Fast prediction of indoor airflow distribution inspired by synthetic image generation artificial intelligence.

Prediction of indoor airflow distribution often relies on high-fidelity, computationally intensive computational fluid dynamics (CFD) simulations. Artificial intelligence (AI) models trained by CFD data can be used for fast and accurate prediction of indoor airflow, but current methods have limitations, such as only predicting limited outputs rather than the entire flow field. Furthermore, conventional AI models are not always designed to predict different outputs based on a continuous input range, and instead make predictions for one or a few discrete inputs. This work addresses these gaps using a conditional generative adversarial network (CGAN) model approach, which is inspired by current state-of-the-art AI for synthetic image generation. We create a new Boundary Condition CGAN (BC-CGAN) model by extending the original CGAN model to generate 2D airflow distribution images based on a continuous input parameter, such as a boundary condition. Additionally, we design a novel feature-driven algorithm to strategically generate training data, with the goal of minimizing the amount of computationally expensive data while ensuring training quality of the AI model. The BC-CGAN model is evaluated for two benchmark airflow cases: an isothermal lid-driven cavity flow and a non-isothermal mixed convection flow with a heated box. We also investigate the performance of the BC-CGAN models when training is stopped based on different levels of validation error criteria. The results show that the trained BC-CGAN model can predict the 2D distribution of velocity and temperature with less than 5% relative error and up to about 75,000 times faster when compared to reference CFD simulations. The proposed feature-driven algorithm shows potential for reducing the amount of data and epochs required to train the AI models while maintaining prediction accuracy, particularly when the flow changes non-linearly with respect to an input.

Partition walls as effective protection from bio-aerosols in classrooms - an experimental investigation.

Introduction: During a pandemic, protective measures to prevent bio-aerosol based infections, such as Corona Virus Infection Disease 19 (COVID 19), are very important. Everyday face masks can only partially block aerosols, and their effectiveness also depend on how well the person is wearing it. They are recommended for classroom situations during high pandemic activity. However, 'unprotected' communication with and among children is fundamental from the pedagogical and psychological point of view for normal psychosocial development and teaching. Partition walls around the persons can theoretically provide substantial standardized mechanical protection against the spread of droplets and aerosols, either as additional protection to face masks or as an alternative. Methods: In the present research, the protection effectiveness of partition walls was investigated. With mannequin heads, fog generators, line lasers and a classroom-like setup with protective walls, flow visualization and aerosol concentration measurements were performed. Additionally, an active fan-suction system was tested to remove the channelled aerosols on top of the partition walls before they reach other persons in the room. Results: It was found that partition walls protect neighbours from bio-aerosol contact regardless of whether they wear masks or not. The combination with standardized room ventilation enforces this effect. Moreover, the experiments performed here clearly showed that partition walls may protect neighbours from bio-aerosols better than suboptimally fitting everyday face masks only. Conclusion: Partition walls are the most effective protection against infectious bio-aerosols in classroom settings and should be combined with standardized ventilation as the preferred method for classrooms during the current COVID 19 pandemic.

Myocardial kinetics of reporter probe 124I-FIAU in isolated perfused rat hearts after in vivo adenoviral transfer of herpes simplex virus type 1 thymidine kinase reporter gene.

UNLABELLED: Reporter gene imaging holds promise for noninvasive monitoring of cardiac gene therapy. We recently demonstrated that (124)I-labeled 2'-fluoro-2'-deoxy-5'-iodo-1beta-d-arabinofuranosyluracil ((124)I-FIAU) is suitable for PET of myocardial expression of herpes simplex virus type 1 thymidine kinase reporter gene (HSV1-tk). In contrast to previous studies in tumors, early specific uptake was followed by rapid washout. Myocardial kinetics of (124)I-FIAU are still poorly understood. This study aimed at a further investigation under controlled conditions using an isolated heart perfusion model. METHODS: Male Wistar rats underwent transthoracic regional injection of replication-defective adenovirus (2.5 x 10(9) plaque-forming units) containing either HSV1-tk (n = 16) or LacZ reporter gene (n = 15) into the inferior wall. Nonmanipulated rats (n = 5) served as further controls. Hearts were excised 2 d later and perfused according to the Langendorff technique with (124)I-FIAU-containing buffer (15 min, followed by 30 min of nonradioactive perfusion). Experiments were performed under baseline conditions and in the presence of thymidine (competitive substrate) or fludarabine (in vitro inhibitor of 5'-nucleotidase). Time-activity curves were acquired by external coincident detectors. The myocardial rate of (124)I-FIAU uptake (K(i)), clearance rate (K(o)), and volume of distribution (V(d) = K(i)/K(o)) were calculated. Subsequently, hearts were subjected to gamma-counting, followed by microtome slicing and autoradiography. RESULTS: The V(d) from Langendorff perfusion significantly correlated with final whole-heart tracer retention (r = 0.88, P = 0.019) and the autoradiographic area of regional myocardial activity (r = 0.89, P = 0.016). HSV1-tk hearts showed higher K(i) and V(d) of (124)I-FIAU compared with that of controls (P < 0.001) and detectable but slower washout compared with that of the LacZ group (P < 0.01). Addition of thymidine to the perfusate inhibited myocardial uptake of (124)I-FIAU by reducing V(d) and K(i) in HSV1-tk and LacZ hearts compared with the baseline. Addition of fludarabine did not result in the expected reduction of washout in HSV1-tk hearts due to inhibition of 5'-nucleotidases (which may dephosphorylate (124)I-FIAU monophosphate). It acted as an uptake inhibitor similar to thymidine, reducing V(d) in HSV1-tk hearts. CONCLUSION: Assessment of specific reporter probe kinetics after regional in vivo reporter gene transfer is feasible using the isolated perfused rat heart preparation. This model allows one to study the effects of pharmacologic interventions and may refine understanding of the reporter probe signal for in vivo imaging. Different nucleoside analogs significantly inhibit (124)I-FIAU uptake, emphasizing the importance of transporter mechanisms for reporter probe kinetics.