Search

Abstract Large language models (LLMs), such as ChatGPT, have demonstrated impressive capabilities in various tasks and attracted increasing interest as a natural language interface across many domains. Recently, large vision-language models (VLMs) that learn rich vision–language correlation from image–text pairs, like BLIP-2 and GPT-4, have been intensively investigated. However, despite these developments, the application of LLMs and VLMs in image quality assessment (IQA), particularly in medical imaging, remains unexplored. This is valuable for objective performance evaluation and potential supplement or even replacement of radiologists’ opinions. To this end, this study introduces IQAGPT, an innovative computed tomography (CT) IQA system that integrates image-quality captioning VLM with ChatGPT to generate quality scores and textual reports. First, a CT-IQA dataset comprising 1,000 CT slices with diverse quality levels is professionally annotated and compiled for training and evaluation. To better leverage the capabilities of LLMs, the annotated quality scores are converted into semantically rich text descriptions using a prompt template. Second, the image-quality captioning VLM is fine-tuned on the CT-IQA dataset to generate quality descriptions. The captioning model fuses image and text features through cross-modal attention. Third, based on the quality descriptions, users verbally request ChatGPT to rate image-quality scores or produce radiological quality reports. Results demonstrate the feasibility of assessing image quality using LLMs. The proposed IQAGPT outperformed GPT-4 and CLIP-IQA, as well as multitask classification and regression models that solely rely on images.

Abstract Flipover, an enhanced dropout technique, is introduced to improve the robustness of artificial neural networks. In contrast to dropout, which involves randomly removing certain neurons and their connections, flipover randomly selects neurons and reverts their outputs using a negative multiplier during training. This approach offers stronger regularization than conventional dropout, refining model performance by (1) mitigating overfitting, matching or even exceeding the efficacy of dropout; (2) amplifying robustness to noise; and (3) enhancing resilience against adversarial attacks. Extensive experiments across various neural networks affirm the effectiveness of flipover in deep learning.

Abstract [6,6]-Phenyl-C61-butyric acid methyl ester (PCBM), a star molecule in the fullerene field, has found wide applications in materials science. Herein, electrosynthesis of buckyballs with fused-ring systems has been achieved through radical α-C−H functionalization of the side-chain ester for both PCBM and its analogue, [6,6]-phenyl-C61-propionic acid methyl ester (PCPM), in the presence of a trace amount of oxygen. Two classes of buckyballs with fused bi- and tricyclic carbocycles have been electrochemically synthesized. Furthermore, an unknown type of a bisfulleroid with two tethered [6,6]-open orifices can also be efficiently generated from PCPM. All three types of products have been confirmed by single-crystal X-ray crystallography. A representative intramolecularly annulated isomer of PCBM has been applied as an additive to inverted planar perovskite solar cells and boosted a significant enhancement of power conversion efficiency from 15.83% to 17.67%.

Abstract The large language model called ChatGPT has drawn extensively attention because of its human-like expression and reasoning abilities. In this study, we investigate the feasibility of using ChatGPT in experiments on translating radiology reports into plain language for patients and healthcare providers so that they are educated for improved healthcare. Radiology reports from 62 low-dose chest computed tomography lung cancer screening scans and 76 brain magnetic resonance imaging metastases screening scans were collected in the first half of February for this study. According to the evaluation by radiologists, ChatGPT can successfully translate radiology reports into plain language with an average score of 4.27 in the five-point system with 0.08 places of information missing and 0.07 places of misinformation. In terms of the suggestions provided by ChatGPT, they are generally relevant such as keeping following-up with doctors and closely monitoring any symptoms, and for about 37% of 138 cases in total ChatGPT offers specific suggestions based on findings in the report. ChatGPT also presents some randomness in its responses with occasionally over-simplified or neglected information, which can be mitigated using a more detailed prompt. Furthermore, ChatGPT results are compared with a newly released large model GPT-4, showing that GPT-4 can significantly improve the quality of translated reports. Our results show that it is feasible to utilize large language models in clinical education, and further efforts are needed to address limitations and maximize their potential.

