RESUMEN

Objective.Automated biopsy needle segmentation in 3D ultrasound images can be used for biopsy navigation, but it is quite challenging due to the low ultrasound image resolution and interference similar to the needle appearance. For 3D medical image segmentation, such deep learning networks as convolutional neural network and transformer have been investigated. However, these segmentation methods require numerous labeled data for training, have difficulty in meeting the real-time segmentation requirement and involve high memory consumption.Approach.In this paper, we have proposed the temporal information-based semi-supervised training framework for fast and accurate needle segmentation. Firstly, a novel circle transformer module based on the static and dynamic features has been designed after the encoders for extracting and fusing the temporal information. Then, the consistency constraints of the outputs before and after combining temporal information are proposed to provide the semi-supervision for the unlabeled volume. Finally, the model is trained using the loss function which combines the cross-entropy and Dice similarity coefficient (DSC) based segmentation loss with mean square error based consistency loss. The trained model with the single ultrasound volume input is applied to realize the needle segmentation in ultrasound volume.Main results.Experimental results on three needle ultrasound datasets acquired during the beagle biopsy show that our approach is superior to the most competitive mainstream temporal segmentation model and semi-supervised method by providing higher DSC (77.1% versus 76.5%), smaller needle tip position (1.28 mm versus 1.87 mm) and length (1.78 mm versus 2.19 mm) errors on the kidney dataset as well as DSC (78.5% versus 76.9%), needle tip position (0.86 mm versus 1.12 mm) and length (1.01 mm versus 1.26 mm) errors on the prostate dataset.Significance.The proposed method can significantly enhance needle segmentation accuracy by training with sequential images at no additional cost. This enhancement may further improve the effectiveness of biopsy navigation systems.

Asunto(s)

RESUMEN

Robot-assisted prostate biopsy is a new technology to diagnose prostate cancer, but its safety is influenced by the inability of robots to sense the tool-tissue interaction force accurately during biopsy. Recently, vision based force sensing (VFS) provides a potential solution to this issue by utilizing image sequences to infer the interaction force. However, the existing mainstream VFS methods cannot realize the accurate force sensing due to the adoption of convolutional or recurrent neural network to learn deformation from the optical images and some of these methods are not efficient especially when the recurrent convolutional operations are involved. This paper has presented a Transformer based VFS (TransVFS) method by leveraging ultrasound volume sequences acquired during prostate biopsy. The TransVFS method uses a spatio-temporal local-global Transformer to capture the local image details and the global dependency simultaneously to learn prostate deformations for force estimation. Distinctively, our method explores both the spatial and temporal attention mechanisms for image feature learning, thereby addressing the influence of the low ultrasound image resolution and the unclear prostate boundary on the accurate force estimation. Meanwhile, the two efficient local-global attention modules are introduced to reduce 4D spatio-temporal computation burden by utilizing the factorized spatio-temporal processing strategy, thereby facilitating the fast force estimation. Experiments on prostate phantom and beagle dogs show that our method significantly outperforms existing VFS methods and other spatio-temporal Transformer models. The TransVFS method surpasses the most competitive compared method ResNet3dGRU by providing the mean absolute errors of force estimation, i.e., 70.4 ± 60.0 millinewton (mN) vs 123.7 ± 95.6 mN, on the transabdominal ultrasound dataset of dogs.

Asunto(s)

RESUMEN

BACKGROUND: Robotic systems are increasingly used to enhance clinical outcomes in prostate intervention. To evaluate the clinical value of the proposed portable robot, the robot-assisted and robot-targeted punctures were validated experimentally. METHOD: The robot registration utilising the electromagnetic tracker achieves coordinate transformation from the ultrasound (US) image to the robot. Subsequently, Transrectal ultrasound (TRUS)-guided phantom trials were conducted for robot-assisted, free-hand, and robot-targeted punctures. RESULTS: The accuracy of robot registration was 0.95 mm, and the accuracy of robot-assisted, free-hand, and robot-targeted punctures was 2.38 ± 0.64 mm, 3.11 ± 0.72 mm, and 3.29 ± 0.83 mm sequentially. CONCLUSION: The registration method has been successfully applied to robot-targeted puncture. Current results indicate that the accuracy of robot-targeted puncture is slightly inferior to that of manual operations. Moreover, in manual operation, robot-assisted puncture improves the accuracy of free-hand puncture. Accuracy superior to 3.5 mm demonstrates the clinical applicability of both robot-assisted and robot-targeted punctures.

