Multitask deep learning on mammography to predict extensive intraductal component in invasive breast cancer.

OBJECTIVES: To develop a multitask deep learning (DL) algorithm to automatically classify mammography imaging findings and predict the existence of extensive intraductal component (EIC) in invasive breast cancer. METHODS: Mammograms with invasive breast cancers from 2010 to 2019 were downloaded for two radiologists performing image segmentation and imaging findings annotation. Images were randomly split into training, validation, and test datasets. A multitask approach was performed on the EfficientNet-B0 neural network mainly to predict EIC and classify imaging findings. Three more models were trained for comparison, including a single-task model (predicting EIC), a two-task model (predicting EIC and cell receptor status), and a three-task model (combining the abovementioned tasks). Additionally, these models were trained in a subgroup of invasive ductal carcinoma. The DeLong test was used to examine the difference in model performance. RESULTS: This study enrolled 1459 breast cancers on 3076 images. The EIC-positive rate was 29.0%. The three-task model was the best DL model with an area under the curve (AUC) of EIC prediction of 0.758 and 0.775 at the image and breast (patient) levels, respectively. Mass was the most accurately classified imaging finding (AUC = 0.915), followed by calcifications and mass with calcifications (AUC = 0.878 and 0.824, respectively). Cell receptor status prediction was less accurate (AUC = 0.625-0.653). The multitask approach improves the model training compared to the single-task model, but without significant effects. CONCLUSIONS: A mammography-based multitask DL model can perform simultaneous imaging finding classification and EIC prediction. CLINICAL RELEVANCE STATEMENT: The study results demonstrated the potential of deep learning to extract more information from mammography for clinical decision-making. KEY POINTS: • Extensive intraductal component (EIC) is an independent risk factor of local tumor recurrence after breast-conserving surgery. • A mammography-based deep learning model was trained to predict extensive intraductal component close to radiologists' reading. • The developed multitask deep learning model could perform simultaneous imaging finding classification and extensive intraductal component prediction.

Risk factors of stent migration in esophageal cancer patients who underwent fully-covered self-expanding metal stents for malignant dysphagia or tracheoesophageal fistula.

BACKGROUND: More than 50% of esophageal cancer patients are diagnosed with advanced diseases and commonly experience dysphagia, some of whom even have tracheoesophageal fistula. Self-expandable metal stent (SEMS) is one of the recommended palliative methods, although complications such as chest pain and stent migration are not uncommon. The goal of this study was to examine the predictors of stent migration. METHODS: We conducted a retrospective cohort study to include patients with esophageal cancer and dysphagia/tracheoesophageal fistula. Clinicopathological information, stent characteristics and patient outcomes were collected for analysis, while side-effects of SEMS were recorded, potential predictors were examined, and patients' nutritional outcomes were compared in the migration and non-migration groups. RESULTS: A total of 54 patients with esophageal cancer who received fully covered SEMS between 2013 and 2022 were included. We found tumor across the esophagogastric junction (adjusted odds ratio (OR) = 32.64, P = 0.01) and the female sex (adjusted OR = 12.5, P = 0.02) were significant predictors for stent migration. There was a decreasing tendency in body mass index/body weight in migration and non-migration groups, but the former had a steeper downslope. CONCLUSION: Fully covered SEMS is a safe and effective strategy to palliate dysphagia or fistula. Tumor across esophagogastric junction and the female sex were higher risk predictors of stent migration. A careful patient selection would optimize the effects of SEMS placement, especially in those with short-expected lifespan.

Recurrent Exacerbations and Evolution into Polymyositis in a Patient with Interstitial Pneumonia with Autoimmune Features: A Case Report and Literature Review.

Interstitial pneumonia with autoimmune features (IPAF) is a new disease entity proposed in 2015. Numerous questions regarding IPAF require clarification, including diagnostic criteria, standard managements for stable disease and exacerbation, and prognosis. We report a case of a 67-year-old Asian woman who presented with progressive dyspnea. Chest computed tomography (CT) scans revealed nonspecific interstitial pneumonia. Serologic testing indicated positive anti-Jo-1 without presence of extrathoracic manifestations. An IPAF diagnosis was made after a multidisciplinary discussion. The patient experienced a severe exacerbation requiring mechanical ventilation, and she was successfully salvaged with methylprednisolone pulse therapy and single-dose cyclophosphamide. During the one-year follow-up, she reported bilateral leg muscle weakness with noticeably elevated serum creatine kinase, suggesting polymyositis. The development of malignancy was also noted 15 months after the initial presentation, and the patient eventually died. This report demonstrated successful salvage treatment with glucocorticoid pulse therapy for IPAF with acute exacerbation. However, the maintenance therapy failed to control disease progression. The treatment strategies for exacerbation and stable disease in IPAF remain unknown and need further studies. Given the high risk of evolution into a defined connective tissue disease (CTD), regular evaluation of the clinical features and biomarkers of CTDs is essential for patients with IPAF.

