METHODS: Positron emission tomography (PET) and computed tomography (CT) image data from 97 patients with LC and 77 patients with TB nodules were collected. One hundred radiomic features were extracted from both PET and CT imaging using the pyradiomics platform, and 2048 deep learning features were obtained through a residual neural network approach. Four models included traditional machine learning model with radiomic features as input (traditional radiomics), a deep learning model with separate input of image features (deep convolutional neural networks [DCNN]), a deep learning model with two inputs of radiomic features and deep learning features (radiomics-DCNN) and a deep learning model with inputs of radiomic features and deep learning features and clinical information (integrated model). The models were evaluated using area under the curve (AUC), sensitivity, accuracy, specificity, and F1-score metrics.
RESULTS: The results of the classification of TB nodules and LC showed that the integrated model achieved an AUC of 0.84 (0.82-0.88), sensitivity of 0.85 (0.80-0.88), and specificity of 0.84 (0.83-0.87), performing better than the other models.
CONCLUSION: The integrated model was found to be the best classification model in the diagnosis of TB nodules and solid LC.
METHODS: For this systematic review and meta-analysis, we searched PubMed, Embase, Scopus, and the Cochrane Library from inception to May 1, 2019, for relevant original research articles without any language restrictions. The literature search and data extraction were done independently by two investigators. Primary outcomes were the prevalence of non-obese or lean people within the NAFLD group and the prevalence of non-obese or lean NAFLD in the general, non-obese, and lean populations; the incidence of NAFLD among non-obese and lean populations; and long-term outcomes of non-obese people with NAFLD. We also aimed to characterise the demographic, clinical, and histological characteristics of individuals with non-obese NAFLD.
FINDINGS: We identified 93 studies (n=10 576 383) from 24 countries or areas: 84 studies (n=10 530 308) were used for the prevalence analysis, five (n=9121) were used for the incidence analysis, and eight (n=36 954) were used for the outcomes analysis. Within the NAFLD population, 19·2% (95% CI 15·9-23·0) of people were lean and 40·8% (36·6-45·1) were non-obese. The prevalence of non-obese NAFLD in the general population varied from 25% or lower in some countries (eg, Malaysia and Pakistan) to higher than 50% in others (eg, Austria, Mexico, and Sweden). In the general population (comprising individuals with and without NAFLD), 12·1% (95% CI 9·3-15·6) of people had non-obese NAFLD and 5·1% (3·7-7·0) had lean NAFLD. The incidence of NAFLD in the non-obese population (without NAFLD at baseline) was 24·6 (95% CI 13·4-39·2) per 1000 person-years. Among people with non-obese or lean NALFD, 39·0% (95% CI 24·1-56·3) had non-alcoholic steatohepatitis, 29·2% (21·9-37·9) had significant fibrosis (stage ≥2), and 3·2% (1·5-5·7) had cirrhosis. Among the non-obese or lean NAFLD population, the incidence of all-cause mortality was 12·1 (95% CI 0·5-38·8) per 1000 person-years, that for liver-related mortality was 4·1 (1·9-7·1) per 1000 person-years, cardiovascular-related mortality was 4·0 (0·1-14·9) per 1000 person-years, new-onset diabetes was 12·6 (8·0-18·3) per 1000 person-years, new-onset cardiovascular disease was 18·7 (9·2-31·2) per 1000 person-years, and new-onset hypertension was 56·1 (38·5-77·0) per 1000 person-years. Most analyses were characterised by high heterogeneity.
INTERPRETATION: Overall, around 40% of the global NAFLD population was classified as non-obese and almost a fifth was lean. Both non-obese and lean groups had substantial long-term liver and non-liver comorbidities. These findings suggest that obesity should not be the sole criterion for NAFLD screening. Moreover, clinical trials of treatments for NAFLD should include participants across all body-mass index ranges.
FUNDING: None.
MATERIALS AND METHODS: From a PSA screening initiative, 161 men were shown to have elevated PSA levels in their blood and underwent prostatic tissue biopsy. DNA was extracted from the blood, and exon 1 of the AR gene amplified by PCR and sequenced. The number of CAG repeat sequences were counted and compared to the immunohistochemical expression of ERG and AR in the matched tumour biopsies.
RESULTS: Of men with elevated PSA, 89 were diagnosed with prostate cancer, and 72 with benign prostatic hyperplasia (BPH). There was no significant difference in the length of the CAG repeat in men with prostate cancer and BPH. The CAG repeat length was not associated with; age, PSA or tumour grade, though a longer CAG repeat was associated with tumour stage. ERG and AR were expressed in 36% and 86% of the cancers, respectively. There was no significant association between CAG repeat length and ERG or AR expression. However, there was a significant inverse relationship between ERG and AR expression. In addition, a significantly great proportion of Indian men had ERG positive tumours, compared to men of Malay or Chinese descent.
CONCLUSIONS: CAG repeat length is not associated with prostate cancer or expression of ERG or AR. However, ERG appears to be more common in the prostate cancers of Malaysian Indian men than in the prostate cancers of other Malaysian ethnicities and its expression in this study was inversely related to AR expression.