Extracting objects of interest from remote sensing imagery is an essential part in various practical applications. The objects that people pay attention to in the remote sensing scene mainly include buildings, roads, vehicles, etc. In this article, extracting the aforementioned objects are collectively referred to as the target extraction task. Arising from object scale variation, appearance similarity between adjacent patches, diversity of imaging orientation, and complexity of background, it is difficult to extract complete objects from cluttered backgrounds. Deep neural network has made great achievement in dense prediction for target extraction. However, most of the previous works are still faced with a formidable challenge in discriminative context feature representation to extract targets of various categories and correctly classify pixels around the boundary. In this article, we propose a target extraction neural network, named discriminative context-aware network, to focus on discriminative high-level context features and preserve spatial information. First, a discriminative context-aware feature module is designed to generate the feature maps in the top layer, which not only captures the rich image context information but also aggregates the contrasted local information at multiple scales. Second, a refine decoder module is adopted to preserve spatial information from low-level layers and enhance the feature representation, leading to precise segmentation results. We conducted extensive experiments on building and road extraction benchmarks, including WHU building dataset and Massachusetts road dataset, together with a self-constructed dataset for vehicle extraction in SAR images. Our method achieves state-of-the-art results with fewer parameters and faster inference.

Abstract The emergence of Y6‐type nonfullerene acceptors has greatly enhanced the power conversion efficiency (PCE) of organic solar cells (OSCs). However, which structural feature is responsible for the excellent photovoltaic performance is still under debate. In this study, two Y6‐like acceptors BDOTP‐1 and BDOTP‐2 were designed. Different from previous Y6‐type acceptors featuring an A–D–Aʹ–D–A structure, BDOTP‐1, and BDOTP‐2 have no electron‐deficient Aʹ fragment in the core unit. Instead, there is an electron‐rich dibenzodioxine fragment in the core. Although this modification leads to a marked change in the molecular dipole moment, electrostatic potential, frontier orbitals, and energy levels, BDOTP acceptors retain similar three‐dimensional packing capability as Y6‐type acceptors due to the similar banana‐shaped molecular configuration. BDOTP acceptors show good performance in OSCs. High PCEs of up to 18.51% (certified 17.9%) are achieved. This study suggests that the banana‐shaped configuration instead of the A–D–Aʹ–D–A structure is likely to be the determining factor in realizing high photovoltaic performance.

Summary: Due to lack of the kernel awareness, some popular deep image reconstruction networks are unstable. To address this problem, here we introduce the bounded relative error norm (BREN) property, which is a special case of the Lipschitz continuity. Then, we perform a convergence study consisting of two parts: (1) a heuristic analysis on the convergence of the analytic compressed iterative deep (ACID) scheme (with the simplification that the CS module achieves a perfect sparsification), and (2) a mathematically denser analysis (with the two approximations: [1] AT is viewed as an inverse A-1 in the perspective of an iterative reconstruction procedure and [2] a pseudo-inverse is used for a total variation operator H). Also, we present adversarial attack algorithms to perturb the selected reconstruction networks respectively and, more importantly, to attack the ACID workflow as a whole. Finally, we show the numerical convergence of the ACID iteration in terms of the Lipschitz constant and the local stability against noise. The bigger picture: For deep tomographic reconstruction to realize its full potential in practice, it is critically important to address the instabilities of deep reconstruction networks, which were identified in a recent PNAS paper. Our analytic compressed iterative deep (ACID) framework has provided an effective solution to address this challenge by synergizing deep learning and compressed sensing through iterative refinement. Here, we provide an initial convergence analysis, describe an algorithm to attack the entire ACID workflow, and establish not only its capability of stabilizing an unstable deep reconstruction network but also its stability against adversarial attacks dedicated to ACID as a whole. Although our theoretical results are under approximations, they shed light on the converging mechanism of ACID, serving as a basis for further investigation.

Summary: A recent PNAS paper reveals that several popular deep reconstruction networks are unstable. Specifically, three kinds of instabilities were reported: (1) strong image artefacts from tiny perturbations, (2) small features missed in a deeply reconstructed image, and (3) decreased imaging performance with increased input data. Here, we propose an analytic compressed iterative deep (ACID) framework to address this challenge. ACID synergizes a deep network trained on big data, kernel awareness from compressed sensing (CS)-inspired processing, and iterative refinement to minimize the data residual relative to real measurement. Our study demonstrates that the ACID reconstruction is accurate, is stable, and sheds light on the converging mechanism of the ACID iteration under a bounded relative error norm assumption. ACID not only stabilizes an unstable deep reconstruction network but also is resilient against adversarial attacks to the whole ACID workflow, being superior to classic sparsity-regularized reconstruction and eliminating the three kinds of instabilities. The bigger picture: Tomographic image reconstruction with deep learning has been a rapidly emerging field since 2016. Recently, a PNAS paper revealed that several well-known deep reconstruction networks are unstable for computed tomography (CT) and magnetic resonance imaging (MRI), and, in contrast, compressed sensing (CS)-inspired reconstruction methods are stable because of their theoretically proven property known as “kernel awareness.” Therefore, for deep reconstruction to realize its full potential and become a mainstream approach for tomographic imaging, it is critically important to stabilize deep reconstruction networks. Here, we propose an analytic compressed iterative deep (ACID) framework to synergize deep learning and compressed sensing through iterative refinement. We anticipate that this integrative model-based data-driven approach will promote the development and translation of deep tomographic image reconstruction networks.

