Multiparametric approach with synthetic MR imaging for diagnosing salivary gland lesions.

PURPOSE: To determine whether synthetic MR imaging can distinguish between benign and malignant salivary gland lesions. METHODS: The study population included 44 patients with 33 benign and 11 malignant salivary gland lesions. All MR imaging was obtained using a 3 Tesla system. The QRAPMASTER pulse sequence was used to acquire images with four TI values and two TE values, from which quantitative images of T1 and T2 relaxation times and proton density (PD) were generated. The Mann-Whitney U test was used to compare T1, T2, PD, and ADC values among the subtypes of salivary gland lesions. ROC analysis was used to evaluate diagnostic capability between malignant tumors (MTs) and either pleomorphic adenomas (PAs) or Warthin tumors (WTs). We further calculated diagnostic accuracy for distinguishing malignant from benign lesions when combining these parameters. RESULTS: PAs demonstrated significantly higher T1, T2, PD, and ADC values than WTs (all p < 0.001). Compared to MTs, PAs had significantly higher T1, T2, and ADC values (all p < 0.001), whereas WTs had significantly lower T1, T2, and PD values (p < 0.001, p = 0.008, and p = 0.003, respectively). T2 and ADC were most effective in differentiating between MTs and PAs (AUC = 0.928 and 0.939, respectively), and T1 and PD values for differentiating between MTs and WTs (AUC = 0.915 and 0.833, respectively). Combining T1 with T2 or ADC achieved accuracy of 86.4% in distinguishing between malignant and benign tumors. Similarly, combining PD with T2 or ADC reached accuracy of 86.4% for differentiating between malignant and benign tumors. CONCLUSIONS: Utilizing a combination of synthetic MRI parameters may assist in differentiating malignant from benign salivary gland lesions.

Machine learning approach using 18F-FDG-PET-radiomic features and the visibility of right ventricle 18F-FDG uptake for predicting clinical events in patients with cardiac sarcoidosis.

OBJECTIVES: To investigate the usefulness of machine learning (ML) models using pretreatment 18F-FDG-PET-based radiomic features for predicting adverse clinical events (ACEs) in patients with cardiac sarcoidosis (CS). MATERIALS AND METHODS: This retrospective study included 47 patients with CS who underwent 18F-FDG-PET/CT scan before treatment. The lesions were assigned to the training (n = 38) and testing (n = 9) cohorts. In total, 49 18F-FDG-PET-based radiomic features and the visibility of right ventricle 18F-FDG uptake were used to predict ACEs using seven different ML algorithms (namely, decision tree, random forest [RF], neural network, k-nearest neighbors, Naïve Bayes, logistic regression, and support vector machine [SVM]) with tenfold cross-validation and the synthetic minority over-sampling technique. The ML models were constructed using the top four features ranked by the decrease in Gini impurity. The AUCs and accuracies were used to compare predictive performances. RESULTS: Patients who developed ACEs presented with a significantly higher surface area and gray level run length matrix run length non-uniformity (GLRLM_RLNU), and lower neighborhood gray-tone difference matrix_coarseness and sphericity than those without ACEs (each, p < 0.05). In the training cohort, all seven ML algorithms had a good classification performance with AUC values of > 0.80 (range: 0.841-0.944). In the testing cohort, the RF algorithm had the highest AUC and accuracy (88.9% [8/9]) with a similar classification performance between training and testing cohorts (AUC: 0.945 vs 0.889). GLRLM_RLNU was the most important feature of the modeling process of this RF algorithm. CONCLUSION: ML analyses using 18F-FDG-PET-based radiomic features may be useful for predicting ACEs in patients with CS.

Adult T-cell Leukemia/Lymphoma (ATL) in the Nasal and Paranasal Cavity: Four Cases Report.

