Ahmad Nasayreh, Nagham Osman, Laila Tahoun, Asem Mansour, Iyad Sultan
Early detection of malignant mammographic lesions is a critical clinical challenge, particularly where large annotated datasets are unavailable. Supervised deep learning achieves near-radiologist performance but requires large datasets with reliable labels and pixel/ROI annotations. Weakly supervised anomaly detection, requiring only image-level binary labels, offers an annotation-efficient alternative, yet prior GAN- and autoencoder-based methods have struggled to preserve fine anatomical detail in the mammographic domain. We present MammoDiff, to our knowledge the first application of classifier-guided denoising diffusion models to weakly supervised mammographic anomaly detection, trained and evaluated on 3,013 mammograms from 806 patients at King Hussein Cancer Center (KHCC), Jordan. MammoDiff encodes input anatomy through deterministic DDIM noise encoding and reconstructs a counterfactual healthy image guided by the gradient signal of a noise-conditioned binary classifier. The anomaly map is the pixel-wise absolute difference between input and reconstruction, and the image-level score is obtained by spatial mean pooling. A 3 × 3 ablation over noise level L ∈ {250, 500, 750} and gradient scale s ∈ {50, 100, 200} was performed on the validation set. The optimal configuration (L = 750, s = 50) achieved AUROC = 0.7246 [95% CI: 0.6685, 0.7825] on the held-out test set (n = 304), compared with a simplified PatchCore-style baseline (AUROC = 0.6634) and a supervised ResNet-50 reference (AUROC = 0.8100), without pixel-level annotation. The proposed Healthy-Shift Index (HSI) was positive for 94.8% of malignant and 87.4% of normal cases, so it is the separation between the class distributions rather than the sign of the index that carries information. This performance is not clinically deployable: at the Youden-optimal threshold the sensitivity is 0.6013, meaning that 61 of 153 malignant cases are missed, well below the approximately 87% targeted by FDA-cleared triage devices, and the study is single-centre without external validation. MammoDiff is therefore presented as a methodological proof-of-concept and a reproducible baseline for future work, not as a screening or triage tool..
Nearly 15% of couples worldwide experience infertility, with Male Infertility (MI) accounting for 30–50% of these cases. Despite this burden, research on MI lags behind female infertility, particularly in understanding how demographic, clinical, lifestyle, and environmental factors interact to shape reproductive outcomes. Traditional approaches typically examine isolated indicators and struggle to capture the high-dimensional, nonlinear patterns present in Electronic Health Record (EHR) data. This study proposes a data-driven, interpretable framework integrating global fertility rate trends with EHR-derived male fertility data to (i) identify key determinants of MI, (ii) compare multiple classical Long Short-Term Memory (LSTM) architectures against a novel Quantum LSTM model embedding a Variational Quantum Circuit (VQC) directly within the LSTM gating mechanism, and (iii) evaluate the feasibility of quantum-enhanced sequential learning on small, heterogeneous healthcare datasets. Two public datasets were used: a global fertility rate time series (1960–2020) and an EHR–derived male fertility dataset (n = 100, 10 features). After preprocessing (scaling, encoding, multicollinearity checks, and class-imbalance correction), several classical LSTM variants and the proposed Quantum LSTM were trained and validated using 5-fold cross-validation, repeated random train–test splits, and bootstrap resampling. Performance was assessed via accuracy, precision, recall, F1-score, and computational time, with statistical significance evaluated using Friedman and Wilcoxon signed-rank tests. SHAP (SHapley Additive exPlanations) was applied for feature-level and model-level interpretability. Among classical models, unilateral LSTM performed most consistently (91.2% accuracy on fertility-rate data; 92.1% on male-fertility data). The Quantum LSTM achieved the highest accuracy on both datasets (92.6% and 93.8% respectively), a modest but statistically significant improvement (Wilcoxon p = 0.031 and p = 0.024; Cohen's d = 0.42–0.48) with narrower bootstrap confidence intervals, indicating greater stability. SHAP analysis identified daily sitting duration and age as the most influential predictors of male fertility outcomes, followed by season and alcohol consumption frequency, with clinical history variables contributing comparatively less. These findings demonstrate the feasibility of embedding quantum circuits within LSTM gating for small-scale healthcare prediction tasks, yielding incremental rather than transformative gains over classical architectures. Observed associations are exploratory and non-causal, and given the limited sample size (n = 100), single-center source, and simulator-based quantum execution, results should be interpreted as hypothesis-generating rather than clinically conclusive. This work establishes a reproducible benchmarking framework combining classical and quantum-enhanced LSTM models with SHAP-based explainability for early MI risk stratification; future work should validate findings on larger, multi-center cohorts and real quantum hardware.
