Improving the diagnostic accuracy for major depressive disorder using machine learning algorithms integrating clinical and near-infrared spectroscopy data.

BACKGROUND: Given that major depressive disorder (MDD) is both biologically and clinically heterogeneous, a diagnostic system integrating neurobiological markers and clinical characteristics would allow for better diagnostic accuracy and, consequently, treatment efficacy. OBJECTIVE: Our study aimed to evaluate the discriminative and predictive ability of unimodal, bimodal, and multimodal approaches in a total of seven machine learning (ML) models-clinical, demographic, functional near-infrared spectroscopy (fNIRS), combinations of two unimodal models, as well as a combination of all three-for MDD. METHODS: We recruited 65 adults with MDD and 68 matched healthy controls, who provided both sociodemographic and clinical information, and completed the HAM-D questionnaire. They were also subject to fNIRS measurement when participating in the verbal fluency task. Using the nested cross validation procedure, the classification performance of each ML model was evaluated based on the area under the receiver operating characteristic curve (ROC), balanced accuracy, sensitivity, and specificity. RESULTS: The multimodal ML model was able to distinguish between depressed patients and healthy controls with the highest balanced accuracy of 87.98 ± 8.84% (AUC = 0.92; 95% CI (0.84-0.99) when compared with the uni- and bi-modal models. CONCLUSIONS: Our multimodal ML model demonstrated the highest diagnostic accuracy for MDD. This reinforces the biological and clinical heterogeneity of MDD and highlights the potential of this model to improve MDD diagnosis rates. Furthermore, this model is cost-effective and clinically applicable enough to be established as a robust diagnostic system for MDD based on patients' biosignatures.

Deep learning architectures analysis for age-related macular degeneration segmentation on optical coherence tomography scans.

BACKGROUND AND OBJECTIVES: Aged people usually are more to be diagnosed with retinal diseases in developed countries. Retinal capillaries leakage into the retina swells and causes an acute vision loss, which is called age-related macular degeneration (AMD). The disease can not be adequately diagnosed solely using fundus images as depth information is not available. The variations in retina volume assist in monitoring ophthalmological abnormalities. Therefore, high-fidelity AMD segmentation in optical coherence tomography (OCT) imaging modality has raised the attention of researchers as well as those of the medical doctors. Many methods across the years encompassing machine learning approaches and convolutional neural networks (CNN) strategies have been proposed for object detection and image segmentation. METHODS: In this paper, we analyze four wide-spread deep learning models designed for the segmentation of three retinal fluids outputting dense predictions in the RETOUCH challenge data. We aim to demonstrate how a patch-based approach could push the performance for each method. Besides, we also evaluate the methods using the OPTIMA challenge dataset for generalizing network performance. The analysis is driven into two sections: the comparison between the four approaches and the significance of patching the images. RESULTS: The performance of networks trained on the RETOUCH dataset is higher than human performance. The analysis further generalized the performance of the best network obtained by fine-tuning it and achieved a mean Dice similarity coefficient (DSC) of 0.85. Out of the three types of fluids, intraretinal fluid (IRF) is more recognized, and the highest DSC value of 0.922 is achieved using Spectralis dataset. Additionally, the highest average DSC score is 0.84, which is achieved by PaDeeplabv3+ model using Cirrus dataset. CONCLUSIONS: The proposed method segments the three fluids in the retina with high DSC value. Fine-tuning the networks trained on the RETOUCH dataset makes the network perform better and faster than training from scratch. Enriching the networks with inputting a variety of shapes by extracting patches helped to segment the fluids better than using a full image.

N-3 fatty acids inhibit vascular calcification via the p38-mitogen-activated protein kinase and peroxisome proliferator-activated receptor-gamma pathways.