Adult T-cell leukemia/lymphoma (ATL) is a form of leukemia caused by the human T-cell leukemia virus type I (HTLV-1). Otolaryngologists often diagnose ATL based on cervical lymphadenopathy or Waldeyer ring lesions. However, there are few reports of ATL occurring in the nasal and paranasal cavity. Here, we report four such cases of ATL. Case 1: An 82-year-old man diagnosed with acute-type ATL with a tumor in the nasal cavity underwent 5 courses of THP-COP, but died after 36 months due to ATL. Case 2: A 62-year-old woman diagnosed with lymphoma-type ATL with a tumor in the frontal sinus was treated with 5 courses of VCAP-AMP-VECP, and has survived for more than 10 years. Case 3: A 64-year-old man diagnosed with lymphoma-type ATL with a tumor in the maxillary sinus underwent 8 courses of VCAP-AMP-VECP and 2 courses of mogamulizumab, but died after 34 months due to ATL. Case 4: A 52-year-old woman diagnosed with lymphoma-type ATL with tumors in both ethmoid sinuses received 2 courses of CHOP, 2 courses of DeVIC, radiotherapy (32 Gy) and 2 courses of mogamulizumab, but died after 9 months due to ATL.

Using Texture Analysis of Neck Computed Tomography Images to Differentiate Primary Hyperparathyroidism From Normal Controls.

OBJECTIVE: To investigate the utility of texture analysis in detecting osseous changes associated with hyperparathyroidism on neck CT examinations compared with control patients and to explore the best regions in the head and neck to evaluate changes in the trabecular architecture secondary to hyperparathyroidism. METHODS: Patients with hyperparathyroidism who underwent a 4D CT of the neck with contrast were included in this study. Age-matched control patients with no history of hyperparathyroidism who underwent a contrast-enhanced neck CT were also included. Mandibular condyles, bilateral mandibular bodies, the body of the C4 vertebra, the manubrium of the sternum, and bilateral clavicular heads were selected for analysis, and oval-shaped regions of interest were manually placed. These segmented areas were imported into an in-house developed texture analysis program, and 41 texture analysis features were extracted. A mixed linear regression model was used to compare differences in the texture analysis features contoured at each of the osseous structures between patients with hyperparathyroidism and age-matched control patients. RESULTS: A total of 30 patients with hyperparathyroidism and 30 age-matched control patients were included in this study. Statistically significant differences in texture features between patients with hyperparathyroidism and control patients in all 8 investigated osseous regions. The sternum showed the greatest number of texture features with statistically significant differences between these groups. CONCLUSIONS: Some CT texture features demonstrated statistically significant differences between patients with hyperparathyroidism and control patients. The results suggest that texture features may discriminate changes in the osseous architecture of the head and neck in patients with hyperparathyroidism.

Differentiating primary central nervous system lymphoma from glioblastoma by time-dependent diffusion using oscillating gradient.

BACKGROUND: This study aimed to elucidate the impact of effective diffusion time setting on apparent diffusion coefficient (ADC)-based differentiation between primary central nervous system lymphomas (PCNSLs) and glioblastomas (GBMs) and to investigate the usage of time-dependent diffusion magnetic resonance imaging (MRI) parameters. METHODS: A retrospective study was conducted involving 21 patients with PCNSLs and 66 patients with GBMs using diffusion weighted imaging (DWI) sequences with oscillating gradient spin-echo (Δeff = 7.1 ms) and conventional pulsed gradient (Δeff = 44.5 ms). In addition to ADC maps at the two diffusion times (ADC7.1 ms and ADC44.5 ms), we generated maps of the ADC changes (cADC) and the relative ADC changes (rcADC) between the two diffusion times. Regions of interest were placed on enhancing regions and non-enhancing peritumoral regions. The mean and the fifth and 95th percentile values of each parameter were compared between PCNSLs and GBMs. The area under the receiver operating characteristic curve (AUC) values were used to compare the discriminating performances among the indices. RESULTS: In enhancing regions, the mean and fifth and 95th percentile values of ADC44.5 ms and ADC7.1 ms in PCNSLs were significantly lower than those in GBMs (p = 0.02 for 95th percentile of ADC44.5 ms, p = 0.04 for ADC7.1 ms, and p < 0.01 for others). Furthermore, the mean and fifth and 95th percentile values of cADC and rcADC were significantly higher in PCNSLs than in GBMs (each p < 0.01). The AUC of the best-performing index for ADC7.1 ms was significantly lower than that for ADC44.5 ms (p < 0.001). The mean rcADC showed the highest discriminating performance (AUC = 0.920) among all indices. In peritumoral regions, no significant difference in any of the three indices of ADC44.5 ms, ADC7.1 ms, cADC, and rcADC was observed between PCNSLs and GBMs. CONCLUSIONS: Effective diffusion time setting can have a crucial impact on the performance of ADC in differentiating between PCNSLs and GBMs. The time-dependent diffusion MRI parameters may be useful in the differentiation of these lesions.