Comparison of LCModel and SAGE in Analysis of Brain Metabolite Concentrations-A study of Patients with Mild Cognitive Impairment.

PURPOSE: The purpose of this study was to compare brain metabolite concentration ratios determined by LCModel and Spectroscopy Analysis by General Electric (SAGE) quantitative methods to elucidate the advantages and disadvantages of each method. MATERIALS AND METHODS: A total of 10 healthy volunteers and 10 patients with mild cognitive impairment (MCI) were recruited in this study. A point-resolved spectroscopy (PRESS) sequence was used to obtain the brain magnetic resonance spectroscopy (MRS) spectra of the volunteers and patients, as well as the General Electric (GE) MRS-HD-sphere phantom. The brain metabolite concentration ratios were estimated based on the peak area obtained from both LCModel and SAGE software. Three brain regions were sampled for each volunteer or patient, and 20 replicates were acquired at different times for the phantom analysis. RESULTS: The metabolite ratios of the GE phantom were estimated to be myo-inositol (mI)/creatine (Cr): 0.70 ± 0.01, choline (Cho)/Cr: 0.37 ± 0.00, N-acetylaspartate (NAA)/Cr: 1.26 ± 0.02, and NAA/mI: 1.81 ± 0.04 by LCModel, and mI/Cr: 0.88 ± 0.15, Cho/Cr: 0.35 ± 0.01, NAA/Cr: 1.33 ± 0.03, and NAA/mI: 1.55 ± 0.26 by SAGE. In the healthy volunteers and MCI patients, the ratios of mI/Cr and Cho/Cr estimated by LCModel were higher than those estimated by SAGE. In contrast, the ratio of NAA/Cr estimated by LCModel was lower than that estimated by SAGE. CONCLUSION: Both methods were acceptable in estimating brain metabolite concentration ratios. However, LCModel was marginally more accurate than SAGE because of its full automation, basis set, and user independency.

Pleural Tags on CT Scans to Predict Visceral Pleural Invasion of Non-Small Cell Lung Cancer That Does Not Abut the Pleura.

PURPOSE: To evaluate the association of pleural tags with visceral pleural invasion of non-small cell lung cancer (NSCLC) that does not abut the pleural surface. MATERIALS AND METHODS: This retrospective study was approved by the institutional review board. Informed consent was waived. The study of NSCLC that does not abut the pleura in 141 patients (44 patients [31.2%] with visceral pleural invasion proved by pathologic analysis and 97 patients [68.8%] without pleural invasion) was conducted at a single tertiary center. The pleural tags were classified into three types (type 1, one or more linear pleural tag; type 2, one or more linear pleural tag with soft tissue component at the pleural end; and type 3, one or more soft tissue cord-like pleural tag) and prioritized into types 3, 2, and 1 when more than one type was present. Diagnostic accuracy, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and positive likelihood ratio (LR) were calculated. RESULTS: In the absence of pleural tags, no pleural invasion was found. The presence of type 2 pleural tags was moderately associated with visceral pleural invasion with the following results: positive LR, 5.06; accuracy, 71%; sensitivity, 36.4%; specificity, 92.8%; PPV, 76.2%; and NPV, 69.6%. Type 1 pleural tags provided weak evidence to rule out visceral pleural invasion (positive LR, 0.38). Type 3 pleural tags indicated minimal increase in the likelihood of visceral pleural invasion (positive LR, 1.68). CONCLUSION: Type 2 pleural tags on conventional CT images can increase the accuracy of early diagnosis of visceral pleural invasion by NSCLC that does not abut the pleura.

Diagnostic Performance of Split-Bolus Portal Venous Phase Dual-Energy CT Urography in Patients With Hematuria.