While micro-CT systems are instrumental in preclinical research, clinical micro-CT imaging has long been desired with cochlear implantation as a primary application. The structural details of the cochlear implant and the temporal bone require a significantly higher image resolution than that (about 0.2 mm ) provided by current medical CT scanners. In this paper, we propose a clinical micro-CT (CMCT) system design integrating conventional spiral cone-beam CT, contemporary interior tomography, deep learning techniques, and the technologies of a micro-focus X-ray source, a photon-counting detector (PCD), and robotic arms for ultrahigh-resolution localized tomography of a freely-selected volume of interest (VOI) at a minimized radiation dose level. The whole system consists of a standard CT scanner for a clinical CT exam and VOI specification, and a robotic micro-CT scanner for a local scan of high spatial and spectral resolution at minimized radiation dose. The prior information from the global scan is also fully utilized for background compensation of the local scan data for accurate and stable VOI reconstruction. Our results and analysis show that the proposed hybrid reconstruction algorithm delivers accurate high-resolution local reconstruction, and is insensitive to the misalignment of the isocenter position, initial view angle and scale mismatch in the data/image registration. These findings demonstrate the feasibility of our system design. We envision that deep learning techniques can be leveraged for optimized imaging performance. With high-resolution imaging, high dose efficiency and low system cost synergistically, our proposed CMCT system has great promise in temporal bone imaging as well as various other clinical applications.

Benzylation of the electrochemically generated dianion from N-p-tolyl-[60]fullerooxazolidinone with benzyl bromide provides three products with different addition patterns. The product distribution can be dramatically altered by varying the reaction conditions. Based on spectral characterizations, these products have been assigned as mono-benzylated 1,4-adduct and bis-benzylated 1,2,3,16- and 1,4,9,25-adducts, respectively. The assigned 1,2,3,16-adduct has been further established by X-ray diffraction analysis. It is believed that the 1,4-adduct is obtained by decarboxylative benzylation of the dianionic species, while bis-benzylated 1,2,3,16- and 1,4,9,25-adducts are achieved via a rearrangement process. In addition, the electrochemical properties of these products have been studied.

Solvent-free mechanical milling is a new, environmentally friendly and cost-effective technology that is now widely used in the field of organic synthesis. The mechanochemical solvent-free synthesis of furoxans from aldoximes was achieved through dimerization of the in situ generated nitrile oxides in the presence of sodium chloride, Oxone and a base. A variety of furoxans was obtained with up to a 92% yield. The present protocol has the advantages of high reaction efficiency and mild reaction conditions.

With the recent advance in chemical modification of fullerenes, electrosynthesis has demonstrated increasing importance in regioselective synthesis of novel fullerene derivatives. Herein, we report successively regioselective synthesis of stable tetra- and hexafunctionalized [60]fullerene derivatives. The cycloaddition reaction of the electrochemically generated dianions from [60]fulleroindolines with phthaloyl chloride regioselectively affords 1,2,4,17-functionalized [60]fullerene derivatives with two attached ketone groups and a unique addition pattern, where the heterocycle is rearranged to a [5,6]-junction and the carbocycle is fused to an adjacent [6,6]-junction. This addition pattern is in sharp contrast with that of the previously reported biscycloadducts, where both cycles are appended to [6,6]-junctions. The obtained tetrafunctionalized compounds can be successively manipulated to 1,2,3,4,9,10-functionalized [60]fullerene derivatives with an intriguing “S”-shaped configuration via a novel electrochemical protonation. Importantly, the stability of tetrafunctionalized [60]fullerene products allows them to be applied in planar perovskite solar cells as efficient electron transport layers.

The solvent-free mechanochemical reaction has aroused increasing interest among scientists. Mechanical ball-milling can implement reactions under mild conditions, shorten the reaction time, and improve the reaction efficiency. Particularly, the most attractive characteristic of mechanochemistry is that it can alter the reaction pathway. However, few such examples have been reported so far. In this paper, we report the reaction of aldoximes with NaCl and Oxone under ball-milling conditions to afford N-acyloxyimidoyl chlorides, which are different from those of the liquid-phase counterpart.