Near-pair patch generative adversarial network for data augmentation of focal pathology object detection models.

Purpose: The limited volume of medical training data remains one of the leading challenges for machine learning for diagnostic applications. Object detectors that identify and localize pathologies require training with a large volume of labeled images, which are often expensive and time-consuming to curate. To reduce this challenge, we present a method to support distant supervision of object detectors through generation of synthetic pathology-present labeled images. Approach: Our method employs the previously proposed cyclic generative adversarial network (cycleGAN) with two key innovations: (1) use of "near-pair" pathology-present regions and pathology-absent regions from similar locations in the same subject for training and (2) the addition of a realism metric (Fréchet inception distance) to the generator loss term. We trained and tested this method with 2800 fracture-present and 2800 fracture-absent image patches from 704 unique pediatric chest radiographs. The trained model was then used to generate synthetic pathology-present images with exact knowledge of location (labels) of the pathology. These synthetic images provided an augmented training set for an object detector. Results: In an observer study, four pediatric radiologists used a five-point Likert scale indicating the likelihood of a real fracture (1 = definitely not a fracture and 5 = definitely a fracture) to grade a set of real fracture-absent, real fracture-present, and synthetic fracture-present images. The real fracture-absent images scored 1.7±1.0, real fracture-present images 4.1±1.2, and synthetic fracture-present images 2.5±1.2. An object detector model (YOLOv5) trained on a mix of 500 real and 500 synthetic radiographs performed with a recall of 0.57±0.05 and an F2 score of 0.59±0.05. In comparison, when trained on only 500 real radiographs, the recall and F2 score were 0.49±0.06 and 0.53±0.06, respectively. Conclusions: Our proposed method generates visually realistic pathology and that provided improved object detector performance for the task of rib fracture detection.

Incorporating Tantalum Oxide Nanoparticles into Implantable Polymeric Biomedical Devices for Radiological Monitoring.

Longitudinal radiological monitoring of biomedical devices is increasingly important, driven by the risk of device failure following implantation. Polymeric devices are poorly visualized with clinical imaging, hampering efforts to use diagnostic imaging to predict failure and enable intervention. Introducing nanoparticle contrast agents into polymers is a potential method for creating radiopaque materials that can be monitored via computed tomography. However, the properties of composites may be altered with nanoparticle addition, jeopardizing device functionality. Thus, the material and biomechanical responses of model nanoparticle-doped biomedical devices (phantoms), created from 0-40 wt% tantalum oxide (TaOx ) nanoparticles in polycaprolactone and poly(lactide-co-glycolide) 85:15 and 50:50, representing non, slow, and fast degrading systems, respectively, are investigated. Phantoms degrade over 20 weeks in vitro in simulated physiological environments: healthy tissue (pH 7.4), inflammation (pH 6.5), and lysosomal conditions (pH 5.5), while radiopacity, structural stability, mechanical strength, and mass loss are monitored. The polymer matrix determines overall degradation kinetics, which increases with lower pH and higher TaOx content. Importantly, all radiopaque phantoms could be monitored for a full 20 weeks. Phantoms implanted in vivo and serially imaged demonstrate similar results. An optimal range of 5-20 wt% TaOx nanoparticles balances radiopacity requirements with implant properties, facilitating next-generation biomedical devices.

Incorporating Radiopacity into Implantable Polymeric Biomedical Devices for Clinical Radiological Monitoring.

Longitudinal radiological monitoring of biomedical devices is increasingly important, driven by risk of device failure following implantation. Polymeric devices are poorly visualized with clinical imaging, hampering efforts to use diagnostic imaging to predict failure and enable intervention. Introducing nanoparticle contrast agents into polymers is a potential method for creating radiopaque materials that can be monitored via computed tomography. However, properties of composites may be altered with nanoparticle addition, jeopardizing device functionality. This, we investigated material and biomechanical response of model nanoparticle-doped biomedical devices (phantoms), created from 0-40wt% TaO x nanoparticles in polycaprolactone, poly(lactide-co-glycolide) 85:15 and 50:50, representing non-, slow and fast degrading systems, respectively. Phantoms degraded over 20 weeks in vitro, in simulated physiological environments: healthy tissue (pH 7.4), inflammation (pH 6.5), and lysosomal conditions (pH 5.5), while radiopacity, structural stability, mechanical strength and mass loss were monitored. The polymer matrix determined overall degradation kinetics, which increased with lower pH and higher TaO x content. Importantly, all radiopaque phantoms could be monitored for a full 20-weeks. Phantoms implanted in vivo and serially imaged, demonstrated similar results. An optimal range of 5-20wt% TaO x nanoparticles balanced radiopacity requirements with implant properties, facilitating next-generation biomedical devices.

Determining Carina and Clavicular Distance-Dependent Positioning of Endotracheal Tube in Critically Ill Patients: An Artificial Intelligence-Based Approach.

Early and accurate prediction of endotracheal tube (ETT) location is pivotal for critically ill patients. Automatic and timely detection of faulty ETT locations from chest X-ray images may avert patients' morbidity and mortality. Therefore, we designed convolutional neural network (CNN)-based algorithms to evaluate ETT position appropriateness relative to four detected key points, including tracheal tube end, carina, and left/right clavicular heads on chest radiographs. We estimated distances from the tube end to tracheal carina and the midpoint of clavicular heads. A DenseNet121 encoder transformed images into embedding features, and a CNN-based decoder generated the probability distributions. Based on four sets of tube-to-carina distance-dependent parameters (i.e., (i) 30-70 mm, (ii) 30-60 mm, (iii) 20-60 mm, and (iv) 20-55 mm), corresponding models were generated, and their accuracy was evaluated through the predicted L1 distance to ground-truth coordinates. Based on tube-to-carina and tube-to-clavicle distances, the highest sensitivity, and specificity of 92.85% and 84.62% respectively, were revealed for 20-55 mm. This implies that tube-to-carina distance between 20 and 55 mm is optimal for an AI-based key point appropriateness detection system and is empirically comparable to physicians' consensus.

Comparison of Colorimetric Analyses to Determine Cortisol in Human Sweat.

Colorimetric analysis, which relies on a chemical reaction to facilitate a change in visible color, is a great strategy for detecting cortisol, which is necessary to diagnose and manage the wide variety of diseases related to the hormone, because it is simple in design, inexpensive, and reliable as a standard cortisol analysis technique. In this study, four different colorimetric cortisol analyses that use various chromogens, which include sulfuric acid, Porter-Silber reagent, Prussian blue, and blue tetrazolium, are studied. Modifications to the classic Porter-Silber method are made by increasing the carbon content of the alcohol and adding gold nanoparticles, which result in a twofold increase in reaction rate and a slight decrease in the limit of detection (LoD). After a comparison of the reaction rate, LoD, dynamic range, characteristic peaks, and color stability of all methods, blue tetrazolium demonstrates a low LoD (97 ng/mL), broad dynamic range (0.05-2 µg/mL), and quick reaction rate (color development as fast as 10 min), which are well within the requirements for human biofluids. Cortisol in artificial saliva and sweat and in human sweat was determined while confirming that no excipients or other biomarkers interfered with the reactions. Twenty-one human sweat samples were tested using blue tetrazolium and revealed a significant difference between male and female apocrine cortisol concentrations and showed a highly significant difference between apocrine and eccrine cortisol concentrations. Colorimetric methods of cortisol can compete with existing electrochemical sensors because of their similar accuracy and detection range in certain wearable biosensor applications. The simplicity of colorimetric methods advances potential applications in skin-interfaced bio-electronics and point-of-care devices.