RESUMEN

Optical coherence tomography (OCT) has found wide application to the diagnosis of ophthalmic diseases, but the quality of OCT images is degraded by speckle noise. The convolutional neural network (CNN) based methods have attracted much attention in OCT image despeckling. However, these methods generally need noisy-clean image pairs for training and they are difficult to capture the global context information effectively. To address these issues, we have proposed a novel unsupervised despeckling method. This method uses the cross-scale CNN to extract the local features and uses the intra-patch and inter-patch based transformer to extract and merge the local and global feature information. Based on these extracted features, a reconstruction network is used to produce the final denoised result. The proposed network is trained using a hybrid unsupervised loss function, which is defined by the loss produced from Nerighbor2Neighbor, the structural similarity between the despeckled results of the probabilistic non-local means method and our method as well as the mean squared error between their features extracted by the VGG network. Experiments on two clinical OCT image datasets show that our method performs better than several popular despeckling algorithms in terms of visual evaluation and quantitative indexes.

Asunto(s)

RESUMEN

It has been reported that RuH2(η2-H2)2(PCy3)2 (1) could mediate CO2 reduction by pinacolborane (HBpin), affording pinBOBpin (7), pinBOCH3 (8), pinBOCHO (9), pinBOCH2OBpin (10), and an unprecedented C2 species pinBOCH2OCHO (11), which meanwhile is converted to the Ru complexes, including the transient 3 (RuH(κ2-O2CH)(CO)(PCy3)2) and 5 (RuH{(µ-H)2Bpin}(CO)(PCy3)2), and the persistent 4 (RuH(κ2-O2CH)(CO)2(PCy3)2) and 6 (RuH2(CO)2(PCy3)2). To gain an insight into the catalysis, a DFT study was carried out. The study identified the key active catalyst to be the hydride 13 (RuH2(CO)(PCy3)2) and characterized the mechanisms leading to the experimentally observed species (3-11). By investigating the experimental system, we learned a new mechanism called σ-π coupling for CO2 decarbonylation. Under this mechanism, CO2 and HBpin first co-coordinate to the Ru center of 13, then σ-π coupling takes place, forming a B-O bond between CO2 and HBpin, Ru-H and Ru-C bonds, and simultaneously breaking the H-Bpin bond, followed by -OBpin group migration to the Ru center, completing the CO2 decarbonylation. An interesting feature regarding the Ru catalysis was the involvement of η1-Hη1-H → η2-H2 and η1-Hη1-Bpin → η2-HBpin reductions, which facilitated the oxidative H-Bpin addition or the coordination mode change of CO2 from η1-O to η2-CO for CO2 activation or σ-π coupling. The facilitation effects could be attributed to the reductions enhancing the electron donations from the Ru center to the antibonding orbitals of the activating bonds.

