We demonstrate the possibility of a domain-adaptation when only a limited amount of MR data is available. The proposed method surpasses the existing compressed sensing algorithms in terms of the image quality and computation time.
Background Mammography is the current standard for breast cancer screening. This study aimed to develop an artificial intelligence (AI) algorithm for diagnosis of breast cancer in mammography, and explore whether it could benefit radiologists by improving accuracy of diagnosis. Methods In this retrospective study, an AI algorithm was developed and validated with 170 230 mammography examinations collected from five institutions in South Korea, the USA, and the UK, including 36 468 cancer positive confirmed by biopsy, 59 544 benign confirmed by biopsy (8827 mammograms) or follow-up imaging (50 717 mammograms), and 74 218 normal. For the multicentre, observer-blinded, reader study, 320 mammograms (160 cancer positive, 64 benign, 96 normal) were independently obtained from two institutions. 14 radiologists participated as readers and assessed each mammogram in terms of likelihood of malignancy (LOM), location of malignancy, and necessity to recall the patient, first without and then with assistance of the AI algorithm. The performance of AI and radiologists was evaluated in terms of LOM-based area under the receiver operating characteristic curve (AUROC) and recall-based sensitivity and specificity. Findings The AI standalone performance was AUROC 0•959 (95% CI 0•952-0•966) overall, and 0•970 (0•963-0•978) in the South Korea dataset, 0•953 (0•938-0•968) in the USA dataset, and 0•938 (0•918-0•958) in the UK dataset. In the reader study, the performance level of AI was 0•940 (0•915-0•965), significantly higher than that of the radiologists without AI assistance (0•810, 95% CI 0•770-0•850; p<0•0001). With the assistance of AI, radiologists' performance was improved to 0•881 (0•850-0•911; p<0•0001). AI was more sensitive to detect cancers with mass (53 [90%] vs 46 [78%] of 59 cancers detected; p=0•044) or distortion or asymmetry (18 [90%] vs ten [50%] of 20 cancers detected; p=0•023) than radiologists. AI was better in detection of T1 cancers (73 [91%] vs 59 [74%] of 80; p=0•0039) or node-negative cancers (104 [87%] vs 88 [74%] of 119; p=0•0025) than radiologists. Interpretation The AI algorithm developed with large-scale mammography data showed better diagnostic performance in breast cancer detection compared with radiologists. The significant improvement in radiologists' performance when aided by AI supports application of AI to mammograms as a diagnostic support tool.
Increasing evidence points to the importance of local protein synthesis for axonal growth and responses to axotomy, yet there is little insight into the functions of individual locally synthesized proteins. We recently showed that expression of a reporter mRNA with the axonally localizing β-actin mRNA 3′UTR competes with endogenous β-actin and GAP-43 mRNAs for binding to ZBP1 and axonal localization in adult sensory neurons (Donnelly et al., 2011). Here, we show that the 3′UTR of GAP-43 mRNA can deplete axons of endogenous β-actin mRNA. We took advantage of this 3′UTR competition to address the functions of axonally synthesized β-actin and GAP-43 proteins. In cultured rat neurons, increasing axonal synthesis of β-actin protein while decreasing axonal synthesis of GAP-43 protein resulted in short highly branched axons. Decreasing axonal synthesis of β-actin protein while increasing axonal synthesis of GAP-43 protein resulted in long axons with few branches. siRNA-mediated depletion of overall GAP-43 mRNA from dorsal root ganglia (DRGs) decreased the length of axons, while overall depletion of β-actin mRNA from DRGs decreased the number of axon branches. These deficits in axon growth could be rescued by transfecting with siRNA-resistant constructs encoding β-actin or GAP-43 proteins, but only if the mRNAs were targeted for axonal transport. Finally, in ovo electroporation of axonally targeted GAP-43 mRNA increased length and axonally targeted β-actin mRNA increased branching of sensory axons growing into the chick spinal cord. These studies indicate that axonal translation of β-actin mRNA supports axon branching and axonal translation of GAP-43 mRNA supports elongating growth.
Although definite radiologic differentiation from malignancy is not clearly possible, we suggest that familiarity with the manifestations of inflammatory pseudotumor can help avoid unnecessary radical surgery before histopathologic proof of malignancy is obtained.
Localized translation of axonal mRNAs contributes to developmental and regenerative axon growth. Although untranslated regions (UTRs) of many different axonal mRNAs appear to drive their localization, there has been no consensus RNA structure responsible for this localization. We recently showed that limited expression of ZBP1 protein restricts axonal localization of both β-actin and GAP-43 mRNAs. β-actin 3′UTR has a defined element for interaction with ZBP1, but GAP-43 mRNA shows no homology to this RNA sequence. Here, we show that an AU-rich element (ARE) in GAP-43’s 3′UTR is necessary and sufficient for its axonal localization. Axonal GAP-43 mRNA levels increase after in vivo injury, and GAP-43 mRNA shows an increased half-life in regenerating axons. GAP-43 mRNA interacts with both HuD and ZBP1, and HuD and ZBP1 coimmunoprecipitate in an RNA-dependent fashion. Reporter mRNA with the GAP-43 ARE competes with endogenous β-actin mRNA for axonal localization and decreases axon length and branching similar to the β-actin 3′UTR competing with endogenous GAP-43 mRNA. Conversely, overexpressing GAP-43 coding sequence with it’s 3′UTR ARE increases axonal elongation and this effect is lost when just the ARE is deleted from GAP-43’s 3′UTR.
After breast conservation therapy in women 50 years or younger, the addition of MRI to annual mammography screening improves detection of early-stage but biologically aggressive breast cancers at acceptable specificity. Results from this study can inform patient decision making on screening methods after breast conservation therapy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.