ABSTRACT
Objective.Deep learning models that aid in medical image assessment tasks must be both accurate and reliable to be deployed within clinical settings. While deep learning models have been shown to be highly accurate across a variety of tasks, measures that indicate the reliability of these models are less established. Increasingly, uncertainty quantification (UQ) methods are being introduced to inform users on the reliability of model outputs. However, most existing methods cannot be augmented to previously validated models because they are not post hoc, and they change a model's output. In this work, we overcome these limitations by introducing a novel post hoc UQ method, termedLocal Gradients UQ, and demonstrate its utility for deep learning-based metastatic disease delineation.Approach.This method leverages a trained model's localized gradient space to assess sensitivities to trained model parameters. We compared the Local Gradients UQ method to non-gradient measures defined using model probability outputs. The performance of each uncertainty measure was assessed in four clinically relevant experiments: (1) response to artificially degraded image quality, (2) comparison between matched high- and low-quality clinical images, (3) false positive (FP) filtering, and (4) correspondence with physician-rated disease likelihood.Main results.(1) Response to artificially degraded image quality was enhanced by the Local Gradients UQ method, where the median percent difference between matching lesions in non-degraded and most degraded images was consistently higher for the Local Gradients uncertainty measure than the non-gradient uncertainty measures (e.g. 62.35% vs. 2.16% for additive Gaussian noise). (2) The Local Gradients UQ measure responded better to high- and low-quality clinical images (p< 0.05 vsp> 0.1 for both non-gradient uncertainty measures). (3) FP filtering performance was enhanced by the Local Gradients UQ method when compared to the non-gradient methods, increasing the area under the receiver operating characteristic curve (ROC AUC) by 20.1% and decreasing the false positive rate by 26%. (4) The Local Gradients UQ method also showed more favorable correspondence with physician-rated likelihood for malignant lesions by increasing ROC AUC for correspondence with physician-rated disease likelihood by 16.2%.Significance. In summary, this work introduces and validates a novel gradient-based UQ method for deep learning-based medical image assessments to enhance user trust when using deployed clinical models.
Subject(s)
Deep Learning , Image Processing, Computer-Assisted , Uncertainty , Humans , Image Processing, Computer-Assisted/methodsABSTRACT
Objective. Manual analysis of individual cancer lesions to assess disease response is clinically impractical and requires automated lesion tracking methodologies. However, no methodology has been developed for whole-body individual lesion tracking, across an arbitrary number of scans, and acquired with various imaging modalities.Approach. This study introduces a lesion tracking methodology and benchmarked it using 2368Ga-DOTATATE PET/CT and PET/MR images of eight neuroendocrine tumor patients. The methodology consists of six steps: (1) alignment of multiple scans via image registration, (2) body-part labeling, (3) automatic lesion-wise dilation, (4) clustering of lesions based on local lesion shape metrics, (5) assignment of lesion tracks, and (6) output of a lesion graph. Registration performance was evaluated via landmark distance, lesion matching accuracy was evaluated between each image pair, and lesion tracking accuracy was evaluated via identical track ratio. Sensitivity studies were performed to evaluate the impact of lesion dilation (fixed versus automatic dilation), anatomic location, image modalities (inter- versus intra-modality), registration mode (direct versus indirect registration), and track size (number of time-points and lesions) on lesion matching and tracking performance.Main results. Manual contouring yielded 956 lesions, 1570 lesion-matching decisions, and 493 lesion tracks. The median residual registration error was 2.5 mm. The automatic lesion dilation led to 0.90 overall lesion matching accuracy, and an 88% identical track ratio. The methodology is robust regarding anatomic locations, image modalities, and registration modes. The number of scans had a moderate negative impact on the identical track ratio (94% for 2 scans, 91% for 3 scans, and 81% for 4 scans). The number of lesions substantially impacted the identical track ratio (93% for 2 nodes versus 54% for ≥5 nodes).Significance. The developed methodology resulted in high lesion-matching accuracy and enables automated lesion tracking in PET/CT and PET/MR.
Subject(s)
Neuroendocrine Tumors , Positron Emission Tomography Computed Tomography , Humans , Tomography, X-Ray Computed/methods , Multimodal Imaging/methods , Positron-Emission Tomography/methods , Neuroendocrine Tumors/diagnostic imaging , Magnetic Resonance Imaging/methodsABSTRACT
Objective. Neuroimaging uncovers important information about disease in the brain. Yet in Alzheimer's disease (AD), there remains a clear clinical need for reliable tools to extract diagnoses from neuroimages. Significant work has been done to develop deep learning (DL) networks using neuroimaging for AD diagnosis. However, no particular model has emerged as optimal. Due to a lack of direct comparisons and evaluations on independent data, there is no consensus on which modality is best for diagnostic models or whether longitudinal information enhances performance. The purpose of this work was (1) to develop a generalizable DL model to distinguish neuroimaging scans of AD patients from controls and (2) to evaluate the influence of imaging modality and longitudinal data on performance.Approach. We trained a 2-class convolutional neural network (CNN) with and without a cascaded recurrent neural network (RNN). We used datasets of 772 (NAD = 364,Ncontrol= 408) 3D18F-FDG PET scans and 780 (NAD = 280,Ncontrol= 500) T1-weighted volumetric-3D MR images (containing 131 and 144 patients with multiple timepoints) from the Alzheimer's Disease Neuroimaging Initiative, plus an independent set of 104 (NAD = 63,NNC = 41)18F-FDG PET scans (one per patient) for validation.Main Results. ROC analysis showed that PET-trained models outperformed MRI-trained, achieving maximum AUC with the CNN + RNN model of 0.93 ± 0.08, with accuracy 82.5 ± 8.9%. Adding longitudinal information offered significant improvement to performance on18F-FDG PET, but not on T1-MRI. CNN model validation with an independent18F-FDG PET dataset achieved AUC of 0.99. Layer-wise relevance propagation heatmaps added CNN interpretability.Significance. The development of a high-performing tool for AD diagnosis, with the direct evaluation of key influences, reveals the advantage of using18F-FDG PET and longitudinal data over MRI and single timepoint analysis. This has significant implications for the potential of neuroimaging for future research on AD diagnosis and clinical management of suspected AD patients.