RESUMEN

Kempe et al. and Milstein et al. have recently advanced the dehydrogenative coupling methodology to synthesize pyrroles from secondary alcohols (e.g., 3) and ß-amino alcohols (e.g., 4), using PNP-Ir (1) and PNN-Ru (2) pincer complexes, respectively. We herein present a DFT study to characterize the catalytic mechanism of these reactions. After precatalyst activation to give active 1A/2A, the transformation proceeds via four stages: 1A/2A-catalyzed alcohol (3) dehydrogenation to give ketone (11), base-facilitated C-N coupling of 11 and 4 to form an imine-alcohol intermediate (18), base-promoted cyclization of 18, and catalyst regeneration via H2 release from 1R/2R. For alcohol dehydrogenations, the bifunctional double hydrogen-transfer pathway is more favorable than that via ß-hydride elimination. Generally, proton-transfer (H-transfer) shuttles facilitate various H-transfer processes in both systems. Notwithstanding, H-transfer shuttles play a much more crucial role in the PNP-Ir system than in the PNN-Ru system. Without H-transfer shuttles, the key barriers up to 45.9 kcal/mol in PNP-Ir system are too high to be accessible, while the corresponding barriers (<32.0 kcal/mol) in PNN-Ru system are not unreachable. Another significant difference between the two systems is that the addition of alcohol to 1A giving an alkoxo complex is endergonic by 8.1 kcal/mol, whereas the addition to 2A is exergonic by 8.9 kcal/mol. The thermodynamic difference could be the main reason for PNP-Ir system requiring lower catalyst loading than the PNN-Ru system. We discuss how the differences are resulted in terms of electronic and geometric structures of the catalysts and how to use the features in catalyst development.

RESUMEN

Density functional theory computations have been applied to gain insight into the CO2 reduction to CH4 with Et3SiH, catalyzed by ammonium hydridoborate 1 ([TMPH](+)[HB(C6F5)3](-), where TMP = 2,2,6,6-tetramethylpiperidine) and B(C6F5)3. The study shows that CO2 is activated through the concerted transfer of H(δ+) and H(δ-) of 1 to CO2, giving a complex (IM2) with a well-formed HCOOH entity, followed by breaking of the O-H bond of the HCOOH entity to return H(δ+) to TMP, resulting in an intermediate 2 ([TMPH](+)[HC(═O)OB(C6F5)3)](-)), with CO2 being inserted into the B-H bond of 1. However, unlike CO2 insertion into transition-metal hydrides, the direct insertion of CO2 into the B-H bond of 1 is inoperative. The computed CO2 activation mechanism agrees with the experimental synthesis of 2 via reacting HCOOH with TMP/B(C6F5)3. Subsequent to the CO2 activation and B(C6F5)3-mediated hydrosilylation of 2 to regenerate the catalyst (1), giving HC(═O)OSiEt3 (5), three hydride-transfer steps take place, sequentially transferring H(δ-) of Et3SiH to 5 to (Et3SiO)2CH2 (6, the product of the first hydride-transfer step) to Et3SiOCH3 (7, the product of the second hydride-transfer step) and finally resulting in CH4. These hydride transfers are mediated by B(C6F5)3 via two SN2 processes without involving 1. B(C6F5)3 acts as a hydride carrier that, with the assistance of a nucleophilic attack of 5-7, first grabs H(δ-) from Et3SiH (the first SN2 process), giving HB(C6F5)3(-), and then leave H(δ-) of HB(C6F5)3(-) to the electrophilic C center of 5-7 (the second SN2 process). The SN2 processes utilize the electrophilic and nucleophilic characteristics possessed by the hydride acceptors (5-7). The hydride-transfer mechanism is different from that in the CO2 reduction to methanol catalyzed by N-heterocyclic carbene (NHC) and PCP-pincer nickel hydride ([Ni]H), where the characteristic of possessing a C═O double bond of the hydride acceptors is utilized for hydride transfer. The mechanistic differences elucidate why the present system can completely reduce CO2 to CH4, whereas NHC and [Ni]H catalysts can only mediate the reduction of CO2 to [Si]OCH3 and catBOCH3, respectively. Understanding this could help in the development of catalysts for selective CO2 reduction to CH4 or methanol.

RESUMEN

Polyols pathway probed: Density functional theory computations reveal that the methyltrioxorhenium-catalyzed deoxydehydration of polyols follows pathway C, which is energetically more favorable than the previously proposed pathways A and B. In addition to serving as solvent/reductant, the alcohol also acts as a shuttle to greatly facilitate various hydrogen-transfer steps.

