Predicting medication adherence using ensemble learning and deep learning models with large scale healthcare data.

Clinical studies from WHO have demonstrated that only 50-70% of patients adhere properly to prescribed drug therapy. Such adherence failure can impact therapeutic efficacy for the patients in question and compromises data quality around the population-level efficacy of the drug for the indications targeted. In this study, we applied various ensemble learning and deep learning models to predict medication adherence among patients. Our contribution to this endeavour involves targeting the problem of adherence prediction for a particularly challenging class of patients who self-administer injectable medication at home. Our prediction pipeline, based on event history, comprises a connected sharps bin which aims to help patients better manage their condition and improve outcomes. In other words, the efficiency of interventions can be significantly improved by prioritizing the patients who are most likely to be non-adherent. The collected data comprising a rich event feature set may be exploited for the purposes of predicting the status of the next adherence state for individual patients. This paper reports on how this concept can be realized through an investigation using a wide range of ensemble learning and deep learning models on a real-world dataset collected from such a system. The dataset investigated comprises 342,174 historic injection disposal records collected over the course of more than 5 years. A comprehensive comparison of different models is given in this paper. Moreover, we demonstrate that the selected best performer, long short-term memory (LSTM), generalizes well by deploying it in a true future testing dataset. The proposed end-to-end pipeline is capable of predicting patient failure in adhering to their therapeutic regimen with 77.35 % accuracy (Specificity: 78.28 %, Sensitivity: 76.42%, Precision: 77.87%,F1 score: 0.7714, ROC AUC: 0.8390).

Influence of key processing parameters and seeding density effects of microencapsulated chondrocytes fabricated using electrohydrodynamic spraying.

Cell delivery and leakage during injection remains a challenge for cell-based intervertebral disc regeneration strategies. Cellular microencapsulation may offer a promising approach to overcome these limitations by providing a protective niche during intradiscal injection. Electrohydrodynamic spraying (EHDS) is a versatile one-step approach for microencapsulation of cells using a high voltage electric field. The primary objective of this work was to characterise key processing parameters such as applied voltage (0, 5, 10 or 15 kV), emitter needle gauge (21, 26 or 30 G), alginate concentration (1%, 2% or 3%) and flow rate (50, 100, 250 or 500 µl min-1) to regulate the size and morphology of alginate microcapsules as well as subsequent cell viability when altering these parameters. The effect of initial cell seeding density (5, 10 and 20 × 106 cells ml-1) on subsequent matrix accumulation of microencapsulated articular chondrocytes was also evaluated. Results showed that increasing alginate concentration and thus viscosity increased overall microcapsule size but also affected the geometry towards ellipsoidal-shaped gels. Altering the electric field strength and needle diameter regulated microcapsule size towards a smaller diameter with increasing voltage and smaller needle diameter. Needle size did not appear to affect cell viability when operating with lower alginate concentrations (1% and 2%), although higher concentrations (3%) and thus higher viscosity hydrogels resulted in diminished viability with decreasing needle diameter. Increasing cell density resulted in decreased cell viability and a concomitant decrease in DNA content, perhaps due to competing nutrient demands as a result of more closely packed cells. However, higher cell densities resulted in increased levels of extracellular matrix accumulated. Overall, this work highlights the potential of EHDS as a controllable and versatile approach to fabricate microcapsules for injectable delivery which can be used in a variety of applications such as drug development or cell therapies.

Tissue Engineering Whole Bones Through Endochondral Ossification: Regenerating the Distal Phalanx.

Novel strategies are urgently required to facilitate regeneration of entire bones lost due to trauma or disease. In this study, we present a novel framework for the regeneration of whole bones by tissue engineering anatomically shaped hypertrophic cartilaginous grafts in vitro that subsequently drive endochondral bone formation in vivo. To realize this, we first fabricated molds from digitized images to generate mesenchymal stem cell-laden alginate hydrogels in the shape of different bones (the temporomandibular joint [TMJ] condyle and the distal phalanx). These constructs could be stimulated in vitro to generate anatomically shaped hypertrophic cartilaginous tissues that had begun to calcify around their periphery. Constructs were then formed into the shape of the distal phalanx to create the hypertrophic precursor of the osseous component of an engineered long bone. A layer of cartilage engineered through self-assembly of chondrocytes served as the articular surface of these constructs. Following chondrogenic priming and subcutaneous implantation, the hypertrophic phase of the engineered phalanx underwent endochondral ossification, leading to the generation of a vascularized bone integrated with a covering layer of stable articular cartilage. Furthermore, spatial bone deposition within the construct could be modulated by altering the architecture of the osseous component before implantation. These findings open up new horizons to whole limb regeneration by recapitulating key aspects of normal bone development.