Differentiating brain metastasis from glioblastoma by time-dependent diffusion MRI.

BACKGROUND: This study was designed to investigate the use of time-dependent diffusion magnetic resonance imaging (MRI) parameters in distinguishing between glioblastomas and brain metastases. METHODS: A retrospective study was conducted involving 65 patients with glioblastomas and 27 patients with metastases using a diffusion-weighted imaging sequence with oscillating gradient spin-echo (OGSE, 50 Hz) and a conventional pulsed gradient spin-echo (PGSE, 0 Hz) sequence. In addition to apparent diffusion coefficient (ADC) maps from two sequences (ADC50Hz and ADC0Hz), we generated maps of the ADC change (cADC): ADC50Hz - ADC0Hz and the relative ADC change (rcADC): (ADC50Hz - ADC0Hz)/ ADC0Hz × 100 (%). RESULTS: The mean and the fifth and 95th percentile values of each parameter in enhancing and peritumoral regions were compared between glioblastomas and metastases. The area under the receiver operating characteristic curve (AUC) values of the best discriminating indices were compared. In enhancing regions, none of the indices of ADC0Hz and ADC50Hz showed significant differences between metastases and glioblastomas. The mean cADC and rcADC values of metastases were significantly higher than those of glioblastomas (0.24 ± 0.12 × 10-3mm2/s vs. 0.14 ± 0.03 × 10-3mm2/s and 23.3 ± 9.4% vs. 14.0 ± 4.7%; all p < 0.01). In peritumoral regions, no significant difference in all ADC indices was observed between metastases and glioblastomas. The AUC values for the mean cADC (0.877) and rcADC (0.819) values in enhancing regions were significantly higher than those for ADC0Hz5th (0.595; all p < 0.001). CONCLUSIONS: The time-dependent diffusion MRI parameters may be useful for differentiating brain metastases from glioblastomas.

The usefulness of machine-learning-based evaluation of clinical and pretreatment 18F-FDG-PET/CT radiomic features for predicting prognosis in patients with laryngeal cancer.

OBJECTIVE: To examine whether machine learning (ML) analyses involving clinical and 18F-FDG-PET-based radiomic features are helpful in predicting prognosis in patients with laryngeal cancer. METHODS: This retrospective study included 49 patients with laryngeal cancer who underwent18F-FDG-PET/CT before treatment, and these patients were divided into the training (n = 34) and testing (n = 15) cohorts.Seven clinical (age, sex, tumor size, T stage, N stage, Union for International Cancer Control stage, and treatment) and 40 18F-FDG-PET-based radiomic features were used to predict disease progression and survival. Six ML algorithms (random forest, neural network, k-nearest neighbors, naïve Bayes, logistic regression, and support vector machine) were used for predicting disease progression. Two ML algorithms (cox proportional hazard and random survival forest [RSF] model) considering for time-to-event outcomes were used to assess progression-free survival (PFS), and prediction performance was assessed by the concordance index (C-index). RESULTS: Tumor size, T stage, N stage, GLZLM_ZLNU, and GLCM_Entropy were the five most important features for predicting disease progression.In both cohorts, the naïve Bayes model constructed by these five features was the best performing classifier (training: AUC = 0.805; testing: AUC = 0.842). The RSF model using the five features (tumor size, GLZLM_ZLNU, GLCM_Entropy, GLRLM_LRHGE and GLRLM_SRHGE) exhibited the highest performance in predicting PFS (training: C-index = 0.840; testing: C-index = 0.808). CONCLUSION: ML analyses involving clinical and 18F-FDG-PET-based radiomic features may help predict disease progression and survival in patients with laryngeal cancer. ADVANCES IN KNOWLEDGE: ML approach using clinical and 18F-FDG-PET-based radiomic features has the potential to predict prognosis of laryngeal cancer.

Extracellular volume fraction derived from equilibrium contrast-enhanced CT as a diagnostic parameter in anterior mediastinal tumors.