Asunto(s)

RESUMEN

Density functional theory computations have been carried out to study the mechanism of hydrogenation-based transformation of dimethyl carbonate to methanol, catalyzed by Ru(II)PNN catalyst. The energetic results show that the catalytic transformation includes three sequential stages consistently involving the catalyst: (stage I) transformation of dimethyl carbonate (3) to methyl formate (5) and methanol; (stage II) transformation of methyl formate 5 to formaldehyde and methanol; (stage III) hydrogenation of formaldehyde to methanol. Stages I and II proceed similarly and follow three steps: hydrogen activation, formation of a hemiacetal intermediate via stepwise hydrogen transfer to dimethyl carbonate in stage I or methyl fomate in stage II, and subsequent decomposition of the hemiacetal intermediate to afford methanol. Hydrogenation via carbonyl insertion into the Ru-H bond is less favorable than the stepwise hydrogen-transfer mechanism. Decomposition of hemiacetal takes places by first breaking the hemiacetal O-H bond to give an alkoxide complex, followed by deprotonation of the benzylic arm ligand to the adjacent methoxy group. Comparing the hydrogenation steps in the three stages, hydrogenation in stage I is most difficult, that in stage II is less difficult, and that in stage III is easiest in terms of both kinetics and thermodynamics. This can be ascribed to the stronger electrophilicity of the carbonyl group in methyl formate or formaldehyde than that in dimethyl carbonate and fewer steric effects between the catalyst and methyl formate or formaldehyde than that between the catalyst and dimethyl carbonate. Thermodynamically, both stages I and II are uphill, but stage III is downhill significantly, which is the driving force for the catalytic transformation. The study indicates that the methanol product could facilitate the hydrogen activation involved in the transformation, implying that transformation could be accelerated by initially adding methanol.

RESUMEN

Frustrated Lewis pairs (FLPs) has been applied to catalytic metal-free hydrogenation. Can the FLP reactivity be used for catalytic hydroamination? Using density functional theory (DFT) calculations, we have explored whether the molecules cat1-cat3, which were previously designed by integrating the dearomatization-aromatization effect and the FLP reactivity, can catalyze the intramolecular hydroaminations of non-activated aminoalkenes to afford nitrogen heterocycles. The study shows that the γ-aminoalkene (am1) hydroamination catalyzed by cat1 proceeds via two steps (aminoalkene N-H bond activation and C-N bond formation) with experimentally accessible energetics, giving the five-membered nitrogen heterocycle product 1,1-dimethylpyrrolidine. The N-H bond activation is reversible. The C-N bond formation step undergoes a concerted mechanism and complies with the Markovnikov addition rule. Possible side reactions which may cause catalyst deactivation were confirmed to be energetically unfavorable. The molecules cat2 and cat3 are less effective than cat1 in catalyzing the am1 hydroamination, but the barriers are not too high. By following the most favorable pathway of the cat1-mediated am1 hydroamination, we further extended the substrate (am1) to other aminoalkenes, including the methyl and phenyl ß-substituted am1 (i.e. am2 and am3, respectively), the benzyl-protected primary aminoalkene (am4), and the δ-aminoalkene (am5). The hydroaminations of am2 and am3 have energetics comparable with am1 hydroamination, the am5 hydroamination is energetically less favorable, and the am4 hydroamination is least favorable but could be realizable by elevating the temperature and pressure. We call experimental efforts to synthesize cat1-cat3 or similar new molecules on the basis of the design strategy.

RESUMEN

An unusual DABCO-catalyzed formal [4 + 2] cycloaddition of ethyl allenoate, as a surrogate of a "1,2-dipole", with various arylidenoxindoles has been developed for the synthesis of dihydropyran-fused indoles. The DFT mechanistic study indicates that the cycloaddition takes place stepwise and the essential role of the catalyst is to raise the HOMO of allenoate.

Asunto(s)