The effect of an artificial intelligence algorithm on chest X-ray interpretation of radiology residents.

OBJECTIVE: Chest X-rays are the most commonly performed diagnostic examinations. An artificial intelligence (AI) system that evaluates the images fast and accurately help reducing workflow and management of the patients. An automated assistant may reduce the time of interpretation in daily practice. We aim to investigate whether radiology residents consider the recommendations of an AI system for their final decisions, and to assess the diagnostic performances of the residents and the AI system. METHODS: Posteroanterior (PA) chest X-rays with confirmed diagnosis were evaluated by 10 radiology residents. After interpretation, the residents checked the evaluations of the AI Algorithm and made their final decisions. Diagnostic performances of the residents without AI and after checking the AI results were compared. RESULTS: Residents' diagnostic performance for all radiological findings had a mean sensitivity of 37.9% (vs 39.8% with AI support), a mean specificity of 93.9% (vs 93.9% with AI support). The residents obtained a mean AUC of 0.660 vs 0.669 with AI support. The AI algorithm diagnostic accuracy, measured by the overall mean AUC, was 0.789. No significant difference was detected between decisions taken with and without the support of AI. CONCLUSION: Although, the AI algorithm diagnostic accuracy were higher than the residents, the radiology residents did not change their final decisions after reviewing AI recommendations. In order to benefit from these tools, the recommendations of the AI system must be more precise to the user. ADVANCES IN KNOWLEDGE: This research provides information about the willingness or resistance of radiologists to work with AI technologies via diagnostic performance tests. It also shows the diagnostic performance of an existing AI algorithm, determined by real-life data.

Opportunities and Challenges for Drug Development: Public-Private Partnerships, Adaptive Designs and Big Data.

Drug development faces the double challenge of increasing costs and increasing pressure on pricing. To avoid that lack of perceived commercial perspective will leave existing medical needs unmet, pharmaceutical companies and many other stakeholders are discussing ways to improve the efficiency of drug Research and Development. Based on an international symposium organized by the Medical School of the University of Duisburg-Essen (Germany) and held in January 2016, we discuss the opportunities and challenges of three specific areas, i.e., public-private partnerships, adaptive designs and big data. Public-private partnerships come in many different forms with regard to scope, duration and type and number of participants. They range from project-specific collaborations to strategic alliances to large multi-party consortia. Each of them offers specific opportunities and faces distinct challenges. Among types of collaboration, investigator-initiated studies are becoming increasingly popular but have legal, ethical, and financial implications. Adaptive trial designs are also increasingly discussed. However, adaptive should not be used as euphemism for the repurposing of a failed trial; rather it requires carefully planning and specification before a trial starts. Adaptive licensing can be a counter-part of adaptive trial design. The use of Big Data is another opportunity to leverage existing information into knowledge useable for drug discovery and development. Respecting limitations of informed consent and privacy is a key challenge in the use of Big Data. Speakers and participants at the symposium were convinced that appropriate use of the above new options may indeed help to increase the efficiency of future drug development.

Electrochemical stability of self-assembled alkylphosphate monolayers on conducting metal oxides.

Alkylphosphate self-assembled monolayers (SAMs) were prepared on Nb-doped SrTiO(3) (Nb-STO) conducting metal oxide substrates. Unlike thiols on gold, the alkylphosphate SAMs on Nb-STO exhibited an electrochemical stability over a wide voltage range from -2 to 2 V. Cyclic voltammetry showed that the SAM modification inhibited the electrochemical activity of the underlying conducting substrate with an efficiency dependent on the chain length. Impedance spectroscopy showed that SAM-modified Nb-STO substrates have a significantly higher resistance than bare substrates.

Monolayer-directed assembly and magnetic properties of FePt nanoparticles on patterned aluminum oxide.

FePt nanoparticles (NPs) were assembled on aluminum oxide substrates, and their ferromagnetic properties were studied before and after thermal annealing. For the first time, phosph(on)ates were used as an adsorbate to form self-assembled monolayers (SAMs) on alumina to direct the assembly of NPs onto the surface. The Al(2)O(3) substrates were functionalized with aminobutylphosphonic acid (ABP) or phosphonoundecanoic acid (PNDA) SAMs or with poly(ethyleneimine) (PEI) as a reference. FePt NPs assembled on all of these monolayers, but much less on unmodified Al(2)O(3), which shows that ligand exchange at the NPs is the most likely mechanism of attachment. Proper modification of the Al(2)O(3) surface and controlling the immersion time of the modified Al(2)O(3) substrates into the FePt NP solution resulted in FePt NPs assembly with controlled NP density. Alumina substrates were patterned by microcontact printing using aminobutylphosphonic acid as the ink, allowing local NP assembly. Thermal annealing under reducing conditions (96%N(2)/4%H(2)) led to a phase change of the FePt NPs from the disordered FCC phase to the ordered FCT phase. This resulted in ferromagnetic behavior at room temperature. Such a process can potentially be applied in the fabrication of spintronic devices.