Subject(s)
Alzheimer Disease , Deep Learning , Alzheimer Disease/diagnostic imaging , Fluorodeoxyglucose F18 , Humans , Magnetic Resonance Imaging/methods , Neuroimaging/methods , Positron-Emission Tomography/methodsABSTRACT
Protein detection is a universal tool critical to many applications in medicine, agriculture, and biotechnology. We developed a novel protein detection method combining light transmission spectroscopy and particle-size analysis of gold nanospheres monovalently functionalized with polyclonal antibodies and applied it to an emerging challenge for such technologiesâthe monitoring of environmental proteins (eProteins) present in natural aquatic systems. These are an underreported source of pollution and include the pseudopersistent Cry toxins that enter aquatic ecosystems from surrounding genetically engineered crops. The assay is capable of detecting proteins in complex matrices, such as water samples collected in the field, making it a competitive assay for eProtein detection. It is sensitive, reaching 1.25 ng mL-1, and we demonstrate its application to the detection of Cry1Ab from subsurface tile-drain and streamwater samples from agricultural waterways. The assay can also be quickly adapted for other protein detection applications in the future.
Subject(s)
Gold , Metal Nanoparticles , Bacterial Proteins/genetics , Ecosystem , Gold/chemistry , Hemolysin Proteins/analysis , Metal Nanoparticles/chemistry , Plants, Genetically Modified/chemistry , Plants, Genetically Modified/metabolism , Spectrum AnalysisABSTRACT
Rapid, sensitive, and quantitative protein detection is critical for many applications in medicine, environmental monitoring, and the food industry. Advancements in detection of proteins include the use of antigen-antibody binding; however, many current methods are time-consuming and have limiting factors such as low sensitivity and the inability to provide absolute values. We present a new high-throughput method for protein detection using light transmission spectroscopy (LTS), which can quantify and size nanoparticles in fluid suspension. LTS can quantify proteins directly and target specific proteins through antigen-antibody binding. This work shows that LTS can distinguish between and quantify bovine serum albumin, its antibody, and the BSA + Ab complex and determine BSA protein concentrations down to 5 µg/mL. We use both Mie and discrete dipole approximation models to provide geometric insight into the binding process.
ABSTRACT
This paper describes light transmission spectroscopy (LTS), a technique for eliminating spectral noise and systematic effects in real-time spectroscopic measurements. In our work, we combine LTS with spectral inversion for the purpose of nanoparticle analysis. This work employs a wideband multi-wavelength light source and grating spectrometers coupled to CCD detectors. The light source ranges from 210 to 2000 nm, the wavelength-dependent light detection system ranges from 200 to 1100 nm with ≤1 nm resolution, and the nanoparticle diameters range from 1 to 3000 nm. The nanoparticles are suspended in pure water or water-based buffer solutions. For testing and calibration purposes, results are presented for nanoparticles composed of polystyrene and gold. Mie theory is used to model the total extinction cross section, and spectral inversion is employed to obtain quantitative particle size distributions, from which information on the size, shape, and number of nanoparticles can be derived. Discussed are the precision, accuracy, resolution, and sensitivity of our results. The LTS technique is quite versatile and can be applied to spectroscopic investigations where wideband, accurate, low-noise, real-time spectra are desired.
ABSTRACT
Functionalized magnetic microspheres are widely used for cell separations, isolation of proteins and other biomolecules, in vitro diagnostics, tissue engineering, and microscale force spectroscopy. We present here the synthesis and characterization of a silicone magnetic microsphere which can be produced in diameters ranging from 0.5 to 50 µm via emulsion polymerization of a silicone ferrofluid precursor. This bottom-up approach to synthesis ensures a uniform magnetic concentration across all sizes, leading to significant advances in magnetic force generation. We demonstrate that in a size range of 5-20 µm, these spheres supply a full order of magnitude greater magnetic force than leading commercial products. In addition, the unique silicone matrix exhibits autofluorescence two orders of magnitude lower than polystyrene microspheres. Finally, we demonstrate the ability to chemically functionalize our silicone microspheres using a standard EDC reaction, and show that our folate-functionalized silicone microspheres specifically bind to targeted HeLa and Jurkat cells. These spheres show tremendous potential for replacing magnetic polystyrene spheres in applications which require either large magnetic forces or minimal autofluorescence, since they represent order-of-magnitude improvements in each. In addition, the unique silicone matrix and proven biocompatibility suggest that they may be useful for encapsulation and targeted delivery of lipophilic pharmaceuticals.