The core challenge is a fundamental tension between security and speed. Modern cryptography relies on public-key algorithms like RSA and elliptic-curve techniques to authenticate devices and establish encryption keys, but a sufficiently large quantum computer could break these schemes outright. Standards bodies, including the US National Institute of Standards and Technology (NIST), have already standardized quantum-resistant replacements, notably ML-KEM for key establishment and ML-DSA for digital signatures. The catch is that post-quantum algorithms tend to produce much larger messages and demand more computation, and brain-computer interfaces are among the most latency-sensitive systems in existence. When the loop from sensing brain activity to delivering feedback determines whether a prosthetic limb responds correctly or a stimulation protocol stays safe, even modest delays matter.

The team, led by Tariq Qayyum, Asadullah Tariq, and Abdelkader Nasreddine Belkacem, published their design in Frontiers in Systems Neuroscience. Their solution rests on an elegant architectural separation: expensive public-key cryptography happens only during session establishment, while the continuous stream of EEG frames and control messages travels on a fast path protected by symmetric authenticated encryption with associated data, or AEAD. Because symmetric encryption is cheap and predictable, the per-packet cost stays constant, and the heavyweight quantum-safe operations never appear on the real-time path. The design supports three flavors of session setup: classical X25519 elliptic-curve exchange, post-quantum ML-KEM-768, and a hybrid that combines both secrets, following IETF migration guidance.

Identity binding adds another layer of assurance. Each headset and hub is provisioned with pinned ML-DSA public keys, so session establishment is tied to stable device identities rather than to certificate chains that would bloat the handshake. The negotiated mode, algorithm list, ephemeral key material, and a hash of the entire transcript are all folded into signed bindings and the key-derivation context. If an attacker attempts to strip out the post-quantum algorithms, downgrade the session to a weaker mode, or splice together two unrelated sessions in a man-in-the-middle attack, the receiver simply fails closed and rejects the connection.

Replay protection is engineered with the same care. Every packet carries a session identifier, an epoch counter, a message type, and a monotonically increasing sequence number, all integrity-protected as associated data. The receiver maintains a sliding replay window that tolerates ordinary wireless packet reordering while deterministically rejecting duplicated or stale traffic. Crucially, packets from one session or key epoch cannot be smuggled into another, a property that matters enormously in closed-loop systems where silent cross-session reuse could masquerade as incorrect commands rather than as an obvious transport failure.

To measure real-world impact, the researchers replayed traffic derived from four public EEG datasets, including motor-imagery benchmarks and an inner-speech dataset, over Bluetooth-like and Wi-Fi-like link profiles in one-hop and two-hop topologies. Each security mode faced identical packetization, loss, delay, and reordering conditions, with at least 400 measurements per data point and bootstrap confidence intervals plus non-parametric paired tests for statistical rigor. The results were striking: in the most constrained Bluetooth-style one-hop configuration, the hybrid post-quantum mode raised the 95th-percentile latency by just 0.45 milliseconds relative to an unsecured baseline, while maintaining 99.0 percent packet delivery. It actually posted the lowest latency percentiles among all secured schemes tested, outperforming TLS 1.3, DTLS, QUIC, WireGuard, and IPsec baselines.

The analysis revealed where the post-quantum cost actually lands: in handshake size, not streaming speed. The hybrid and post-quantum modes require roughly 4.6 kilobytes of handshake transcript compared with a fraction of a kilobyte for classical setups, and measured cryptographic handshake processing stayed sub-millisecond on the test host. A companion projection for wearable-class Cortex-M4 microcontrollers, derived from published PQM4 cycle benchmarks and 10 to 20 milliwatt active-CPU power profiles, estimated ML-KEM-768 operations at around 10 to 11 milliseconds and a fraction of a millijoule each. Because those operations occur only at setup or during infrequent public-key refreshes, they pose no burden on the continuous data stream.

Safety engineering runs throughout the design. Cryptographic failures such as authentication errors, replay detection, or session-state violations trigger an immediate transition to a safe state. Timing anomalies, by contrast, are handled by a hysteretic watchdog: one missed critical update raises a warning, two consecutive misses suppress new actuation and enter a degraded state, and three consecutive misses activate SafeMode, disabling stimulation and placing motor or prosthetic control in a neutral, stopped, or locked state. For the evaluated 20 to 40 millisecond update periods, this bounds the communication contribution to time-to-safe-state at 60 to 120 milliseconds plus local shutdown time, a deliberate balance between tolerating ordinary wireless hiccups and preventing unsafe operation under sustained jamming or delay.

