Deep learning in ultrasound(US) imaging aims to construct foundational models that accurately reflect the modality's unique characteristics. Nevertheless, the limited datasets and narrow task types have restricted this field in recent years. To address these challenges, we introduce US-MTD120K, a multi-task ultrasound dataset with 120,354 real-world two-dimensional images. This dataset covers three standard plane recognition and two diagnostic tasks in ultrasound imaging, providing a rich basis for model training and evaluation. We detail the data collection, distribution, and labelling processes, ensuring a thorough understanding of the dataset's structure. Furthermore, we conduct extensive benchmark tests on 27 state-of-the-art methods from both supervised and self-supervised learning(SSL) perspectives. In the realm of supervised learning, we analyze the sensitivity of two main feature computation methods to ultrasound images at the representational level, highlighting that models which judiciously constrain global feature computation could potentially serve as a viable analytical approach for US image analysis. In the context of self-supervised learning, we delved into the modelling process of self-supervised learning models for medical images and proposed an improvement strategy, named MoCo-US, a solution that addresses the excessive reliance on pretext task design from the input side. It achieves competitive performance with minimal pretext task design and enhances other SSL methods simply. The dataset and the code will be available at https://github.com/JsongZhang/CDOA-for-UMTD .
1 March 2025 in International Journal of Women's Health
Background: The relationship between Polycystic ovary syndrome (PCOS) and immune dysregulation, along with metabolic disturbances, remains unclear. This study used Mendelian Randomization (MR) to investigate causal relationships between immune cells, PCOS, and possible metabolite mediators. Methods: We explored the genetic-level relationship between immune cells and PCOS, focusing on metabolites as potential mediators. Data from genome-wide association studies (GWAS) included 731 immune cell types (n=3757), 1400 plasma metabolites (n=8299), and PCOS cases (n=797) versus controls (n=140,558). Bidirectional MR analysis examined immune-PCOS relationships, while two-step MR and mediation analysis identified metabolites as potential mediators. The inverse variance-weighted (IVW) method was used for primary causal assessment, with sensitivity analysis validating results. Results: We identified a total of 33 immune cells that were associated with increased or decreased risk of PCOS (P < 0.05), and these immune cells were also associated with alterations in certain metabolite levels (P < 0.05). Among them, 12 immune cells were found to influence the occurrence of PCOS through the mediation of 17 metabolites. Notably, the effects of Myeloid DC %DC, NKT AC, CD20 on CD20- CD38-, CD25 on memory B cell, and HLA DR on CD33dim HLA DR+ CD11b+ were mediated by multiple metabolites on PCOS development. Similarly, histidine betaine (hercynine) levels and alpha-ketoglutarate to ornithine ratio mediated the association of more than one immune cell with PCOS. Conclusion: This study highlights 12 immune cells impacting PCOS through 17 metabolites, advancing the understanding of immune mechanisms in PCOS risk and suggesting potential therapeutic approaches targeting immune modulation.