESULTS: Of 174 patients, 109 (62.6%) harbored EGFR mutations, 21L858R was the most common mutation type [55.9% (61/109)]. The mutant/wild-type model was identified in the training (AUC, 0.77) and validation (AUC, 0.71) groups. The 19/21 mutation site model had an AUC of 0.82 and 0.73 in the training and validation groups, respectively. The C-index of the CPH model was 0.757. The survival time between targeted therapy and chemotherapy for patients with EGFR mutations was significantly different (P=0.03). CONCLUSIONS: Radiomic features based on (18F)-FDG PET/CT combined with clinicopathological factors could reflect genetic differences and predict EGFR mutation type and prognosis.

Radiomics analysis of dual-energy CT-derived iodine maps for diagnosing metastatic cervical lymph nodes in patients with papillary thyroid cancer.

OBJECTIVE: To investigate the value of radiomics analysis of dual-energy computed tomography (DECT)-derived iodine maps for preoperative diagnosing cervical lymph nodes (LNs) metastasis in patients with papillary thyroid cancer (PTC). METHODS: Two hundred and fifty-five LNs (143 non-metastatic and 112 metastatic) were enrolled and allocated to training and validation sets (7:3 ratio). Radiomics features were extracted from arterial and venous phase iodine maps, respectively. Radiomics signature was constructed based on reproducible features using the least absolute shrinkage and selection operator (LASSO) logistic regression algorithm with 10-fold cross-validation. Logistic regression modeling was employed to build models based on CT image features (model 1), radiomics signature (model 2), and the combined (model 3). A nomogram was plotted for the combined model and decision curve analysis was applied for clinical use. Diagnostic performance was assessed and compared. Internal validation was performed on an independent set containing 78 LNs. RESULTS: Model 3 showed optimal diagnostic performance in both training (AUC = 0.933) and validation set (AUC = 0.895), followed by model 2 (training set, AUC = 0.910; validation set, AUC = 0.847). Both these two models outperformed model 1 in both training (AUC = 0.763) (p < 0.05) and validation set (AUC = 0.728) (p < 0.05). CONCLUSION: Radiomics analysis of DECT-derived iodine maps showed better diagnostic performance than qualitative evaluation of CT image features in preoperative diagnosing cervical LN metastasis in PTC patients. Radiomics signature integrated with CT image features can serve as a promising imaging biomarker for the differentiation. KEY POINTS: • Conventional CT image features have limited value for the diagnosis of metastatic LNs in PTC patients. • Radiomics analysis of dual-energy CT-derived iodine maps significantly outperformed qualitative CT image features in differentiating metastatic from non-metastatic LNs. • Radiomics signature integrated with qualitative CT image features can serve as a useful tool in judging LNs status, thus aiding clinical decision-making.

Stress dynamic myocardial CT perfusion for symptomatic patients with intermediate- or high-risk of coronary artery disease: Optimization and incremental improvement between the absolute and relative myocardial blood flow analysis.

BACKGROUND: The optimization of myocardial CT perfusion (CTP) assessment remains inconsistent and uncertain. Our aim was to explore the superior analysis selection and incremental improvement of myocardial blood flow (MBF) assessment on CTP in diagnosing hemodynamically significant coronary artery disease (CAD). METHODS: Sixty patients (43 men and 17 women; 61.38 ± 8.01 years) were prospectively recruited and underwent stress dynamic myocardial CTP examinations. Absolute and relative MBF was used for ischemia evaluation with the invasive coronary angiography and fractional flow reserve were used as the reference standard. Areas under the receiver operating characteristic curves (AUCs) and cutoff values were calculated and compared. RESULTS: There were 151 vessels in 60 patients finally enrolled for analysis. The sensitivity, specificity, PPV, NPV and diagnostic accuracy for the absolute MBF value and relative MBF ratio were 82.76%, 98.92%, 97.96%, 90.20%, and 92.72% and 74.14%, 93.56%, 87.76%, 85.29%, and 86.09%, respectively. The absolute MBF value was superior than the relative MBF ratio in detecting ischemia (AUC, 0.955 [95%CI: 0.919-0.990] vs.0.906 [95%CI:0.857-0.954])(P = 0.02). For territories with both sensitivity and specificity ≤90%, the diagnostic accuracy increased from 79.1% to 88.4% when the specific data were assessed using the absolute MBF value instead of the relative MBF ratio. CONCLUSIONS: The absolute MBF value from the endocardial myocardium on stress dynamic myocardial CTP showed superior diagnostic performance compared to the relative MBF ratio for the detection of myocardial ischemia in intermediate-to-high risk patients. The absolute MBF value provides an incremental benefit toward diagnostic performance for the relative MBF ratio evaluation.

The Performance of a Dual-Energy CT Derived Radiomics Model in Differentiating Serosal Invasion for Advanced Gastric Cancer Patients After Neoadjuvant Chemotherapy: Iodine Map Combined With 120-kV Equivalent Mixed Images.