PURPOSE: To assess the usefulness of extracellular volume (ECV) fraction derived from equilibrium contrast-enhanced CT (CECT) for diagnosing anterior mediastinal tumors. METHOD: This study included 161 histologically confirmed anterior mediastinal tumors (55 low-risk thymomas, 57 high-risk thymomas, 32 thymic carcinomas, and 17 malignant lymphomas) that were assessed by pretreatment CECT. ECV fraction was calculated using measurements obtained within the lesion and the aorta on unenhanced and equilibrium phase CECT. ECV fraction was compared among anterior mediastinal tumors using one-way ANOVA or t-test. Receiver-operating characteristic (ROC) curve analysis was performed to evaluate the ability of ECV fraction to differentiate thymic carcinomas/lymphomas from thymomas. RESULTS: ECV fraction differed significantly among the anterior mediastinal tumors (p < 0.001). ECV fraction of thymic carcinomas was significantly higher than those of low-risk thymomas, high-risk thymomas, and lymphomas (p < 0.001, p < 0.001, and p = 0.006, respectively). ECV fraction of lymphomas was significantly higher than that of low-risk thymomas (p < 0.001). ECV fraction was significantly higher in thymic carcinomas/lymphomas than in thymomas (40.1 % vs. 27.7 %, p < 0.001). The optimal cutoff value to differentiate thymic carcinomas/lymphomas from thymomas was 38.5 % (AUC, 0.805; 95 %CI, 0.736-0.863). CONCLUSIONS: ECV fraction derived from equilibrium CECT is helpful in diagnosing anterior mediastinal tumors. High ECV fraction is indicative of thymic carcinomas/lymphomas, particularly thymic carcinomas.

Cine MR feature tracking analysis for diagnosing thymic epithelial tumors: a feasibility study.

BACKGROUND: To assess the feasibility of the cine MR feature tracking technique for the evaluation of cardiovascular-induced morphological deformation in the diagnosis of thymic epithelial tumors (TETs). METHODS: Our study population consisted of 43 patients with pathologically proven TETs including 10 low-grade thymomas, 23 high-grade thymomas, and 10 thymic carcinomas. Cine MR images were acquired using a balanced steady-state free precession sequence with short periods of breath-hold in the axial and oblique planes in the slice with the largest lesion cross-sectional area. The tumor margin was manually delineated in the diastolic phase and was automatically tracked for all other cardiac phases. The change rates of the long-to-short diameter ratio (∆LSR) and tumor area (∆area) associated with pulsation were compared between the three pathological groups using the Kruskal-Wallis H test and the Mann-Whitney U test. A receiver-operating characteristic (ROC) curve analysis was performed to assess the ability of each parameter to differentiate thymic carcinomas from thymomas. RESULTS: ∆LSR and ∆area were significantly different among the three groups in the axial plane (p = 0.028 and 0.006, respectively) and in the oblique plane (p = 0.034 and 0.043, respectively). ∆LSR and ∆area values were significantly lower in thymic carcinomas than in thymomas in the axial plane (for both, p = 0.012) and in the oblique plane (p = 0.015 and 0.011, respectively). The area under the ROC curves for ∆LSR and ∆area for the diagnosis of thymic carcinoma ranged from 0.755 to 0.764. CONCLUSIONS: Evaluation of morphological deformation using cine-MR feature tracking analysis can help diagnose histopathological subtypes of TETs and identify thymic carcinomas preoperatively.

Left atrial CT volume and CHA2DS2-VASc score predict early pulmonary vein stump thrombus after left upper lobectomy.

The purpose of this study is to clarify the feasibility of left atrial (LA) volume measurement and CHA2DS2-VASc score for predicting the development of pulmonary vein (PV) stump thrombus after left upper lobectomy (LUL). The study population comprised 50 patients who underwent LUL for pulmonary lesions. All patients were evaluated for the development of PV stump thrombus at 7 days after LUL. LA volume was measured using preoperative CT and the CHA2DS2-VASc score was evaluated. LA volume and CHA2DS2-VASc score were compared between patients with and without the development of PV stump thrombus using the Mann-Whitney U test. Receiver-operating characteristic (ROC) curve analysis was performed to evaluate the accuracy of prediction of PV stump thrombus development. PV stump thrombus was detected in 17 (33.4%) of the 50 patients. LA volume was significantly greater in patients who developed PV stump thrombus than in those without thrombus (79.7 ± 19.4 vs. 66.6 ± 17.0 mL, p = 0.040). CHA2DS2-VASc score was significantly higher in patients with PV stump thrombosis than in those without thrombus (3.4 ± 1.5 vs. 2.5 ± 1.5, p = 0.039). Area under the ROC curve values for predicting PV stump thrombus were 0.679, 0.676, and 0.714 for LA volume, CHA2DS2-VASc score, and their combination, respectively. In conclusion, LA volume measured using preoperative CT and CHA2DS2-VASc score may help predict the development of PV stump thrombus after LUL.