The loss of homeostasis of the local microenvironment is one of the processes encountered in chronic wound repair and is therefore a big challenge in clinical practice. Endogenous electric fields, being important biophysical cues in the wound bed, guide the direction of movement of the reparative cells, control the inflammatory process and help in the regeneration of the tissues. The classical external electrical stimulation equipment is strongly limited in the portability that can be offered and in the biocompatibility that can be ensured, which limits the wider clinical application of electroceutical approaches for wound care. Some recent developments in the field of electroactive materials give new methodologies for wound therapy by using electrical stimulation. In this article, first, the physiological process of wound closure and the origin and the properties of the endogenous electric fields are summarized. Then, it concentrates on three types of electroactive materials: conductive hydrogels that allow the effective conduction and enhancement of electrical signals through flexible interfaces; self-powered materials that generate their own power, e.g., by means of piezoelectricity or triboelectricity, transforming ambient mechanical energy into therapeutic electrical fields; and multifunctional composite materials that combine electrical stimulation together with antibacterial, exudate control, regulated drug delivery and immunomodulatory functions in a single system. Then the review explains in an ordered fashion the different mechanisms how the electric field can help wound repair, from the generation of directional signaling cues to the modulation of the inflammatory, proliferative and remodeling stages and also to the antimicrobial actions. Finally, it discusses existing barriers, i.e., the conflict between standard and personalized electrical stimulation parameters, the lack of long-term durability assessments and comprehensive biosafety evaluations. The review ends with some outlooks on the progress of future integrated closed-loop intelligent repair systems with sensing, decision-making and responsive functionalities.
Conventional protocols for cancer treatment are usually accompanied by serious limitations, which can be addressed using innovative, minimally invasive methods, such as photothermal therapy (PTT). But, using this method as a standalone therapy is usually accompanied by uneven heat distribution within tumor tissues, and the risk of cancer recurrence. Integrating PTT with complementary therapeutic modalities can help prevent these issues. The development of nanophotothermal systems (NPTSs) as multipurpose, multimodal platforms for smart drug delivery has greatly broadened the therapeutic landscape of cancer treatment. Additionally, recent developments in bio-based and biocompatible nanomaterials have further expanded the potential of NPTSs in clinical applications. Furthermore, NPTSs combine therapeutic functionalities and diagnostic imaging, i.e., tracking modalities, such as photoacoustic, magnetic resonance, or fluorescence imaging, to provide real-time monitoring of drug delivery and treatment response. Translation of preclinical trials into clinical applications for these systems remains challenging due to issues such as toxicity, biodegradability, scalability, standardization, and regulatory approval. This narrative review analyzes two of the most versatile natural polysaccharides, chitosan (CS) and hyaluronic acid (HA), which have complementary physicochemical and biological properties. Covering the advantages of CS and HA for the application in multifuntional and multimodal nanophotothermal systems, their complementary effects, as well as their functional implications in therapeutic settings, this work aims to create a framework that will serve as a basis for further research in optimizing the design of multifunctional drug-delivery nanosystems, and at the same time, facilitate the transition of photothermal therapy from the bench to clinical use.
Saumya Jamuar, Richard Palmer, Haur Yueh Lee, Faith Li-Ann Chia, Choo Beng Goh, Stuart Lee, Petra Helmholz, Shermaine Chan, Gareth Baynam
People living with rare diseases (PLWRD) often face significant challenges in receiving timely and accurate diagnoses, leading to what is known as a diagnostic odyssey. Digital phenotyping (DP) offers a promising solution by leveraging advanced technology to capture unique digital signatures associated with various rare diseases. This innovative approach not only aids in the identification of these conditions but also facilitates the detection of digital biomarkers (DBM). By utilizing DP and DBMs, we can improve both the diagnosis and management of RDs, ultimately leading to better health outcomes for affected participants. To identify whether DBMs can be identified by DP utilizing 3D facial imaging techniques in outpatient settings in participants with RDs. This multicentre observational study was conducted in 3 hospitals in Singapore. The eligible participants were male and female RD participants of various age groups. The study duration was 4 years and 8 months. Twenty participants of Chinese genetic ancestry were photographed using a 3D camera. Additionally, two participants with hereditary angioedema (HAE) were photographed during acute stages of disease flare-ups. The obtained facial scans of participants (that included participants with HAE in non-acute phase) were plotted using Artificial Intelligence-powered software - Cliniface. The growth curves and facial landmarks obtained were compared against the growth curves of normal RD-unaffected individuals of Chinese genetic ancestry. The two participants with HAE were photographed qualitatively over a longer period of time, and their scans were plotted, yielding growth curves. Distinct facial markers such as periorbital swelling were identified in two qualitatively assessed HAE participants during flare-up stages. This provides an opportunity to explore and validate whether these facial signatures in a disease condition can be assigned as DBM for HAE. and relevance: This study explores the utility of 3D facial analysis as a DBM in rare diseases such as HAE. These findings offer preliminary objective evidence that supports existing subjective reports of facial features in the literature. Additionally, while DP's diagnostic capabilities may be limited, it successfully identified DBM, which could facilitate disease monitoring in conditions such as HAE.