OBJECTIVE: The purpose of this study is to evaluate the diagnostic performance of split-bolus portal venous phase dual-energy CT (DECT) urography in patients with hematuria. MATERIALS AND METHODS: True unenhanced and split-bolus portal venous phase contrast-enhanced weighted-average images were obtained in 171 patients with hematuria. Virtual unenhanced and iodine-overlay images were reconstructed from contrast-enhanced 80-kVp and tin-filtered 140-kVp scans. Images were independently reviewed by two radiologists who were blinded to the final diagnoses in two separate reading sessions: virtual unenhanced and iodine-overlay images (single phase) in the first session and true unenhanced and contrast-enhanced weighted-average images (dual phase) in the second session (mean ± SD, 52 ± 8 days later). Sensitivity, specificity, and accuracy of mass detection were calculated from the data of both reading sessions. The number of calculi detected on virtual unenhanced images was compared with the number detected on true unenhanced images. The difference in radiation dose between the single- and dual-phase protocols was calculated. The statistical significance was determined by ANOVA. RESULTS: The sensitivity, specificity, and accuracy were 98.7%, 98.9%, and 98.8%, respectively, for the single-phase approach to malignant mass detection and 98.7%, 97.9%, and 98.3%, respectively, for the dual-phase approach (p > 0.05 for all comparisons). The overall sensitivity of stone detection was 86.7% (39/45) for virtual unenhanced images. Omitting the unenhanced scan reduced the mean radiation dose from 15.4 to 6.7 mSv. CONCLUSION: The diagnostic performance of both the single- and dual-phase approaches of portal venous phase split-bolus DECT urography is equally good in patients with hematuria, and single-phase acquisition has the added benefit of radiation reduction.

Split-Bolus Portal Venous Phase Dual-Energy CT Urography: Protocol Design, Image Quality, and Dose Reduction.

OBJECTIVE: The purpose of this study is to evaluate the image quality of split-bolus portal venous phase urography and the potential reduction of radiation dose by using a second-generation dual-source dual-energy CT (DECT) scanner. MATERIALS AND METHODS: DECT urography was performed in 84 patients. Unenhanced CT was performed 20 minutes after drinking 800 mL of water. The split-bolus protocol consisted of a sequence of injections, as follows: 200 mL of normal saline (2.0 mL/s), 50 mL of contrast medium (2.5 mL/s) at 0 second, 70 mL of contrast medium (2.5 mL/s) at 360 seconds, and a saline flush of 25 mL. The scan was started at 420 seconds. Virtual unenhanced images were reconstructed from contrast-enhanced images. The mean CT density and signal-to-noise ratio (SNR) of the renal parenchyma, vessels, upper urinary tract, normal reference tissues, and tumors were measured for image quantitative analysis. Image quality and opacification of the collecting systems were rated by two radiologists using 3- or 4-point scales. RESULTS: The SNR of all measured sites, except the renal pelvis, showed a statistically significant correlation (p < 0.001) between the true unenhanced and virtual unenhanced images. The overall sensitivity of stone detection was 87.5% (28/32) in virtual unenhanced images. Image quality of the renal parenchyma, arteries, and veins was excellent in 59.5%, 75.0%, and 97.6% of cases, respectively. Opacification of the intrarenal collecting systems, proximal, middle, and distal ureters, and bladder was complete in 92.9%, 83.9%, 78.6%, 77.4%, and 26.2% of patients, respectively. Omitting the unenhanced scan can reduce the mean radiation dose from 15.6 to 6.7 mSv. CONCLUSION: Portal venous phase split-bolus DECT urography provides sufficient image quality with potential to reduce radiation exposure.

Lowering radiation dose during dedicated colorectal cancer MDCT: comparison of low tube voltage and sinogram-affirmed iterative reconstruction at 80 kVp versus blended dual-energy images in a population of patients with low body mass index.

PURPOSE: To assess the diagnostic accuracy, cancer staging, image quality, and radiation dose of 80-kVp computed tomography (CT) images for patients with colorectal cancers (CRCs) using sinogram-affirmed iterative reconstruction (SAFIRE). METHODS: Sixty-four consecutive patients (mean weight 62.5  ±  11.3 kg, mean BMI 24.1  ±  3.3 kg/m(2)) with known CRC underwent dual-energy CT. Data were reconstructed as a weighted average (WA) 120-kVp dataset. Both filtered back projection (FBP) and SAFIRE were applied to reconstruct the WA 120-Kvp (Protocol A, B) and 80-kVp (Protocol C, D) image sets. The image noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) of the cancers, the normal reference tissues, and the effective dose for each protocol were assessed. The cancer detection, staging, and image quality were evaluated. Analysis of variance was used for statistical analysis. RESULTS: Compared with the FBP datasets at WA 120-kVp (Protocol A) and 80-kVp (Protocol C), the SAFIRE-reconstructed images (Protocols B, D) demonstrated significantly lower image noise (P  <  0.0083). Protocol D yielded significantly higher CNRs and SNRs for the CRCs and normal reference tissues than did Protocols A and C (P  <  0.0083). Protocol D also exhibited a significantly higher CNR for the CRC and some normal reference tissues than did Protocol B (P  <  0.0083). For hypovascular liver metastases (n  =  10), Protocol D yielded better SNRs and significantly higher CNRs than did Protocol A (P  <  0.0083). Overall, accuracy for tumor staging and liver metastasis was 95.3% (61/64) and 100%, respectively, in all of the 4 protocols. The mean effective dose decreased 41% from the WA 120-kVp to the 80-kVp protocols (6.23 vs. 3.68 mSv). CONCLUSIONS: The 80-kVp technique with SAFIRE provided high SNR, high CNR, and good accuracy for staging in nonobese patients with CRC. Our study results should be extrapolated to patient populations with a high BMI with caution. Further studies of high BMI patients are therefore warranted.

