To reduce the diagnostic burden on radiologists, recent studies have explored automatic chest X-ray (CXR) report generation via artificial intelligence. Yet, achieving robust cross-modal alignment between medical images and textual reports remains a major challenge. In this paper, we propose CyclicAligner, a knowledge-enhanced cyclical alignment framework for CXR report generation. CyclicAligner adopts a novel cyclical training paradigm with four tightly coupled tasks to effectively learn cross-modal semantic alignment: (1) an image-to-text generation task that aligns visual semantics with clinical findings, (2) a text-to-text reconstruction task that strengthens language modeling, (3) a hybrid-to-text reconstruction task that mixes vision and language tokens for text reconstruction, and (4) a traceback-alignment task that re-encodes texts generated by the image-to-text branch for text reconstruction and aligns the reconstructed text with the reference. To further enhance cross-modal understanding, we integrate domain-specific medical entity knowledge extracted from a pre-trained encoder to enrich both vision and language tokens. Moreover, CyclicAligner jointly predicts medical tags and narrative reports within a unified auto-regressive pipeline, where the tags serve as auxiliary semantic anchors that guide the report generation. Extensive experiments on public datasets demonstrate the effectiveness of our method for clinical-coherent CXR report generation. The related code is available at https://github.com/yangyan22/CyclicAligner.

Recent studies highlight the beneficial effects of instructional videos (IVs) and intelligent tutoring systems (ITS) on the oral development and learning attitudes of learners of English as a Foreign Language (EFL). Despite these advances, studies investigating the combined impact Of Al-driven tutoring systems with instructional videos on EFL learners' listening and speaking skills remain sparse. This study aims to fill this research gap by employing a 15-week quasi-experimental design with 43 adult participants, divided into two groups. The experimental group (Group IVIT) utilized instructional videos integrated with an Al-driven tutoring system (iTutor), whereas the control group (Group IV) engaged with instructional videos alone. Comparative analysis demonstrated significant enhancements in listening and speaking proficiencies within Group IVIT. which notably outperformed Group IV in speaking skill development. Additionally, learners in Group IVIT reported more positive perceptions of the opportunities for speaking provided by IVIT and experienced less anxiety related to speaking tasks. These findings emphasize the significant potential of integrating intelligent tutoring systems with traditional learning modalities to accelerate language acquisition in adult EFL learners.

Automatic detection of cervical lesion cells or cell clumps using cervical cytology images is critical to computer-aided diagnosis (CAD) for accurate, objective, and efficient cervical cancer screening. Recently, many methods based on modern object detectors were proposed and showed great potential for automatic cervical lesion detection. Although effective, several issues still hinder further performance improvement of such known methods, such as large appearance variances between single-cell and multi-cell lesion regions, neglecting normal cells, and visual similarity among abnormal cells. To tackle these issues, we propose a new task decomposing and cell comparing network, called TDCC-Net, for cervical lesion cell detection. Specifically, our task decomposing scheme decomposes the original detection task into two subtasks and models them separately, which aims to learn more efficient and useful feature representations for specific cell structures and then improve the detection performance of the original task. Our cell comparing scheme imitates clinical diagnosis of experts and performs cell comparison with a dynamic comparing module (normal-abnormal cells comparing) and an instance contrastive loss (abnormal-abnormal cells comparing). Comprehensive experiments on a large cervical cytology image dataset confirm the superiority of our method over state-of-the-art methods.

In the field of remote sensing, data acquired from a single sensor usually can't meet the needs of some special applications, because the information extracted from the data are often incomplete and limited. Data fusing can solve this problem, but it will lead to the redundant information. In this paper, we proposed a novel manifold learning approach to perform dimensionality reduction for the fusing optical and SAR data. And three typical manifold learning models, namely, ISOMAP, local linear embedding (LLE) and principle component analysis (PCA), were utilized to test the robustness of our method by comparing with the land cover classification results. Our experimental results showed that our proposed method obtained the best land cover classification results among these approaches for the fusing optical and SAR data.