The development of an automated machine learning pipeline for the detection of Alzheimer's Disease.

Although Alzheimer's disease is the most prevalent form of dementia, there are no treatments capable of slowing disease progression. A lack of reliable disease endpoints and/or biomarkers contributes in part to the absence of effective therapies. Using machine learning to analyze EEG offers a possible solution to overcome many of the limitations of current diagnostic modalities. Here we develop a logistic regression model with an accuracy of 81% that addresses many of the shortcomings of previous works. To our knowledge, no other study has been able to solve the following problems simultaneously: (1) a lack of automation and unbiased removal of artifacts, (2) a dependence on a high level of expertise in data pre-processing and ML for non-automated processes, (3) the need for very large sample sizes and accurate EEG source localization using high density systems, (4) and a reliance on black box ML approaches such as deep neural nets with unexplainable feature selection. This study presents a proof-of-concept for an automated and scalable technology that could potentially be used to diagnose AD in clinical settings as an adjunct to conventional neuropsychological testing, thus enhancing efficiency, reproducibility, and practicality of AD diagnosis.

Synthesis of fracture radiographs with deep neural networks.

PURPOSE: We describe a machine learning system for converting diagrams of fractures into realistic X-ray images. We further present a method for iterative, human-guided refinement of the generated images and show that the resulting synthetic images can be used during training to increase the accuracy of deep classifiers on clinically meaningful subsets of fracture X-rays. METHODS: A neural network was trained to reconstruct images from programmatically created line drawings of those images. The images were then further refined with an optimization-based technique. Ten physicians were recruited into a study to assess the realism of synthetic radiographs created by the neural network. They were presented with mixed sets of real and synthetic images and asked to identify which images were synthetic. Two classifiers were trained to detect humeral shaft fractures: one only on true fracture images, and one on both true and synthetic images. RESULTS: Physicians were 49.63% accurate in identifying whether images were synthetic or real. This is close to what would be expected by pure chance (i.e. random guessing). A classifier trained only on real images detected fractures with 67.21% sensitivity when no fracture fixation hardware was present. A classifier trained on both real images and synthetic images was 75.54% sensitive. CONCLUSION: Our method generates X-rays realistic enough to be indistinguishable from real X-rays. We also show that synthetic images generated using this method can be used to increase the accuracy of deep classifiers on clinically meaningful subsets of fracture X-rays.

A Novel, Adaptable Laryngeal Mask to Facilitate a Percutaneous Dilatational Tracheostomy: Proof-of-Concept Prototype Demonstration on a Mannequin Model and Cadaver.

BACKGROUND: Most percutaneous dilatational tracheostomy (PDT) mortalities result from airway-related complications. Improved airway pressure management and gas delivery are targets for innovation. This study describes an adaptable laryngeal mask (ALM) designed to remove the bronchoscope from the endotracheal tube (ETT) and place it in a separate lumen. Airflow and device efficacy were evaluated during PDTs with an ALM on mannequins and cadavers, respectively. METHODS: Procedures were completed by a single physician using an 8.0 mm ETT and the Ciaglia Blue Rhino method on simulation mannequins (TruCorp AirSim Traci) and fresh-frozen cadavers. Mannequin simulation tested the respiratory capabilities of an ALM utilizing a BioPac spirometer and a Maquet Servo ventilator. Qualitative analysis on device efficacy was performed on 2 fresh-frozen cadavers (1 male, 1 female). RESULTS: Preliminary ventilation testing on a PDT-able mannequin using the ALM showed an increase in airflow reaching the lungs compared with a deflated ETT. During mannequin and cadaver testing, the ALM was placed over the in situ ETT effectively, thereby removing the bronchoscope from the ETT while maintaining a continuous visual of the incision site. Both mannequin and cadaveric testing using an ALM enabled a single physician to safely perform the PDT procedure with minimal assistance. CONCLUSIONS: Initial testing using an ALM during PDT on mannequins and cadavers showed an improvement in airflow and the removal of the bronchoscope from the ETT, respectively. Further studies using the ALM in a patient population compared with standard techniques would be useful.

Enhancing Tabletop X-Ray Phase Contrast Imaging with Nano-Fabrication.

X-ray phase-contrast imaging is a promising approach for improving soft-tissue contrast and lowering radiation dose in biomedical applications. While current tabletop imaging systems adapt to common x-ray tubes and large-area detectors by employing absorptive elements such as absorption gratings or monolithic crystals to filter the beam, we developed nanometric phase gratings which enable tabletop x-ray far-field interferometry with only phase-shifting elements, leading to a substantial enhancement in the performance of phase contrast imaging. In a general sense the method transfers the demands on the spatial coherence of the x-ray source and the detector resolution to the feature size of x-ray phase masks. We demonstrate its capabilities in hard x-ray imaging experiments at a fraction of clinical dose levels and present comparisons with the existing Talbot-Lau interferometer and with conventional digital radiography.

Electrodeposition of Gold to Conformally Fill High Aspect Ratio Nanometric Silicon Grating Trenches: A Comparison of Pulsed and Direct Current Protocols.

Filling high-aspect-ratio trenches with gold is a frequent requirement in the fabrication of x-ray optics as well as micro-electronic components and other fabrication processes. Conformal electrodeposition of gold in sub-micron-width silicon trenches with an aspect ratio greater than 35 over a grating area of several square centimeters is challenging and has not been described in the literature previously. A comparison of pulsed plating and constant current plating led to a gold electroplating protocol that reliably filled trenches for such structures.

Fabrication of 200 nm period hard X-ray phase gratings.

Far field X-ray grating interferometry achieves extraordinary phase sensitivity in imaging weakly absorbing samples, provided that the grating period is within the transverse coherence length of the X-ray source. Here we describe a cost-efficient process to fabricate large area, 100 nm half-pitch hard X-ray phase gratings with an aspect ratio of 32. The nanometric gratings are suitable for ordinary compact X-ray sources having low spatial coherence, as demonstrated by X-ray diffraction experiments.