Cost-Effectiveness of Low-Dose Computed Tomography Screenings for Lung Cancer in High-Risk Populations: A Markov Model.

Background: Domestic and foreign studies on lung cancer have been oriented to the medical efficacy of low-dose computed tomography (LDCT), but there is a lack of studies on the costs, value and cost-effectiveness of the treatment. There is a scarcity of conclusive evidence regarding the cost-effectiveness of LDCT within the specific context of Taiwan. This study is designed to address this gap by conducting a comprehensive analysis of the cost-effectiveness of LDCT and chest X-ray (CXR) as screening methods for lung cancer. Methods: Markov decision model simulation was used to estimate the cost-effectiveness of biennial screening with LDCT and CXR based on a health provider perspective. Inputs are based on probabilities, health status utility (quality-adjusted life years (QALYs)), costs of lung cancer screening, diagnosis, and treatment from the literatures, and expert opinion. A total of 1,000 simulations and five cycles of Markov bootstrapping simulations were performed to compare the incremental cost-utility ratio (ICUR) of these two screening strategies. Probability and one-way sensitivity analyses were also performed. Results: The ICUR of early lung cancer screening compared LDCT to CXR is $-24,757.65/QALYs, and 100% of the probability agree to adopt it under a willingness-to-pay (WTP) threshold of the Taiwan gross domestic product (GDP) per capita ($35,513). The one-way sensitivity analysis also showed that ICUR depends heavily on recall rate. Based on the prevalence rate of 39.7 lung cancer cases per 100,000 people in 2020, it could be estimated that LDCT screening for high-risk populations could save $17,154,115. Conclusion: LDCT can detect more early lung cancers, reduce mortality and is cost-saving than CXR in a long-term simulation of Taiwan's healthcare system. This study provides valuable insights for healthcare decision-makers and suggests analyzing cost-effectiveness for additional variables in future research.

Low-dose CT screening among never-smokers with or without a family history of lung cancer in Taiwan: a prospective cohort study.