Feasibility of iodine concentration and extracellular volume fraction measurement derived from the equilibrium phase dual-energy CT for differentiating thymic epithelial tumors.

PURPOSE: To assess the diagnostic feasibility of iodine concentration (IC) and extracellular volume (ECV) fraction measurement using the equilibrium phase dual-energy CT (DECT) for the evaluation of thymic epithelial tumors (TETs). MATERIALS AND METHODS: This study included 33 TETs (11 low-risk thymomas, 11 high-risk thymomas, and 11 thymic carcinomas) that were assessed by pretreatment DECT. IC was measured during the equilibrium phases and ECV fraction was calculated using IC of the thymic lesion and the aorta. IC and ECV fraction were compared among TET subtypes using the Kruskal-Wallis H test and Mann-Whitney U test. Receiver-operating characteristic (ROC) curve analysis was performed to evaluate the ability of IC and ECV fraction to diagnose thymic carcinoma. RESULTS: IC during the equilibrium phase and ECV fraction differed among the three TET groups (both p < 0.001). IC during the equilibrium phase and ECV fraction was significantly higher in thymic carcinomas than in thymomas (1.9 mg/mL vs. 1.2 mg/mL, p < 0.001; 38.2% vs. 25.9%, p < 0.001; respectively). The optimal cutoff values of IC during the equilibrium phase and of ECV fraction to diagnose thymic carcinoma were 1.5 mg/mL (AUC, 0.955; sensitivity, 100%; specificity, 90.9%) and 26.8% (AUC, 0.888; sensitivity, 100%; specificity, 72.7%), respectively. CONCLUSION: IC and ECV fraction measurement using DECT are helpful in diagnosing TETs. High IC during the equilibrium phase and high ECV fraction are suggestive of thymic carcinoma.

Hemodynamic features underlying pulmonary vein stump thrombus formation after left upper lobectomy: four-dimensional flow magnetic resonance imaging study.

BACKGROUND: We previously reported that to-and-fro blood movement near the pulmonary vein stump was particularly prominent after left upper lobectomy compared with other lobectomy, which may be the cause of the high susceptibility of pulmonary vein thrombus after left upper lobectomy. The aim of the present study was to compare the hemodynamics in patients who developed pulmonary vein thrombus after left upper lobectomy with those in patients who did not develop pulmonary vein thrombus using four-dimensional flow magnetic resonance imaging (4D MRI). METHODS: This was a retrospective evaluation of a prospectively collected clinical and radiological database of 37 patients who underwent 4D MRI 7 days after left upper lobectomy for lung cancer (n=37). We obtained two parameters by 4D MRI: the grade of to-and-fro blood movement and the flow energy loss around the pulmonary vein stump. The length of the pulmonary vein stump, a known risk factor for pulmonary vein thrombus, was also measured. RESULTS: According to a scatterplot of the grade of to-and-fro blood movement versus the flow energy loss, patients with pulmonary vein thrombus (n=15) were concentrated in an area which appears to be a 'dangerous' hemodynamic condition. There were few patients without pulmonary vein thrombus in this 'dangerous' area, except for one who unfortunately developed delayed pulmonary vein thrombus and cerebral infarction. We proposed a formula using the 4D MRI-derived parameters based on a stepwise multiple regression analysis that was more closely associated with the development of pulmonary vein thrombus than the length of the pulmonary vein stump (area under the receiver operating characteristics curve: 0.918 vs. 0.705, P=0.0500). CONCLUSIONS: We proposed the existence of a 'dangerous' hemodynamic condition responsible for pulmonary vein thrombus formation. 4D MRI before the development of pulmonary vein thrombus may help identify patients requiring preventive therapy against pulmonary vein thrombus and subsequent thromboembolic complications.

