In radiomics studies, researchers usually need to develop a supervised machine learning model to map image features onto the clinical conclusion. A classical machine learning pipeline consists of several steps, including normalization, feature selection, and classification. It is often tedious to find an optimal pipeline with appropriate combinations. We designed an open-source software package named FeAture Explorer (FAE). It was programmed with Python and used NumPy, pandas, and scikit-learning modules. FAE can be used to extract image features, preprocess the feature matrix, develop different models automatically, and evaluate them with common clinical statistics. FAE features a user-friendly graphical user interface that can be used by radiologists and researchers to build many different pipelines, and to compare their results visually. To prove the effectiveness of FAE, we developed a candidate model to classify the clinical-significant prostate cancer (CS PCa) and non-CS PCa using the PROSTATEx dataset. We used FAE to try out different combinations of feature selectors and classifiers, compare the area under the receiver operating characteristic curve of different models on the validation dataset, and evaluate the model using independent test data. The final model with the analysis of variance as the feature selector and linear discriminate analysis as the classifier was selected and evaluated conveniently by FAE. The area under the receiver operating characteristic curve on the training, validation, and test dataset achieved results of 0.838, 0.814, and 0.824, respectively. FAE allows researchers to build radiomics models and evaluate them using an independent testing dataset. It also provides easy model comparison and result visualization. We believe FAE can be a convenient tool for radiomics studies and other medical studies involving supervised machine learning.
Reducing interface nonradiative recombination is important for realizing highly efficient perovskite solar cells. In this work, we develop a synergistic bimolecular interlayer (SBI) strategy via 4-methoxyphenylphosphonic acid (MPA) and 2-phenylethylammonium iodide (PEAI) to functionalize the perovskite interface. MPA induces an in-situ chemical reaction at the perovskite surface via forming strong P-O-Pb covalent bonds that diminish the surface defect density and upshift the surface Fermi level. PEAI further creates an additional negative surface dipole so that a more n-type perovskite surface is constructed, which enhances electron extraction at the top interface. With this cooperative surface treatment, we greatly minimize interface nonradiative recombination through both enhanced defect passivation and improved energetics. The resulting p-i-n device achieves a stabilized power conversion efficiency of 25.53% and one of the smallest nonradiative recombination induced Voc loss of only 59 mV reported to date. We also obtain a certified efficiency of 25.05%. This work sheds light on the synergistic interface engineering for further improvement of perovskite solar cells.
8 March 2019 in Proceedings of the National Academy of Sciences
For nearly 100 y, homochiral ferroelectrics were basically multicomponent simple organic amine salts and metal coordination compounds. Single-component homochiral organic ferroelectric crystals with high-Curie temperature ( Tc ) phase transition were very rarely reported, although the first ferroelectric Rochelle salt discovered in 1920 is a homochiral metal coordination compound. Here, we report a pair of single-component organic enantiomorphic ferroelectrics, ( R )-3-quinuclidinol and ( S )-3-quinuclidinol, as well as the racemic mixture ( Rac )-3-quinuclidinol. The homochiral ( R )- and ( S )-3-quinuclidinol crystallize in the enantiomorphic-polar point group 6 ( C6 ) at room temperature, showing mirror-image relationships in vibrational circular dichroism spectra and crystal structure. Both enantiomers exhibit 622 F 6-type ferroelectric phase transition with as high as 400 K [above that of BaTiO 3 ( Tc = 381 K)], showing very similar ferroelectricity and related properties, including sharp step-like dielectric anomaly from 5 to 17, high saturation polarization (7 μC/cm 2 ), low coercive field (15 kV/cm), and identical ferroelectric domains. Their racemic mixture ( Rac )-3-quinuclidinol, however, adopts a centrosymmetric point group 2/ m ( C2h ), undergoing a nonferroelectric high-temperature phase transition. This finding reveals the enormous benefits of homochirality in designing high- Tc ferroelectrics, and sheds light on exploring homochiral ferroelectrics with great application.