BACKGROUND: In Taiwan, lung cancers occur predominantly in never-smokers, of whom nearly 60% have stage IV disease at diagnosis. We aimed to assess the efficacy of low-dose CT (LDCT) screening among never-smokers, who had other risk factors for lung cancer. METHODS: The Taiwan Lung Cancer Screening in Never-Smoker Trial (TALENT) was a nationwide, multicentre, prospective cohort study done at 17 tertiary medical centres in Taiwan. Eligible individuals had negative chest radiography, were aged 55-75 years, had never smoked or had smoked fewer than 10 pack-years and stopped smoking for more than 15 years (self-report), and had one of the following risk factors: a family history of lung cancer; passive smoke exposure; a history of pulmonary tuberculosis or chronic obstructive pulmonary disorders; a cooking index of 110 or higher; or cooking without using ventilation. Eligible participants underwent LDCT at baseline, then annually for 2 years, and then every 2 years up to 6 years thereafter, with follow-up assessments at each LDCT scan (ie, total follow-up of 8 years). A positive scan was defined as a solid or part-solid nodule larger than 6 mm in mean diameter or a pure ground-glass nodule larger than 5 mm in mean diameter. Lung cancer was diagnosed through invasive procedures, such as image-guided aspiration or biopsy or surgery. Here, we report the results of 1-year follow-up after LDCT screening at baseline. The primary outcome was lung cancer detection rate. The p value for detection rates was estimated by the χ2 test. Univariate and multivariable logistic regression analyses were used to assess the association between lung cancer incidence and each risk factor. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of LDCT screening were also assessed. This study is registered with ClinicalTrials.gov, NCT02611570, and is ongoing. FINDINGS: Between Dec 1, 2015, and July 31, 2019, 12 011 participants (8868 females) were enrolled, of whom 6009 had a family history of lung cancer. Among 12 011 LDCT scans done at baseline, 2094 (17·4%) were positive. Lung cancer was diagnosed in 318 (2·6%) of 12 011 participants (257 [2·1%] participants had invasive lung cancer and 61 [0·5%] had adenocarcinomas in situ). 317 of 318 participants had adenocarcinoma and 246 (77·4%) of 318 had stage I disease. The prevalence of invasive lung cancer was higher among participants with a family history of lung cancer (161 [2·7%] of 6009 participants) than in those without (96 [1·6%] of 6002 participants). In participants with a family history of lung cancer, the detection rate of invasive lung cancer increased significantly with age, whereas the detection rate of adenocarcinoma in situ remained stable. In multivariable analysis, female sex, a family history of lung cancer, and age older than 60 years were associated with an increased risk of lung cancer and invasive lung cancer; passive smoke exposure, cumulative exposure to cooking, cooking without ventilation, and a previous history of chronic lung diseases were not associated with lung cancer, even after stratification by family history of lung cancer. In participants with a family history of lung cancer, the higher the number of first-degree relatives affected, the higher the risk of lung cancer; participants whose mother or sibling had lung cancer were also at an increased risk. A positive LDCT scan had 92·1% sensitivity, 84·6% specificity, a PPV of 14·0%, and a NPV of 99·7% for lung cancer diagnosis. INTERPRETATION: TALENT had a high invasive lung cancer detection rate at 1 year after baseline LDCT scan. Overdiagnosis could have occurred, especially in participants diagnosed with adenocarcinoma in situ. In individuals who do not smoke, our findings suggest that a family history of lung cancer among first-degree relatives significantly increases the risk of lung cancer as well as the rate of invasive lung cancer with increasing age. Further research on risk factors for lung cancer in this population is needed, particularly for those without a family history of lung cancer. FUNDING: Ministry of Health and Welfare of Taiwan.

Elevated intramuscular adipose tissue content with a high Ishak fibrosis stage (>3) had a negative effect on liver regeneration in cirrhotic patients undergoing portal vein embolization.

This study investigated the relationship between body composition parameters and changes in future liver remnant volume (FLRV) in hepatocellular carcinoma (HCC) patients undergoing portal vein embolization (PVE) in preparation for right hepatectomy. This retrospective study enrolled 21 patients between May 2013 and October 2020. Body composition parameters, including skeletal muscle attenuation (SMA), skeletal muscle mass index (SMI), intramuscular adipose tissue content (IMAC), and visceral-to-subcutaneous adipose tissue area ratio (VSR), were measured by computed tomography (CT) prior to PVE. Liver volumetry was measured before and at least 5 weeks after PVE. The mean interval between two CT volumetries was 9.1 ± 4.9 weeks, the mean value of increase in FLRV (ΔFLRV) was 236.0 ± 118.3 cm3 , the ratio of increased FLRV (ΔFLRV%) was 55.7 ± 29.4%, and the rate of increased FLRV was 31.0 ± 18.8 (cm3 /week). Subjects with high IMAC showed significantly lower (p = 0.044) ΔFLRV% than those with normal IMAC. Furthermore, ΔFLRV% was linearly reduced (p for trend = 0.043) among those with low Ishak fibrosis stage (<3) + normal IMAC (76.1 ± 36.8%), those with low Ishak fibrosis stage (<3) + high IMAC or high Ishak fibrosis stage (>3) + normal IMAC (54.0 ± 24.1%), and those with high Ishak fibrosis stage (>3) + low IMAC (28.7 ± 1.6%) (p for trend = 0.043). Our data indicated that high IMAC with a high Ishak fibrosis stage (>3) had a significant negative effect on ΔFLRV%.

Unique Pulmonary Hypertension in Young Children: A Case Series Study.

Pediatric pulmonary hypertension (PH) has a similar clinical presentation to the adult disease but is associated with several additional disorders and challenges that require a specific approach for their fulminant course. With improved care for premature infants, various forms of pulmonary vascular disease have been found in children that did not previously exist. Pediatric PH can begin in utero, resulting in pulmonary vascularity growth abnormalities that may persist into adulthood. Here, we retrospectively reviewed several unique pediatric PH cases from 2000 to 2020 at Kaohsiung Medical University Hospital, Taiwan, a tertiary teaching hospital. Their comorbidities varied and included surfactant dysfunction, bronchopulmonary dysplasia, premature closure of the ductus arteriosus, high levels of renin and aldosterone, and Swyer-James-Macleod syndrome. Their clinical profiles, radiological characteristics, echocardiography, pulmonary angiogram, and therapeutic regimens were recorded. Further, because the underlying causes of pediatric PH were complex and markedly different according to age, adult PH classification may not be applicable to pediatric PH in all settings. We also classified these cases using different systems, including the Panama classification and the Sixth World Symposium on PH, and compared their advantages and disadvantages.