Fish oil supplementation is associated with lower risk of coronary artery disease in humans, and it has been shown to reduce ectopic calcification in an animal model. However, whether N-3 fatty acids, active ingredients of fish oil, have direct effects on calcification of vascular cells is not clear. In this report, we investigated the effects of eicosapentaenoic acid and docosahexaenoic acid (DHA) on osteoblastic differentiation and mineralization of calcifying vascular cells (CVCs), a subpopulation of bovine aortic medial cells that undergo osteoblastic differentiation and form calcified matrix in vitro. Results showed that N-3 fatty acids inhibited alkaline phosphatase (ALP) activity and mineralization of vascular cells, suggesting that they directly affect osteoblastic differentiation in vascular cells. By Western blot analysis, DHA activated p38-mitogen-activated protein kinase (MAPK) but not extracellular-regulated kinase (ERK) or Akt. An inhibitor of p38-MAPK partially reversed the inhibitory effects of DHA on osteoblastic differentiation and mineralization. Transient transfection experiments showed that DHA also activated peroxisome proliferator-activated receptor-gamma (PPAR-gamma). Both p38-MAPK activator and PPAR-gamma agonists reproduced the inhibitory effects of DHA on CVC mineralization. Pretreatment with DHA also inhibited interleukin-6-induced ALP activity and mineralization. Together, these results suggest that N-3 fatty acids directly inhibit vascular calcification, and that the inhibitory effects are mediated by the p38-MAPK and PPAR-gamma pathways.

Fusing Results of Several Deep Learning Architectures for Automatic Classification of Normal and Diabetic Macular Edema in Optical Coherence Tomography.

Diabetic Macular Edema (DME) is a severe eye disease that can lead to irreversible blindness if it is left untreated. DME diagnosis still relies on manual evaluation from opthalmologists, thus the process is time consuming and diagnosis may be subjective. This paper presents two novel DME detection frameworks: (1) combining features from three pre-trained Convolutional Neural Networks: AlexNet, VggNet and GoogleNet and performing feature space reduction using Principal Component Analysis and (2) a majority voting scheme based on a plurality rule between classifications from AlexNet, VggNet and GoogleNet. Experiments were conducted using Optical Coherence Tomography datasets retrieved from the Singapore Eye Research Institute and the Chinese University Hong Kong. The results are evaluated using a Leave-Two-Patients-Out Cross Validation at the volume level. This method improves DME classification with an accuracy of 93.75%, which is similar to the best algorithms so far on the same data sets.

Implementation of wireless power transfer and communications for an implantable ocular drug delivery system.

A wireless power transfer and communication system based on near-field inductive coupling has been designed and implemented. The feasibility of using such a system to remotely control drug release from an implantable drug delivery system is addressed. The architecture of the wireless system is described and the signal attenuation over distance in both water and phosphate buffered saline is studied. Additionally, the health risk due to exposure to radio frequency (RF) radiation is examined using a biological model. The experimental results demonstrate that the system can trigger the release of drug within 5 s, and that such short exposure to RF radiation does not produce any significant (

Development of a miniaturised drug delivery system with wireless power transfer and communication.

The development of an implantable system designed to deliver drug doses in a controlled manner over an extended time period is reported. Key performance parameters are the physical size, the power consumption and also the ability to perform wireless communications to enable the system to be externally controlled and interrogated. The system has been designed to facilitate wireless power transfer, which is very important for miniaturisation as it removes the need for a battery.

Adaptive, integrated sensor processing to compensate for drift and uncertainty: a stochastic 'neural' approach.

An adaptive stochastic classifier based on a simple, novel neural architecture--the Continuous Restricted Boltzmann Machine (CRBM) is demonstrated. Together with sensors and signal conditioning circuits, the classifier is capable of measuring and classifying (with high accuracy) the H+ ion concentration, in the presence of both random noise and sensor drift. Training on-line, the stochastic classifier is able to overcome significant drift of real incomplete sensor data dynamically. As analogue hardware, this signal-level sensor fusion scheme is therefore suitable for real-time analysis in a miniaturised multisensor microsystem such as a Lab-in-a-Pill (LIAP).