Small lesion classification on abbreviated breast MRI: training can improve diagnostic performance and inter-reader agreement.

OBJECTIVE: To determine whether the diagnostic performance and inter-reader agreement for small lesion classification on abbreviated breast MRI (AB-MRI) can be improved by training, and can achieve the level of full diagnostic protocol MRI (FDP-MRI). METHODS: This retrospective study enrolled 1165 breast lesions (≤ 2 cm; 409 malignant and 756 benign) from 1165 MRI examinations for reading test. Twelve radiologists were assigned into a trained group and a non-trained group. They interpreted each AB-MRI twice, which was extracted from FDP-MRI. After the first read, the trained group received a structured training for AB-MRI interpretation while the non-trained group did not. FDP-MRIs were interpreted by the trained group after the second read. BI-RADS category for each lesion was compared to the standard of reference (histopathological examination or follow-up) to calculate diagnostic accuracy. Inter-reader agreement was assessed using multirater k analysis. Diagnostic accuracy and inter-reader agreement were compared between the trained and non-trained groups, between the first and second reads, and between AB-MRI and FDP-MRI. RESULTS: After training, the diagnostic accuracy of AB-MRI increased from 77.6 to 84.4%, and inter-reader agreement improved from 0.410 to 0.579 (both p < 0.001), which were higher than those of the non-trained group (accuracy, 84.4% vs 78.0%; weighted k, 0.579 vs 0.461; both p < 0.001). The post-training accuracy and inter-reader agreement of AB-MRI were lower than those of FDP-MRI (accuracy, 84.4% vs 92.8%; weighted k, 0.579 vs 0.602; both p < 0.001). CONCLUSIONS: Training can improve the diagnostic performance and inter-reader agreement for small lesion classification on AB-MRI; however, it remains inferior to those of FDP-MRI. KEY POINTS: • Training can improve the diagnostic performance for small breast lesions on AB-MRI. • Training can reduce inter-observer variation for breast lesion classification on AB-MRI, especially among junior radiologists. • The post-training diagnostic performance and inter-reader agreement of AB-MRI remained inferior to those of FDP-MRI.

Value of digital mammography in predicting lymphovascular invasion of breast cancer.

BACKGROUND: Lymphovascular invasion (LVI) has never been revealed by preoperative scans. It is necessary to use digital mammography in predicting LVI in patients with breast cancer preoperatively. METHODS: Overall 122 cases of invasive ductal carcinoma diagnosed between May 2017 and September 2018 were enrolled and assigned into the LVI positive group (n = 42) and the LVI negative group (n = 80). Independent t-test and χ2 test were performed. RESULTS: Difference in Ki-67 between the two groups was statistically significant (P = 0.012). Differences in interstitial edema (P = 0.013) and skin thickening (P = 0.000) were statistically significant between the two groups. Multiple factor analysis showed that there were three independent risk factors for LVI: interstitial edema (odds ratio [OR] = 12.610; 95% confidence interval [CI]: 1.061-149.922; P = 0.045), blurring of subcutaneous fat (OR = 0.081; 95% CI: 0.012-0.645; P = 0.017) and skin thickening (OR = 9.041; 95% CI: 2.553-32.022; P = 0.001). CONCLUSIONS: Interstitial edema, blurring of subcutaneous fat, and skin thickening are independent risk factors for LVI. The specificity of LVI prediction is as high as 98.8% when the three are used together.

Comparison of the clinical and MRI features of patients with hepatic hemangioma, epithelioid hemangioendothelioma, or angiosarcoma.