Long-Term Study on Therapeutic Strategy for Treatment of Eisenmenger Syndrome Patients: A Case Series Study.

Eisenmenger syndrome (ES) refers to congenital heart diseases (CHD) with reversal flow associated with increased pulmonary pressure and irreversible pulmonary vascular remodeling. Previous reports showed limited therapeutic strategies in ES. In this study, 5 ES patients (2 males and 3 females), who had been followed regularly at our institution from 2010 to 2019, were retrospectively reviewed. We adopted an add-on combination of sildenafil, bosentan, and iloprost and collected the clinical characteristics and outcomes as well as findings of echocardiography, computed tomography, pulmonary perfusion-ventilation scans, positron emission tomography, and biomarkers. The age of diagnosis in these ES patients ranged from 23 to 54 years (38.2 ± 11.1 years; mean ± standard deviation), and they were followed for 7 to 17 years. Their mean pulmonary arterial pressure and pulmonary vascular resistance index were 56.4 ± 11.3 mmHg and 24.7 ± 8.5 WU.m2, respectively. Intrapulmonary arterial thrombosis was found in 4 patients, ischemic stroke was noted in 2 patients, and increased glucose uptake of the right ventricle was observed in 4 patients. No patient mortality was seen within 5 years of follow-up. Subsequently, 2 patients died of right ventricular failure, 1 died of sepsis related to brain abscess, and another died of sudden death. The life span of these patients was 44-62 years. Although these patients showed longer survival, the beneficial data on specific-target pharmacologic interventions in ES is still preliminary. Thus, larger trials are warranted, and the study of cardiac remodeling in ES from various CHD should be explored.

Integration of Clinical and CT-Based Radiomic Features for Pretreatment Prediction of Pathologic Complete Response to Neoadjuvant Systemic Therapy in Breast Cancer.

The purpose of the present study was to examine the potential of a machine learning model with integrated clinical and CT-based radiomics features in predicting pathologic complete response (pCR) to neoadjuvant systemic therapy (NST) in breast cancer. Contrast-enhanced CT was performed in 329 patients with breast tumors (n = 331) before NST. Pyradiomics was used for feature extraction, and 107 features of seven classes were extracted. Feature selection was performed on the basis of the intraclass correlation coefficient (ICC), and six ICC thresholds (0.7−0.95) were examined to identify the feature set resulting in optimal model performance. Clinical factors, such as age, clinical stage, cancer cell type, and cell surface receptors, were used for prediction. We tried six machine learning algorithms, and clinical, radiomics, and clinical−radiomics models were trained for each algorithm. Radiomics and clinical−radiomics models with gray level co-occurrence matrix (GLCM) features only were also built for comparison. The linear support vector machine (SVM) regression model trained with radiomics features of ICC ≥0.85 in combination with clinical factors performed the best (AUC = 0.87). The performance of the clinical and radiomics linear SVM models showed statistically significant difference after correction for multiple comparisons (AUC = 0.69 vs. 0.78; p < 0.001). The AUC of the radiomics model trained with GLCM features was significantly lower than that of the radiomics model trained with all seven classes of radiomics features (AUC = 0.85 vs. 0.87; p = 0.011). Integration of clinical and CT-based radiomics features was helpful in the pretreatment prediction of pCR to NST in breast cancer.

Zone-based analysis for automated detection of abnormalities in chest radiographs.

PURPOSE: The aim of this study was to develop an automated method for detection of local texture-based and density-based abnormalities in chest radiographs. METHODS: The method was based on profile analysis to detect abnormalities in chest radiographs. In the method, one density-based feature, Density Symmetry Index, and two texture-based features, Roughness Maximum Index and Roughness Symmetry Index, were used to detect abnormalities in the lung fields. In each chest radiograph, the lung fields were divided into four zones initially and then the method was applied to each zone separately. For each zone, Density Symmetry Index was obtained from the projection profile of each zone, and Roughness Maximum Index and Roughness Symmetry Index were obtained by measuring the roughness of the horizontal profiles via moving average technique. Linear discriminant analysis was used to classify normal and abnormal cases based on the three indices. The discriminant performance of the method was evaluated using ROC analysis. RESULTS: The method was evaluated on a database of 250 normal and 250 abnormal chest images. In the optimized conditions, the zone-based performance Az of the method for zones 1, 2, 3, and 4 were 0.917, 0.897, 0.892, and 0.814, respectively, and the case-based performance Az of the method was 0.842. Our previous method for detection of gross abnormalities was also evaluated on the same database. The case-based performance of our previous method was 0.689. CONCLUSIONS: In comparing the previous method and the new method proposed in this study, there was a great improvement by the new method for detection of local texture-based and density-based abnormalities. The new method combined with the previous one has potential for screening abnormalities in chest radiographs.

