Background: With the rapid advances in artificial intelligence—particularly convolutional neural networks—researchers now exploit CT, PET/CT and other imaging modalities to predict epidermal growth factor receptor (EGFR) mutation status in non-small-cell lung cancer (NSCLC) non-invasively, rapidly and repeatably. End-to-end deep-learning models simultaneously perform feature extraction and classification, capturing not only traditional radiomic signatures such as tumour density and texture but also peri-tumoural micro-environmental cues, thereby offering a higher theoretical performance ceiling than hand-crafted radiomics coupled with classical machine learning. Nevertheless, the need for large, well-annotated datasets, the domain shifts introduced by heterogeneous scanning protocols and preprocessing pipelines, and the “black-box” nature of neural networks all hinder clinical adoption. To address fragmented evidence and scarce external validation, we conducted a systematic review to appraise the true performance of deep-learning and radiomics models for EGFR prediction and to identify barriers to clinical translation, thereby establishing a baseline for forthcoming multicentre prospective studies.Methods: Following PRISMA 2020, we searched PubMed, Web of Science and IEEE Xplore for studies published between 2018 and 2024. Fifty-nine original articles met the inclusion criteria. QUADAS-2 was applied to the eight studies that developed models using real-world clinical data, and details of external validation strategies and performance metrics were extracted systematically.Results: The pooled internal area under the curve (AUC) was 0.78 for radiomics–machine-learning models and 0.84 for deep-learning models. Only 17 studies (29%) reported independent external validation, where the mean AUC fell to 0.77, indicating a marked domain-shift effect. QUADAS-2 showed that 31% of studies had high risk of bias in at least one domain, most frequently in Index Test and Patient Selection.Conclusion: Although deep-learning models achieved the best internal performance, their reliance on single-centre data, the paucity of external validation and limited code availability preclude their use as stand-alone clinical decision tools. Future work should involve multicentre prospective designs, federated learning, decision-curve analysis and open sharing of models and data to verify generalisability and facilitate clinical integration.

Designing a chaotic system with a simple structure and complex dynamic behavior is one of the main tasks of chaotic cryptography. This paper designs a new 1D chaotic system called 1D two‐parameters‐sin‐cos (1D‐TPSC). Compared with high‐dimensional chaotic systems, the 1D‐TPSC has a simple structure and is easy to implement with software. The Lyapunov exponent analyzes the parameter space of the 1D‐TPSC in a chaotic state. Furthermore, using sensitivity analysis, cobweb plot, and bifurcation diagram to verify that the sequence generated by 1D‐TPSC has good performance. In addition, the 1D‐TPSC has also been applied in chaotic image encryption. Arnold mapping is used to scramble the plaintext, and random XOR is used to diffuse the scrambled image. Simulation experiments show that the method can remarkably resist standard attack methods.

In view of the uncertainty of model parameters, the influence of external disturbances and sensor noise on the flight of coaxial rotor aircraft during autonomous flight, a robust backstepping sliding mode control algorithm for the position and attitude feedback control system is studied to solve the trajectory tracking problem of an aircraft in the case of unknown external interference. In this study, a non-linear dynamic model based on a disturbed coaxial rotor aircraft was established for an unknown flight. Then, a non-linear robust backstepping sliding mode controller was designed, which was divided into two sub-controllers: the attitude controller and the position controller of the coaxial rotor aircraft. In the controller, virtual control was introduced to construct the Lyapunov function to ensure the stability of each subsystem. The effectiveness of the proposed controller was verified through numerical simulation. Finally, the effectiveness of the backstepping sliding mode control algorithm was verified by flight experiments.

Faces are widely used in information recognition and other fields. Due to the openness of the Internet, ensuring that face information is not stolen by criminals is a hot issue. The traditional encryption method only encrypts the whole area of the image and ignores some features of the face. This paper proposes a double-encrypted face image encryption algorithm. The contour features of the face are extracted, followed by two rounds of encryption. The first round of encryption algorithm encrypts the identified face image, and the second round of encryption algorithm encrypts the entire image. The encryption algorithm used is scrambling and diffusion at the same time, and the keystream of the cryptosystem is generated by 2D SF-SIMM. The design structure of this cryptosystem increases security, and the attacker needs to crack two rounds of the algorithm to get the original face image.

Owing to continual advances in computational fluid dynamics, simulations of fluids have emerged as an important means of analyzing the aerodynamic characteristics of flapping-wing flight. This study establishes an aerodynamic model of bionic flapping-wing flight and uses dynamic hybrid grid technology to divide its flow field and other parts. We formulate equations to control the discretized flow field and combine the Navier–Stokes equation (N–S equation) with the theory of dynamic vortices to solve the flow field of the motion of flapping wings. We analyzed the lift and drag generated by the flapping wing at different wind speeds, amplitudes, frequencies, and chordal torsion angles. The results show the following: (1) wind speed had a significant influence on the lift resistance of the flapping wing. When the wind speed was 1 m/s, the lift force was 1.544 N and increased 4.3 times to 6.636 N when the wind speed was 5 m/s. The resistance increased from 0.39 to 0.88 N. (2) Changes in the amplitude of flutter had little effect on the average lift resistance. When the amplitude was increased from 15° to 45°, the lift force increased to only 0.757 N and drag changed by little. (3) The increase in the flapping frequency improved flight lift. When the frequency was increased from 1 to 5 Hz, the lift increased by 2.9 N (1.78 times) and resistance increased by only 0.08 N. (4) Increasing the chord torsion angle increased flight lift. The lifts at $\mathrm{\beta }$ = 5° and 15° were 6.636 and 6.654 N, respectively, 0.85 and 0.87 N greater than those at $\mathrm{\beta }$ = 0°. As the torsion angle continued to increase, the lift decreased, while the resistance increased more quickly. When $\mathrm{\beta }$ = 15°, the resistance increased by 1.53 N, 9.6 times larger than that at 0°. Increasing the flight speed and flapping frequency can increase flight lift. A small increase in the chord torsion angle increased flight lift, but an excessively large angle led to a substantial increase in drag. Increasing the amplitude of flapping can increase the instantaneous lift generated but has a smaller effect on the average lift. By revealing the influence of different parameters on the lifting resistance of a flapping wing during flight, this paper provides a theoretical foundation for the design and control of bionic flapping-wing aircraft.