Dual-Energy CT-Derived Electron Density for Diagnosing Metastatic Mediastinal Lymph Nodes in Non-Small Cell Lung Cancer: Comparison With Conventional CT and FDG PET/CT Findings.

BACKGROUND. Accurate nodal staging is essential to guide treatment selection in patients with non-small cell lung cancer (NSCLC). To our knowledge, measurement of electron density (ED) using dual-energy CT (DECT) is unexplored for this purpose. OBJECTIVE. The purpose of our study was to assess the utility of ED from DECT in diagnosing metastatic mediastinal lymph nodes in patients with NSCLC in comparison with conventional CT and FDG PET/CT. METHODS. This retrospective study included 57 patients (36 men, 21 women; mean age, 68.4 ± 8.9 [SD] years) with NSCLC and surgically resected mediastinal lymph nodes who underwent preoperative DECT and FDG PET/CT. The patients had a total of 117 resected mediastinal lymph nodes (33 metastatic, 84 nonmetastatic). Two radiologists independently reviewed the morphologic features of nodes on the 120-kVp images and also measured the iodine concentration (IC) and ED of nodes using maps generated from DECT data; consensus was reached for discrepancies. Two different radiologists assessed FDG PET/CT examinations in consensus for positive node uptake. Diagnostic performance was evaluated for individual and pairwise combinations of features. RESULTS. The sensitivity, specificity, and accuracy for nodal metastasis were 15.2%, 98.8%, and 75.2% for the presence of necrosis, respectively; 54.5%, 85.7%, and 76.9% for short-axis diameter greater than 8.5 mm; 63.6%, 73.8%, and 70.9% for long-axis diameter greater than 13.0 mm; 51.5%, 79.8%, and 71.8% for attenuation on 120-kVp images of 95.8 HU or less; 87.9%, 58.3%, and 66.7% for ED of 3.48 × 1023/cm3 or less; and 66.7%, 75.0%, and 72.6% for positive FDG uptake. Among pairwise combinations of features, accuracy was highest for the combination of ED and short-axis diameter (accuracy, 82.9%; sensitivity, 54.5%; specificity, 94.0%) and the combination of ED and positive FDG uptake (accuracy, 82.1%; sensitivity, 60.6%; specificity, 90.5%); these accuracies were greater than those for the individual features (p < .05). The remaining combinations exhibited accuracies ranging from 74.4% to 77.8%. Interobserver agreement analysis showed an intraclass correlation coefficient of 0.90 for ED. IC was not significantly different between metastatic and nonmetastatic nodes (p = .18) and was excluded from the diagnostic performance analysis. CONCLUSION. ED derived from DECT may help diagnose metastatic lymph nodes in NSCLC given decreased ED in metastatic nodes. CLINICAL IMPACT. ED may complement conventional CT findings and FDG uptake on PET/CT in diagnosing metastatic nodes.

Preoperative CT Findings for Predicting Acute Exacerbation of Interstitial Pneumonia After Lung Cancer Surgery: A Multicenter Case-Control Study.