Cerebral white matter hyperintensities predict functional stroke outcome.

BACKGROUND: Growing evidence suggests that white matter hyperintensities (WMHs) are implicated in stroke recurrence and mortality, and their location can be a critical factor. This study evaluated the impact of periventricular WMHs (PVWMHs) and subcortical WMHs (SWMHs) on poststroke functional outcomes. METHODS: Brain MRI was performed on 187 acute ischemic stroke patients (57.8% male; mean age = 64.3 years) recruited from the Kaohsiung Municipal Hsiao-Kang Hospital from February 2007 to January 2008. A Fazekas score >or=2 in the periventrcular or subcortical white matter was taken as presence of WMHs. Demographic data and risk factors for stroke were assessed. Functional stroke outcomes were evaluated 30 days after stroke using the Barthel Index (BI) and the modifiedRankin Scale (mRS). RESULTS: WMHs were inversely linked to favorable functional outcome measured by mRS (p = 0.001) and BI (p = 0.003). Evaluating different locations, PVWMHs were associated with unfavorable functional outcomes (p = 0.002, mRS; p = 0.001, BI). SWMHs were related to mRS (p = 0.026) but not BI (p = 0.069). After controlling other stroke risk factors, infarct volumes and initial stroke severity, PVWMHs were a significant indicator for both mRS (OR = 2.76; 95% CI = 1.03-7.40) and BI (OR = 3.07; 95% CI = 1.13-8.40), but SWMHs were not. CONCLUSIONS: Unfavorable functional stroke outcome is associated with MRI WMHs. In terms of location, PVWMHs but not SWMHs are related to functional stroke outcome.

Automatic assessment of bone age in Taiwanese children: A comparison of the Greulich and Pyle method and the Tanner and Whitehouse 3 method.

Proper bone age assessment is crucial for the clinical diagnosis and evaluation of treatment responses. We investigated the applicability of Greulich and Pyle (GP), and Tanner and Whitehouse 3 (TW3) methods for children in modern Taiwan, using computer-aided diagnosis. Hand and wrist radiographs were obtained from 611 children (3-17 years) who came to our emergency department due to trauma. Ages 0 to 2 years old were excluded because of a limited number of cases. Skeletal maturation was assessed using the BoneXpert (version 2.5.4.1 automated software), which determines GP and TW3 bone age. The two scoring systems were evaluated for comparing the chronological ages in each subgroup. In boys, mean GP bone age vs mean chronological ages were delayed for ages 3 to 11 and advanced for age 12 to 17. In girls, mean GP bone age vs mean chronological ages was delayed for ages 4 to 8 and 17, and advanced for ages 3 and 9 to 17. In boys, the mean TW3 bone ages vs mean chronological ages were delayed for ages 5 to 10 except age 8, and advanced for ages 3 to 4, 8, and 11 to 15. In girls, the mean TW3 bone ages vs mean chronological ages were delayed for ages 4 to 12, and advanced for ages 3 and 13 to 14. By using the BoneXpert automatic software, we established bone age reference standards for children in Taiwan. Clinical application of GP and TW3 scoring methods can be adjusted according to our results to better assess bone age.

Cost-effectiveness of stereotactic vacuum-assisted biopsy for nonpalpable breast lesions.