Active attack testing underscored the framework’s resilience. In modeled injection, replay, and man-in-the-middle session-splicing scenarios, the hybrid mode produced the lowest false-acceptance rates, with identity pinning and transcript binding preventing every modeled session-splicing attempt across the full range of link-quality conditions. The researchers are candid about limitations: denial-of-service and jamming cannot be defeated by cryptography alone, and risks such as compromised hubs, cloud compromise, and supply-chain attacks are addressed through architectural controls like key isolation, signed updates, revocation registries, and audit logging rather than through experimentally validated attack tests. The evaluation, they stress, is communication-layer evidence, with hardware-specific effects left for device-integration studies.

The significance of the work extends well beyond one prototype. As neural devices move toward clinical approval, regulatory frameworks increasingly demand systematic cybersecurity risk management, secure updates, and documented residual risk. A harvest-now, decrypt-later adversary could record encrypted EEG sessions today and attempt decryption once quantum computers mature, making post-quantum readiness a long-horizon privacy requirement rather than a distant concern, particularly for data as sensitive as brain signals, which prior research has shown can leak private attributes. PQ-NeuroLink demonstrates that the quantum-safe transition need not come at the price of real-time performance, providing a reproducible, open foundation for the next generation of secure, low-latency brain-computer interface deployments.

Subject of Research: Latency-aware post-quantum secure communication for distributed EEG brain-computer interfaces

Article Title: PQ-Neurolink: latency-aware post-quantum-ready secure communication for distributed brain–computer interfaces

Article References: Qayyum, T., Tariq, A., & Belkacem, A. N. (2026). PQ-Neurolink: latency-aware post-quantum-ready secure communication for distributed brain–computer interfaces. Frontiers in Systems Neuroscience, 20, Article 1891041. https://doi.org/10.3389/fnsys.2026.1891041

Image Credits: AI Generated

DOI: 10.3389/fnsys.2026.1891041

Keywords: brain-computer interface, post-quantum cryptography, EEG, ML-KEM, ML-DSA, secure communication, latency, hybrid key exchange, authenticated encryption, closed-loop systems, neurosecurity, wearable devices

The scheme, named CNNHSIE, short for Convolutional Neural Network and Hyperchaotic System-based Image Encryption, is built around three core components. The first is a hyperchaotic system, a class of dynamical systems that exhibit not just one but multiple positive Lyapunov exponents, meaning their trajectories diverge exponentially in several independent directions at once. This extreme sensitivity to initial conditions makes hyperchaotic sequences ideal for generating pseudo-random matrices that look statistically indistinguishable from true noise to any observer who does not possess the exact initial values. In CNNHSIE, the hyperchaotic system produces the random matrix that drives both the scrambling of pixel positions and the alteration of pixel values, two operations that cryptographers call permutation and diffusion.

Permutation and diffusion are the twin pillars of image encryption. Permutation rearranges the positions of pixels so that the spatial structure of the original image is destroyed, while diffusion changes the actual pixel values so that even a single-bit change in the plaintext cascades through the entire ciphertext. Traditional schemes often perform these operations in separate, sequential passes, which can be computationally expensive and can leave exploitable structure behind. The Iranian researchers tackled this bottleneck with two novel techniques. The first, called HSPT for high-speed permutation technique, reorders pixels with an efficiency that dramatically reduces encryption time compared with conventional sorting-based or swapping-based permutation strategies. The second, DPDT for dynamic pixel diffusion technique, modifies pixel values in a way that adapts dynamically during the encryption process rather than following a fixed pattern, making it substantially harder for an attacker to model or reverse the transformation.

What sets CNNHSIE apart from many previous chaos-based cryptosystems is its incorporation of a convolutional neural network into the key-generation process. Specifically, the authors employ the VGG16 architecture, a deep convolutional network originally developed for large-scale image classification, to extract features from the plain image itself. These extracted features are then used to compute the public key. This design choice has an elegant consequence: the encryption key material is tied to the unique statistical fingerprint of each individual image, so even minor differences between two plaintext images lead to different key components. An attacker who obtains the ciphertext for one image gains no reusable shortcut for decrypting another, because the network-derived key material is image-specific and changes from one input to the next.

The researchers evaluated CNNHSIE rigorously using standard test images from the USC-SIPI image database, a widely used benchmark repository maintained by the University of Southern California, along with public datasets drawn from Radiopaedia, an open-access radiology resource, and Kaggle, the popular machine-learning data platform. The inclusion of medical imaging datasets is significant, because medical images are among the most sensitive classes of data transmitted over networks, and regulatory frameworks around the world impose strict confidentiality requirements on them. A scheme that performs well on both natural photographic images and clinical scans has a broader practical reach than one tested only on generic benchmarks.