BACKGROUND: Comparisons of hepatic epithelioid hemangioendothelioma (HEHE), hepatic hemangioma, and hepatic angiosarcoma (HAS) have rarely been reported. The purpose of our study was to analyze the clinical and magnetic resonance imaging (MRI) findings of these conditions. METHODS: A total of 57 patients (25 with hemangioma, 13 with HEHE, and 19 with HAS) provided hepatic vascular endothelial cell data between June 2006 and May 2017. RESULTS: The proportions of cases with circumscribed margins were 88% (22/25), 84.6% (11/13), and 31.6% (6/19) for hemangioma, HEHE, and HAS, respectively (P < 0.001). HAS lesions were less likely to have circumscribed margins. The proportions of lesions with hemorrhaging were 4% (1/25), 30.8% (4/13), and 36.8% (7/19) for hemangioma, HEHE, and HAS, respectively (P = 0.014). HEHE and HAS cases were more likely to show heterogeneous signals on T1-weighted (T1WI) MRI. HEHE and HAS cases were more likely to show heterogeneous signals on T2-weighted (T2WI) MRI. Centripetal enhancement was the most common pattern in vascular tumors, with proportions of 100, 46.2% (6/13), and 68.4% (13/19) for hemangioma, HEHE, and HAS, respectively. The difference in enhancement pattern between HEHE and HAS was not significant, but rim enhancement was more common for HEHE (46.2%, 6/13). CONCLUSIONS: Our study revealed clinical and imaging differences between HEHE and HAS. The platelet count (PLT) and coagulation function of the HAS group decreased, whereas the alpha-fetoprotein (AFP) level increased. The 5-year survival rate for HAS was significantly lower than that of HEHE. A higher malignancy degree indicated a more blurred lesion margin, easier occurrence of hemorrhaging, and more heterogeneous T1WI and T2WI signals.

Selection of an endogenous 2,3-butanediol pathway in Escherichia coli by fermentative redox balance.

Fermentative redox balance has long been utilized as a metabolic evolution platform to improve efficiency of NADH-dependent pathways. However, such system relies on the complete recycling of NADH and may become limited when the target pathway results in excess NADH stoichiometrically. In this study, endogenous capability of Escherichia coli for 2,3-butanediol (2,3-BD) synthesis was explored using the anaerobic selection platform based on redox balance. To address the issue of NADH excess associated with the 2,3-BD pathway, we devised a substrate-decoupled system where a pathway intermediate is externally supplied in addition to the carbon source to decouple NADH recycling ratio from the intrinsic pathway stoichiometry. In this case, feeding of the 2,3-BD precursor acetoin effectively restored anaerobic growth of the mixed-acid fermentation mutant that remained otherwise inhibited even in the presence of a functional 2,3-BD pathway. Using established 2,3-BD dehydrogenases as model enzyme, we verified that the redox-based selection system is responsive to NADPH-dependent reactions but with lower sensitivity. Based on this substrate-decoupled selection scheme, we successfully identified the glycerol/1,2-propanediol dehydrogenase (Ec-GldA) as the major enzyme responsible for the acetoin reducing activity (kcat/Km≈0.4mM-1s-1) observed in E. coli. Significant shift of 2,3-BD configuration upon withdrawal of the heterologous acetolactate decarboxylase revealed that the endogenous synthesis of acetoin occurs via diacetyl. Among the predicted diacetyl reductase in E. coli, Ec-UcpA displayed the most significant activity towards diacetyl reduction into acetoin (Vmax≈6U/mg). The final strain demonstrated a meso-2,3-BD production titer of 3g/L without introduction of foreign genes. The substrate-decoupled selection system allows redox balance regardless of the pathway stoichiometry thus enables segmented optimization of different reductive pathways through enzyme bioprospecting and metabolic evolution.

Engineering cofactor flexibility enhanced 2,3-butanediol production in Escherichia coli.

Enzymatic reduction of acetoin into 2,3-butanediol (2,3-BD) typically requires the reduced nicotinamide adenine dinucleotide (NADH) or its phosphate form (NADPH) as electron donor. Efficiency of 2,3-BD biosynthesis, therefore, is heavily influenced by the enzyme specificity and the cofactor availability which varies dynamically. This work describes the engineering of cofactor flexibility for 2,3-BD production by simultaneous overexpression of an NADH-dependent 2,3-BD dehydrogenase from Klebsiella pneumoniae (KpBudC) and an NADPH-specific 2,3-BD dehydrogenase from Clostridium beijerinckii (CbAdh). Co-expression of KpBudC and CbAdh not only enabled condition versatility for 2,3-BD synthesis via flexible utilization of cofactors, but also improved production stereo-specificity of 2,3-BD without accumulation of acetoin. With optimization of medium and fermentation condition, the co-expression strain produced 92 g/L of 2,3-BD in 56 h with 90% stereo-purity for (R,R)-isoform and 85% of maximum theoretical yield. Incorporating cofactor flexibility into the design principle should benefit production of bio-based chemical involving redox reactions.