Using AI/ML to predict blending performance and process sensitivity for Continuous Direct Compression (CDC)

Utilising three artificial intelligence (AI)/machine learning (ML) tools, this study explores the prediction of fill level in inclined linear blenders at steady state by mapping a wide range of bulk powder characteristics to processing parameters. Predicting fill levels enables the calculation of blade passes (strain), known from existing literature to enhance content uniformity. We present and train three AI/ML models, each demonstrating unique predictive capabilities for fill level. These models collectively identify the following rank order of feature importance: RPM, Mixing Blade Region (MB) size, Wall Friction Angle (WFA), and Feed Rate (FR). Random Forest Regression, a machine learning algorithm that constructs a multitude of decision trees at training time and outputs the mode of the classes (classification) or mean prediction (regression) of the individual trees, develops a series of individually useful decision trees. but also allows the extraction of logic and breakpoints within the data. A novel tool which utilises smart optimisation and symbolic regression to model complex systems into simple, closed-form equations, is used to build an accurate reduced-order model. Finally, an Artificial Neural Network (ANN), though less interrogable emerges as the most accurate fill level predictor, with an r2 value of 0.97. Following training on single-component mixtures, the models are tested with a four-component powdered paracetamol formulation, mimicking an existing commercial drug product. The ANN predicts the fill level of this formulation at three RPMs (250, 350 and 450) with a mean absolute error of 1.4%. Ultimately, the modelling tools showcase a framework to better understand the interaction between process and formulation. The result of this allows for a first-time-right approach for formulation development whilst gaining process understanding from fewer experiments. Resulting in the ability to approach risk during product development whilst gaining a greater holistic understanding of the processing environment of the desired formulation.

Successful Formulation Window for the design of pharmaceutical tablets with required mechanical properties

Pharmaceutical tablet formulations combine the active ingredient with processing aids and functional components. This paper evaluates compressibility based predictive models for binary and ternary formulations to establish an acceptable range of tablet compression parameters that satisfy prescribed quality target criteria for tablets including minimum tablet strength and processing constraints such as maximum ejection stress and maximum compaction pressure. The concept of Successful Formulation Window (SFW) is introduced. A methodology is proposed to determine the SFW for a given formulation based on compaction simulator data collected for individual formulation components. The methodology is validated for binary and ternary mixtures and lubricated formulations. The SFW analysis was developed to support tablet formulation design to meet mechanical requirements.

Application of Positron Emission Particle Tracking (PEPT) for the evaluation of powder behaviour in an incline linear blender for Continuous Direct Compression (CDC)

Positron Emission Particle Tracking (PEPT) is a non-invasive measurement technique which offers the ability to track the motion of individual particles with high temporal and spatial resolution, and thus build up an understanding of the bulk behaviour of a system from its microscopic (particle level) dynamics. Using this measurement technique, we have developed a series of novel metrics to better understand the behaviours of powders during the steady-state operation of a continuous blender system. Results are presented concerning the response of particle motion to processing parameters (mixing blade configuration and RPM), quantifying the motion in terms of predicted mixing performance. It was found that both increasing rpm and increasing hold-up mass (by selecting fewer transport blades and more mixing blades) provided improved mixing conditions. Interestingly, under specific conditions, there is evidence of convection-like mixing occurring at the interface of the transport and mixing region. This suggests the existence of a potential 'folding region' whereby powder is transported up the barrel (and away from the powder bulk bed) before being reconstituted back into the bulk mass. The results also provide valuable experimental data for the development, calibration and validation of future Discrete Element Method (DEM) simulations.

Acoustic emission monitoring from a lab scale high shear granulator--a novel approach

A new approach to the monitoring of granulation processes using passive acoustics together with precise control over the granulation process has highlighted the importance of particle-particle and particle-bowl collisions in acoustic emission. The results have shown that repeatable acoustic results could be obtained but only when a spray nozzle water addition system was used. Acoustic emissions were recorded from a transducer attached to the bowl and an airborne transducer. It was found that the airborne transducer detected very little from the granulation and only experienced small changes throughout the process. The results from the bowl transducer showed that during granulation the frequency content of the acoustic emission shifted towards the lower frequencies. Results from the discrete element model indicate that when larger particles are used the number of collisions the particles experience reduces. This is a result of the volume conservation methodology used in this study, therefore larger particles results in less particles. These simulation results coupled with previous theoretical work on the frequency content of an impacting sphere explain why the frequency content of the acoustic emissions reduces during granule growth. The acoustic system used was also clearly able to identify when large over-wetted granules were present in the system, highlighting its benefit for detecting undesirable operational conditions. High-speed photography was used to study if visual changes in the granule properties could be linked with the changing acoustic emissions. The high speed photography was only possible towards the latter stages of the granulation process and it was found that larger granules produced a higher magnitude of acoustic emission across a broader frequency range.

Comorbidity of postural orthostatic tachycardia syndrome and chronic fatigue syndrome in an Australian cohort

OBJECTIVE

Patients with chronic fatigue syndrome (CFS) are frequently diagnosed with comorbid postural orthostatic tachycardia syndrome (POTS), suggesting a shared pathogenesis. The aim of this study was to examine the relationship between demographic characteristics, autonomic functioning and fatigue levels amongst CFS patients with and without comorbid POTS.

DESIGN AND SETTING

All patients presenting to the CFS Discovery Clinic between 2009 and 2012 completed a 20-min standing task as part of their initial assessment. Heart rate and pulse pressure were recorded at baseline, at 2-min intervals poststanding, at the end of the task and following a recovery period. Average heart rate and pulse pressure variability were calculated from this data. Age, gender, length of illness and self-reported fatigue scores were also recorded. POTS patients were diagnosed by an orthostatic increase in heart rate >30 beats per min, concomitant symptoms of orthostatic intolerance and no orthostatic hypotension. Differences in autonomic functioning between POTS and CFS patients were compared using independent samples t-tests, whilst logistic and linear regressions were performed to examine the contribution of autonomic functioning to task completion and perceived fatigue, respectively.

RESULTS

Comorbidity of CFS and POTS (CFS-POTS) was observed in 11% (33/306) of patients. CFS-POTS patients were significantly younger (P < 0.001), had a shorter length of illness (P = 0.034), experienced greater task difficulty (P = 0.002) and were able to stand for significantly shorter periods compared to the CFS-only patients (P < 0.001). CFS-POTS patients experienced significantly lower baseline diastolic blood pressure (P = 0.002), significantly higher heart rate and lower pulse pressures at each standing measurement. Early heart rate changes (P = 0.002) and overall heart rate change (P < 0.001) were significant predictors of completion status, whereas heart rate variability (P < 0.001) and female gender (P < 0.001) were significant predictors of increased perceived task difficulty.

CONCLUSIONS

Haemodynamic and demographic differences between CFS-POTS and CFS-only patients suggest that the former group reflects a distinct subgroup of the CFS population. The findings highlight the utility of screening younger patients with fatigue for POTS, and identified heart rate variability as an important marker of fatigue for CFS patients in general.