The headline results are striking. For the standard 1024 by 1024 pixel test image known as Male, CNNHSIE achieved a unified average changing intensity, or UACI, of 33.4809 percent, a number that sits almost exactly at the theoretical ideal of 33.46 percent for a perfectly random-looking ciphertext. The number of pixels change rate, or NPCR, reached 99.6140 percent, meaning that when the encryption key was perturbed by even a minuscule amount, more than 99.6 percent of all ciphertext pixels changed value, the hallmark of extreme key sensitivity. The information entropy of the encrypted image measured 7.9998 bits out of a theoretical maximum of 8 bits for an 8-bit image, indicating that the distribution of pixel values in the ciphertext is essentially uniform and carries virtually no exploitable information about the original.

Correlation analysis told a similarly compelling story. Adjacent pixels in a natural image are almost always highly correlated, because neighboring pixels tend to have similar brightness values. Effective encryption must drive this correlation to zero in every direction, horizontal, vertical and diagonal. CNNHSIE produced ciphertexts with near-zero correlation coefficients across all tested directions, confirming that the permutation and diffusion stages thoroughly destroyed the statistical structure of the plaintext. Perhaps most importantly for real-world deployment, the scheme achieved these security metrics while encrypting the large 1024 by 1024 Male image in just 0.1216 seconds, a speed that places it firmly in the territory of real-time applications such as live video streaming, secure telemedicine consultations and on-the-fly transmission of surveillance imagery.

The authors report that CNNHSIE demonstrated the ability to resist a variety of common attacks, including the chosen-plaintext and known-plaintext attacks that plague weaker image cryptosystems. In a chosen-plaintext attack, an adversary deliberately encrypts specially crafted images to probe the internal workings of the algorithm; image-specific key material derived from the VGG16 feature extractor makes such probing far less informative, because the relationship between plaintext and key changes with every input. The combination of hyperchaotic pseudo-randomness, dynamic diffusion and neural key generation creates multiple independent layers of defense, so a weakness in any single layer does not compromise the whole system. High key sensitivity further ensures that brute-force attempts to guess the chaotic initial conditions face an astronomically large search space with no smooth gradient to follow.

This work sits within a rapidly growing research frontier that merges deep learning with chaos-based cryptography. Earlier efforts have explored DNA encoding schemes combined with fractional-order hyperchaotic systems, Latin square constructions, Joseph traversal and discrete Fourier transforms, and several groups have experimented with neural networks for key generation or ciphertext post-processing. Zefreh himself has a track record in this area, with prior publications on Latin square-based encryption and parallel diffusion-confusion cryptosystems. What distinguishes the new contribution is the explicit pairing of a proven, off-the-shelf feature extractor with high-speed permutation and dynamic diffusion, achieving a balance of security metrics and raw throughput that many earlier designs pursued separately rather than simultaneously.

The practical implications extend across the sectors where image confidentiality matters most. In healthcare, secure transmission of radiological images could protect patient privacy without adding perceptible latency to clinical workflows. In military and defense contexts, real-time encryption of reconnaissance imagery could reduce the window of vulnerability during transmission. In finance, encrypted document images could move through cloud infrastructure with stronger guarantees against tampering. The researchers note that data and code are available upon request from the corresponding author, which opens the door for independent verification and adoption by other laboratories. As digital image traffic continues its exponential growth, schemes like CNNHSIE, which treat speed and security as joint requirements rather than competing trade-offs, are likely to shape the next generation of multimedia security systems, and the marriage of chaotic dynamics with deep neural networks may prove to be one of the defining patterns of that evolution.

Subject of Research: A CNN and hyperchaotic system-based image encryption algorithm using high-speed permutation and dynamic diffusion

Article Title: CNNHSIE: CNN and hyperchaotic-based image encryption algorithm using high-speed permutation and dynamic diffusion

Article References: Zarei Zefreh, E., & Heydari, M. (2026). CNNHSIE: CNN and hyperchaotic-based image encryption algorithm using high-speed permutation and dynamic diffusion. Neural Computing and Applications, 38(17), Article 722. https://doi.org/10.1007/s00521-026-12419-y

Image Credits: AI Generated

DOI: 10.1007/s00521-026-12419-y

Keywords: image encryption, convolutional neural network, hyperchaotic system, VGG16, permutation, diffusion, cryptography, information entropy, NPCR, UACI, medical imaging security, deep learning