PURPOSE: To examine the potential cost-savings of stereotactic vacuum-assisted biopsy (SVAB) over open surgical biopsy (OSB) in diagnosis of nonpalpable lesions on mammography and to estimate the cost-saving effect on lesions at different levels of malignant probability. METHODS: This retrospective study was approved by our Institutional Review Board. We retrospectively reviewed 276 (33.8 %) SVAB and 541 (66.2 %) OSB medical records at a medical center. Direct costs included patients' self-paid and national health insurance claim charges. Indirect costs were calculated using sick days, average salary, and age-adjusted employment rate. One-way and two-way sensitivity analyses were conducted. Lesion classification was determined by the assessment categories of Breast Imaging Reporting and Data System (BI-RADS), 4th or 5th editions. RESULTS: SVAB decreased the direct cost by $90.3 (10.1 %) per diagnosis. The indirect cost was decreased by $560.2 (96.0 %). Overall, SVAB saved 43.9 % of resource utilization for each biopsy. Taking the cost of the subsequent malignant surgery into account, from the healthcare providers' perspective, SVAB was cost-effective if a lesion had less than 19 % likelihood of malignancy. From the societal perspective, SVAB reduced productivity loss for all the lesions. Based on the positive predictive value of the BI-RADS categories, SVAB was more suitable for the lesions of category 4A and category 3, resulting in greater savings in both medical and societal resources. CONCLUSIONS: SVAB is a cost-effective diagnostic option for nonpalpable breast lesions. The cost-saving effect is greater for the lesions of category 4A and category 3.

Post-infectious Bronchiolitis Obliterans: HRCT, DECT, Pulmonary Scintigraphy Images, and Clinical Follow-up in Eight Children.

Background: Bronchiolitis obliterans (BO), first mentioned in 1901, is a severe and rare chronic lung disease in children. BO has various etiologies and the most common in children is post-infectious BO (PIBO). High resolution CT (HRCT) is an often-used image tool for the diagnosis of BO, and pulmonary scintigraphy is an alternative tool that can functionally evaluate BO. Recently, dual-energy computed tomography (DECT) have also been applied to BO for its accuracy and safety. Here we described the characteristics of HRCT, pulmonary scintigraphy, DECT, and the clinical profiles of patients with PIBO. Methods: This is a retrospective and descriptive study. Data were collected from patients diagnosed with PIBO from 2014 to 2019 in the Pediatric Cardiopulmonary Outpatient Clinics of Kaohsiung Medical University Hospital. The diagnosis was based on clinical, chest X-ray, and HRCT findings. Clinical profile, radiological characteristics, and images of pulmonary scintigraphy were documented. Results: Eight children (4 boys and 4 girls) were diagnosed with PIBO at a mean age of 25.8 months (range 15 to 41 months). Two of our patients developed pulmonary hypertension. The most common HRCT finding is mosaic pattern, where match ventilation/perfusion (V/Q) defects is a general feature in pulmonary scintigraphy. DECT pulmonary blood vasculature images revealed various degrees of decreased perfusion and is compatible with the decreased perfusion on pulmonary scintigraphy. Conclusion: The therapeutic strategy of PIBO is still lacking of standardization. HRCT and V/Q scans are important image tools in diagnosis and follow-up of BO. DECT may be used in BO patients as it has no additional radiation exposure and add value on functional information of HRCT.

Association between Geriatric Nutrition Risk Index and Skeletal Muscle Mass Index with Bone Mineral Density in Post-Menopausal Women Who Have Undergone Total Thyroidectomy.

BACKGROUND: Osteoporosis is highly prevalent in postmenopausal women and may result in fractures and disabilities. Total thyroidectomy has also been associated with loss of bone mass. The aim of this cross-sectional study was to evaluate associations among nutritional status, skeletal muscle index and markers of bone turnover to bone mineral density in postmenopausal women who had undergone total thyroidectomy. METHODS: Fifty postmenopausal women who had undergone total thyroidectomy were included. Body composition was measured using dual-energy X-ray absorptiometry (DXA). The Geriatric Nutritional Risk Index (GNRI) was calculated using baseline body weight and serum albumin level. Skeletal muscle mass index was calculated as the appendicular skeletal muscle mass (ASM) divided by the height squared and assessed using DXA. RESULTS: Multivariate stepwise linear regression analysis showed that a low GNRI was significantly associated with low lumbar spine bone mineral density (BMD) and T-score, and that a low ASM/height2 was significantly associated with low femoral neck BMD and T-score. A low vitamin D level was significantly associated with low femoral neck BMD and T-score and low total hip BMD and T-score. A high bone alkaline phosphatase (ALP) level was significantly associated with low femoral neck T-score and low total hip BMD and T-score. A low insulin-like growth factor-1 (IGF-1) was significantly associated with low total hip BMD and T-score. CONCLUSION: In the postmenopausal women who had undergone total thyroidectomy in this study, BMD was positively associated with GNRI, skeletal muscle mass index, and levels of vitamin D and serum IGF-1, and inversely associated with bone ALP level. Nutritional status, skeletal muscle mass index and bone turnover biomarkers can be used to early identify patients with a high risk of osteoporosis in this high-risk group.