Predicting bone metastasis-free survival in non-small cell lung cancer from preoperative CT via deep learning.

Accurate prediction of bone metastasis-free survival (BMFS) after complete surgical resection in patients with non-small cell lung cancer (NSCLC) may facilitate appropriate follow-up planning. The aim of this study was to establish and validate a preoperative CT-based deep learning (DL) signature to predict BMFS in NSCLC patients. We performed a retrospective analysis of 1547 NSCLC patients who underwent complete surgical resection, followed by at least 36 months of monitoring at two hospitals. We constructed a DL signature from multiparametric CT images using 3D convolutional neural networks, and we integrated this signature with clinical-imaging factors to establish a deep learning clinical-imaging signature (DLCS). We evaluated performance using Harrell's concordance index (C-index) and the time-dependent receiver operating characteristic. We also assessed the risk of bone metastasis (BM) in NSCLC patients at different clinical stages using DLCS. The DL signature successfully predicted BM, with C-indexes of 0.799 and 0.818 for the validation cohorts. DLCS outperformed the DL signature with corresponding C-indexes of 0.806 and 0.834. Ranges for area under the curve at 1, 2, and 3 years were 0.820-0.865 for internal and 0.860-0.884 for external validation cohorts. Furthermore, DLCS successfully stratified patients with different clinical stages of NSCLC as high- and low-risk groups for BM (p < 0.05). CT-based DL can predict BMFS in NSCLC patients undergoing complete surgical resection, and may assist in the assessment of BM risk for patients at different clinical stages.

Deep learning signature based on multiphase enhanced CT for bladder cancer recurrence prediction: a multi-center study.

Background: Accurate stratification of recurrence risk for bladder cancer (BCa) is essential for precise individualized therapy. This study aimed to develop and validate a model for predicting the risk of recurrence in BCa patients postoperatively using 3-phase enhanced CT images. Methods: We retrospectively enrolled 874 BCa patients across four centers between January 2006 and December 2021. Patients from one center were used as training set, while the remaining patients went into the validation set. We trained a deep learning (DL) model based on convolutional neural networks using 3-phase enhanced CT images. The resulting prediction scores were entered into Cox regression analysis to obtain DL scores and construct a DL signature. DL scores and clinical features were then used as deep learning radioclinical signature. The predictive performance of DL signature was assessed according to concordance index and area under curve compared with deep learning radioclinical signature, clinical model and a widely accepted staging grading system. Recurrence-free survival (RFS) and overall survival (OS) were also predicted in order to further assess survival benefits. Findings: DL signature showed strong power for predicting recurrence (concordance index, 0.869; area under curve, 0.889) in validation set, outperforming other models and system. In addition, we divided RFS and OS into high and low risk groups by selecting appropriate cutoff values for DL signature, and calculated cumulative recurrence risk rates for both groups. Interpretation: Our proposed DL signature shows promising potential as clinical aid for predicting postoperative recurrence risk in BCa and for stratifying the risk of RFS and OS, which can be applied to guide personalized precision therapy. Funding: There are no sources of funding for this manuscript.

Fault Diagnosis Method for Imbalanced Data Based on Multi-Signal Fusion and Improved Deep Convolution Generative Adversarial Network.

The realization of accurate fault diagnosis is crucial to ensure the normal operation of machines. At present, an intelligent fault diagnosis method based on deep learning has been widely applied in mechanical areas due to its strong ability of feature extraction and accurate identification. However, it often depends on enough training samples. Generally, the model performance depends on sufficient training samples. However, the fault data are always insufficient in practical engineering as the mechanical equipment often works under normal conditions, resulting in imbalanced data. Deep learning-based models trained directly with the imbalanced data will greatly reduce the diagnosis accuracy. In this paper, a diagnosis method is proposed to address the imbalanced data problem and enhance the diagnosis accuracy. Firstly, signals from multiple sensors are processed by the wavelet transform to enhance data features, which are then squeezed and fused through pooling and splicing operations. Subsequently, improved adversarial networks are constructed to generate new samples for data augmentation. Finally, an improved residual network is constructed by introducing the convolutional block attention module for enhancing the diagnosis performance. The experiments containing two different types of bearing datasets are adopted to validate the effectiveness and superiority of the proposed method in single-class and multi-class data imbalance cases. The results show that the proposed method can generate high-quality synthetic samples and improve the diagnosis accuracy presenting great potential in imbalanced fault diagnosis.

Comparison of optical quality and distinct macular thickness in femtosecond laser-assisted versus phacoemulsification cataract surgery.

BACKGROUND: Optical quality and macular thickness changing optical quality is rarely reported after femtosecond laser-assisted cataract surgery (FLACS). In current research, we evaluated optical quality recovery and distinct macular thickness changes after FLACS and phacoemulsification cataract surgery (PCS). METHODS: A total of 100 cataract patients (100 eyes) were included (50 eyes for the FLACS group and 50 eyes for the PCS group). Modulation transfer function (MTF), point spread function (PSF) and dysfunctional lens index (DLI) were measured by a ray-tracing aberrometer (iTrace). Uncorrected distance visual acuity (UDVA) and corrected distance visual acuity (CDVA) were also assessed pre-operation,1 week and 1 month after surgery. The MTF values at spatial frequencies of 5, 10, 15, 20, 25 and 30 cycles/degree (c/d) were selected. We used optical coherence tomography (OCT) to assess the macular thickness of different regions pre-operatively and1month after the surgery. RESULTS: In PCS group, we found the statistically significant differences between pre-operation and post-operation in DLI (p < 0.0001), PSF (strehl ratio, SR) (p = 0.027) and MTF (p = 0.028), but not intraocular pressure (IOP) (p = 0.857). The differences between pre-operation and post-operation for DLI (p = 0.031), SR (p = 0.01) and IOP (p = 0.03), but not MTF (p = 0.128) were also found in FLACS group. The differences were statistically significant when the spatial frequencies were at 5, 10 and 25 (p = 0.013, 0.031 and 0.048) between pre-operation and post-operation in PCS group but not FLACS group at 1 month. In PCS group, we found the differences between pre-operation and post-operation in nasal inter macular ring thickness (NIMRT) (p = 0.03), foveal volume (FV) (p = 0.034) and average retinal thickness (ART) (p = 0.025) but not FLACS group at 1 month. CONCLUSION: FLACS is safe that did not cause significant increase of macular thickness in current study. However, it also cannot produce better optical quality. In contrast, PCS can produce macular thickness changes, but better optical quality recovery. The slightly retinal change may not affect optical quality.

Entropy Exchange and Thermodynamic Properties of the Single Ion Cooling Process.

A complete quantum cooling cycle may be a useful platform for studying quantum thermodynamics just as the quantum heat engine does. Entropy change is an important feature which can help us to investigate the thermodynamic properties of the single ion cooling process. Here, we analyze the entropy change of the ion and laser field in the single ion cooling cycle by generalizing the idea in Reference (Phys. Rev. Lett. 2015, 114, 043002) to a single ion system. Thermodynamic properties of the single ion cooling process are discussed and it is shown that the Second and Third Laws of Thermodynamics are still strictly held in the quantum cooling process. Our results suggest that quantum cooling cycles are also candidates for the investigation on quantum thermodynamics besides quantum heat engines.