BACKGROUND. Acute exacerbation (AE) is a life-threatening complication of inter-stitial pneumonia (IP). Thoracic surgery may trigger AE. OBJECTIVE. The purpose of this study is to explore the role of preoperative CT findings in predicting postoperative AE in patients with IP and lung cancer. METHODS. This retrospective case-control study included patients from 22 institutions who had IP and underwent thoracic surgery for lung cancer. AE was diagnosed on the basis of symptoms and imaging findings noted within 30 days after surgery and the absence of alternate causes. For each patient with AE, two control patients without AE were identified. After exclusions, the study included 92 patients (78 men and 14 women; 31 with AE [the AE group] and 61 without AE [the no-AE group]; mean age, 72 years). Two radiologists independently reviewed preoperative thin-slice CT examinations for pulmonary findings and resolved differences by consensus. The AE and no-AE groups were compared using the Fisher exact and Mann-Whitney U tests. Multivariable logistic regression was performed. Interreader agreement was assessed by kappa coefficients. RESULTS. A total of 94% of patients in the AE group underwent segmentectomy or other surgery that was more extensive than wedge resection versus 75% in the no-AE group (p = .046). The usual IP pattern was present in 58% of the AE group versus 74% of the no-AE group (p = .16). According to subjective visual scoring, the mean (± SD) ground-glass opacity (GGO) extent was 6.3 ± 5.4 in the AE group versus 3.9 ± 3.8 in the no-AE group (p = .03), and the mean consolidation extent was 0.5 ± 1.2 in the AE group versus 0.1 ± 0.3 in the no-AE group (p = .009). Mean pulmonary trunk diameter was 28 ± 4 mm in the AE group versus 26 ± 3 mm in the no-AE group (p = .02). In a model of CT features only, independent predictors of AE (p < .05) were GGO extent (odds ratio [OR], 2.8), consolidation extent (OR, 9.4), and pulmonary trunk diameter (OR, 4.2); this model achieved an AUC of 0.75, a PPV of 71%, and an NPV of 77% for AE. When CT and clinical variables were combined, undergoing segmentectomy or more extensive surgery also independently predicted AE (OR, 8.2; p = .02). CONCLUSION. The presence of GGO, consolidation, and pulmonary trunk enlargement on preoperative CT predicts AE in patients with IP who are undergoing lung cancer surgery. CLINICAL IMPACT. Patients with IP and lung cancer should be carefully managed when predictive CT features are present. Wedge resection, if possible, may help reduce the risk of AE in these patients. TRIAL REGISTRATION. University Hospital Medical Information Clinical Trial Registry UMIN000029661.

Application of a machine learning approach to characterization of liver function using 99mTc-GSA SPECT/CT.

PURPOSE: To assess the utility of a machine-learning approach for predicting liver function based on technetium-99 m-galactosyl serum albumin (99mTc-GSA) single photon emission computed tomography (SPECT)/CT. METHODS: One hundred twenty-eight patients underwent a 99mTc-GSA SPECT/CT-based liver function evaluation. All were classified into the low liver-damage or high liver-damage group. Four clinical (age, sex, background liver disease and histological type) and 8 quantitative 99mTc-GSA SPECT/CT features (receptor index [LHL15], clearance index [HH15], liver-SUVmax, liver-SUVmean, heart-SUVmax, metabolic volume of liver [MVL], total lesion GSA [TL-GSA, liver-SUVmean × MVL] and SUVmax ratio [liver-SUVmax/heart-SUVmax]) were obtained. To predict high liver damage, a machine learning classification with features selection based on Gini impurity and principal component analysis (PCA) were performed using a support vector machine and a random forest (RF) with a five-fold cross-validation scheme. To overcome imbalanced data, stratified sampling was used. The ability to predict high liver damage was evaluated using a receiver operating characteristic (ROC) curve analysis. RESULTS: Four indices (LHL15, HH15, heart SUVmax and SUVmax ratio) yielded high areas under the ROC curves (AUCs) for predicting high liver damage (range: 0.89-0.93). In a machine learning classification, the RF with selected features (heart SUVmax, SUVmax ratio, LHL15, HH15, and background liver disease) and PCA model yielded the best performance for predicting high liver damage (AUC = 0.956, sensitivity = 96.3%, specificity = 90.0%, accuracy = 91.4%). CONCLUSION: A machine-learning approach based on clinical and quantitative 99mTc-GSA SPECT/CT parameters might be useful for predicting liver function.

Adding Delayed Phase Images to Dual-Phase Contrast-Enhanced CT Increases Sensitivity for Small Pancreatic Ductal Adenocarcinoma.