OBJECTIVES: The aim was to determine whether the dual-energy CT radiomics model derived from an iodine map (IM) has incremental diagnostic value for the model based on 120-kV equivalent mixed images (120 kVp) in preoperative restaging of serosal invasion with locally advanced gastric cancer (LAGC) after neoadjuvant chemotherapy (NAC). METHODS: A total of 155 patients (110 in the training cohort and 45 in the testing cohort) with LAGC who had standard NAC before surgery were retrospectively enrolled. All CT images were analyzed by two radiologists for manual classification. Volumes of interests (VOIs) were delineated semi-automatically, and 1,226 radiomics features were extracted from every segmented lesion in both IM and 120 kVp images, respectively. Spearman's correlation analysis and the least absolute shrinkage and selection operator (LASSO) penalized logistic regression were implemented for filtering unstable and redundant features and screening out vital features. Two predictive models (120 kVp and IM-120 kVp) based on 120 kVp selected features only and 120 kVp combined with IM selected features were established by multivariate logistic regression analysis. We then build a combination model (ComModel) developed with IM-120 kVp signature and ycT. The performance of these three models and manual classification were evaluated and compared. RESULT: Three radiomics models showed great predictive accuracy and performance in both the training and testing cohorts (ComModel: AUC: training, 0.953, testing, 0.914; IM-120 kVp: AUC: training, 0.953, testing, 0.879; 120 kVp: AUC: training, 0.940, testing, 0.831). All these models showed higher diagnostic accuracy (ComModel: 88.9%, IM-120 kVp: 84.4%, 120 kVp: 80.0%) than manual classification (68.9%) in the testing group. ComModel and IM-120 kVp model had better performances than manual classification both in the training (both p<0.001) and testing cohorts (p<0.001 and p=0.034, respectively). CONCLUSIONS: Dual-energy CT-based radiomics models demonstrated convincible diagnostic performance in differentiating serosal invasion in preoperative restaging for LAGC. The radiomics features derived from IM showed great potential for improving the diagnostic capability.

Machine Learning-Based Computational Models Derived From Large-Scale Radiographic-Radiomic Images Can Help Predict Adverse Histopathological Status of Gastric Cancer.

INTRODUCTION: Adverse histopathological status (AHS) decreases outcomes of gastric cancer (GC). With the lack of a single factor with great reliability to preoperatively predict AHS, we developed a computational approach by integrating large-scale imaging factors, especially radiomic features at contrast-enhanced computed tomography, to predict AHS and clinical outcomes of patients with GC. METHODS: Five hundred fifty-four patients with GC (370 training and 184 test) undergoing gastrectomy were retrospectively included. Six radiomic scores (R-scores) related to pT stage, pN stage, Lauren & Borrmann (L&B) classification, World Health Organization grade, lymphatic vascular infiltration, and an overall histopathologic score (H-score) were, respectively, built from 7,000+ radiomic features. R-scores and radiographic factors were then integrated into prediction models to assess AHS. The developed AHS-based Cox model was compared with the American Joint Committee on Cancer (AJCC) eighth stage model for predicting survival outcomes. RESULTS: Radiomics related to tumor gray-level intensity, size, and inhomogeneity were top-ranked features for AHS. R-scores constructed from those features reflected significant difference between AHS-absent and AHS-present groups (P < 0.001). Regression analysis identified 5 independent predictors for pT and pN stages, 2 predictors for Lauren & Borrmann classification, World Health Organization grade, and lymphatic vascular infiltration, and 3 predictors for H-score, respectively. Area under the curve of models using those predictors was training/test 0.93/0.94, 0.85/0.83, 0.63/0.59, 0.66/0.63, 0.71/0.69, and 0.84/0.77, respectively. The AHS-based Cox model produced higher area under the curve than the eighth AJCC staging model for predicting survival outcomes. Furthermore, adding AHS-based scores to the eighth AJCC staging model enabled better net benefits for disease outcome stratification. DISCUSSION: The developed computational approach demonstrates good performance for successfully decoding AHS of GC and preoperatively predicting disease clinical outcomes.

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.

Correlations Study Between 18F-FDG PET/CT Metabolic Parameters Predicting Epidermal Growth Factor Receptor Mutation Status and Prognosis in Lung Adenocarcinoma.

Purpose: This study assessed the ability of metabolic parameters from 18Fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) and clinicopathological data to predict epidermal growth factor receptor (EGFR) expression/mutation status in patients with lung adenocarcinoma and to develop a prognostic model based on differences in EGFR expression status, to enable individualized targeted molecular therapy. Patients and Methods: Metabolic parameters and clinicopathological data from 200 patients diagnosed with lung adenocarcinoma between July 2009 and November 2016, who underwent 18F-FDG PET/CT and EGFR mutation testing, were retrospectively evaluated. Multivariate logistic regression was applied to significant variables to establish a prediction model for EGFR mutation status. Overall survival for both mutant and wild-type EGFR was analyzed to establish a multifactor Cox regression model. Results: Of the 200 patients, 115 (58%) exhibited EGFR mutations and 85 (42%) were wild-type. Among selected metabolic parameters, metabolic tumor volume (MTV) demonstrated a significant difference between wild-type and mutant EGFR mutation status, with an area under the receiver operating characteristic curve (AUC) of 0.60, which increased to 0.70 after clinical data (smoking status) were combined. Survival analysis of wild-type and mutant EGFR yielded mean survival times of 34.451 (95% CI 28.654-40.249) and 53.714 (95% CI 44.331-63.098) months, respectively. Multivariate Cox regression revealed that mutation type, tumor stage, and thyroid transcription factor-1 (TTF-1) expression status were the main factors influencing patient prognosis. The hazard ratio for mutant EGFR was 0.511 (95% CI 0.303-0.862) times that of wild-type, and the risk of death was lower for mutant EGFR than for wild-type. The risk of death was lower in TTF-1-positive than in TTF-1-negative patients. Conclusion: 18F-FDG PET/CT metabolic parameters combined with clinicopathological data demonstrated moderate diagnostic efficacy in predicting EGFR mutation status and were associated with prognosis in mutant and wild-type EGFR non-small-cell lung cancer (NSCLC), thus providing a reference for individualized targeted molecular therapy.