Magnetic particle imaging (MPI) has demonstrated its advantages of high sensitivity and temporal resolution in various preclinical applications. However, during the imaging process, the signal is susceptible to different noises, resulting in severe stripe artifacts in reconstructed MPI images. This phenomenon will be further aggravated in scenarios with low-concentration particles, which is a standard practice in biological applications, thereby seriously hindering the identification of key information. To solve this problem, we propose a joint optimization approach called diffusion denoising model for MPI (DDMPI) that integrates diffusion model with Transformer to remove the artifacts directly from MPI images obtained in the low-concentration scenarios. In DDMPI, a latent encoder generates prior features containing the relevance mapping between the contents and the artifacts within MPI images, and a conditional latent diffusion model optimizes these prior features. A U-shape Transformer module incorporates the prior features by a hierarchical integration module and utilizes a strip-self-attention module to capture the spatial distribution of the artifacts. Ablation experiments demonstrate the effectiveness of these modules in DDMPI. Extensive experiments, including simulation, phantom and in vivo experiments, demonstrate that DDMPI effectively removes artifacts and recovers fine details. Additionally, DDMPI is independent of the primary image reconstruction methods of various scanning devices. Thus, DDMPI can be not only practically applied to in vivo imaging but also flexibly combined with various existing MPI devices to effectively improve the imaging quality and provide critical information about diseases.

Continuous monitoring of blood pressure (BP) waveform is challenging in clinical applications due to the invasive nature of traditional techniques. As a result, there is a growing focus on the estimation of continuous BP waveforms from photoplethysmography (PPG) signals obtained through affordable wearable Internet of Medical Things (IoMT) devices. To address this demand, we introduce diffusion for continuous noninvasive BP (DiffCNBP), a lightweight model that employs a joint optimization approach incorporating sequence learning, diffusion modeling, and conditional embedding. The sequence learning module comprises stacked Transformer encoders to capture the temporal information of the PPG signals. These dynamically related features are then fed into a lightweight diffusion module based on structured state-space sequence encoders to learn detailed variation features associated with vascular dynamics. Additionally, the conditional embedding module introduces constraints to incorporate physiologically specific prior information for model estimation, thereby enhancing the fidelity of the estimated BP waveform. Our proposed method was validated using subject-wise fivefold cross-validation on a multisource database of 1415 subjects. This database included data from intensive care unit IoMT applications, where finger-based PPG sensors were employed alongside invasive BP sensors. Experimental results demonstrate that DiffCNBP outperforms other state-of-the-art methods with an average root-mean-square error of 4.36 mmHg for waveform estimation. The mean error ± standard deviation error of systolic and diastolic BP was$0.67~\pm ~4.29$mmHg and$0.37~\pm ~2.46$mmHg, respectively, meeting the clinical standards. Furthermore, we demonstrated the robust long-term trend-tracking ability of DiffCNBP on resource-constrained devices, indicating its potential IoMT deployment in clinical settings.