BACKGROUND. Contrast-enhanced CT performed for pancreatic ductal adeno-carcinoma (PDAC) detection traditionally uses a dual-phase (pancreatic and portal venous) protocol. However, PDAC may exhibit isoattenuation in these phases, hindering detection. OBJECTIVE. The purpose of this study was to assess the impact on diagnostic performance in detection of small PDAC when a delayed phase is added to dual-phase contrast-enhanced CT. METHODS. A database of 571 patients who underwent triple-phase (pancreatic, portal venous, and delayed) contrast-enhanced MDCT between January 2017 and March 2020 for suspected pancreatic tumor was retrospectively reviewed. A total of 97 patients had pathologically confirmed small PDAC (mean size, 22 mm; range, 7-30 mm). Twenty control patients had no pancreatic tumor suspected on CT, on initial MRI and follow-up CT, or on MRI after 12 months or longer. Three radiologists independently reviewed dual-phase and triple-phase images. Two additional radiologists assessed tumors' visual attenuation on each phase, reaching consensus for differences. Performance of dual- and triple-phase images were compared using ROC analysis, McNemar test, and Fisher exact test. RESULTS. AUC was higher (p < .05) for triple-phase than dual-phase images for all observers (observer 1, 0.97 vs 0.94; observer 2, 0.97 vs 0.94; observer 3, 0.97 vs 0.95). Sensitivity was higher (p < .001) for triple-phase than dual-phase images for all observers (observer 1, 74.2% [72/97] vs 59.8% [58/97]; observer 2, 88.7% [86/97] vs 71.1% [69/97]; observer 3, 86.6% [84/97] vs 72.2% [70/97]). Specificity, PPV, and NPV did not differ between image sets for any reader (p ≥ .05). Seventeen tumors showed pancreatic phase visual isoattenuation, of which nine showed isoattenuation and eight hyperattenuation in the delayed phase. Of these 17 tumors, 16 were not detected by any observer on dual-phase images; of these 16, six were detected by at least two observers and five by at least one observer on triple-phase images. Visual attenuation showed excellent interob-server agreement (κ = 0.89-0.96). CONCLUSION. Addition of a delayed phase to pancreatic and portal venous phase CT increases sensitivity for small PDAC without loss of specificity, partly related to delayed phase hyperattenuation of some small PDACs showing pancreatic phase isoattenuation. CLINICAL IMPACT. Addition of a delayed phase may facilitate earlier PDAC detection and thus improved prognosis.

Pulmonary hamartoma: Feasibility of dual-energy CT detection of intranodular fat.

We have reported 2 cases of pulmonary hamartoma focusing on detecting intranodular fat, which is one of CT features suggestive of pulmonary hamartoma, using dual-energy CT analyses. For patient 1, a 73-year-old man was pointed out to have a nodular opacity on chest radiograph of pretreatment workup for retinal detachment. In patient 2, a 66-year-old woman with uterine carcinoma admitted for preoperative assessment. Both patients underwent dual-energy CT examination and the pulmonary lesions exhibited a downward-sloping curve at lower X-ray energies on attenuation curve of virtual monochromatic images, which suggested fatty tissue. Dual-energy CT analysis can help diagnose pulmonary hamartoma with detection of intralesional fatty tissue.

Differentiating malignant from benign salivary gland lesions: a multiparametric non-contrast MR imaging approach.

The purpose of this study is to determine whether multiparametric non-contrast MR imaging including diffusion-weighted imaging (DWI), arterial spin labeling (ASL), and amide proton transfer (APT) weighted imaging can help differentiate malignant from benign salivary gland lesions. The study population consisted of 42 patients, with 31 benign and 11 malignant salivary gland lesions. All patients were evaluated using DWI, three-dimensional pseudo-continuous ASL, and APT-weighted imaging on 3 T MR imaging before treatment. Apparent diffusion coefficient (ADC), tumor blood flow (TBF), and APT-related signal intensity (APTSI) values within the lesion were compared between the malignant and benign lesions by Mann-Whitney U test. For each parameter, optimal cutoff values were chosen using a threshold criterion that maximized the Youden index for predicting malignant lesions. The performance of ADC, TBF, APTSI, individually and combined, was evaluated in terms of diagnostic ability for malignant lesions. Diagnostic performance was compared by McNemar test. APTSI was significantly higher in malignant lesions (2.18 ± 0.89%) than in benign lesions (1.57 ± 1.09%, p = 0.047). There was no significant difference in ADC or TBF between benign and malignant lesions (p = 0.155 and 0.498, respectively). The accuracy of ADC, TBF, and APTSI for diagnosing malignant lesions was 47.6%, 50.0%, and 66.7%, respectively; whereas the accuracy of the three parameters combined was 85.7%, which was significantly higher than that of each parameter alone (p = 0.001, 0.001, and 0.008, respectively). Therefore, the combination of ADC, TBF, and APTSI can help differentiate malignant from benign salivary gland lesions.