Deep learning for FTIR histology: leveraging spatial and spectral features with convolutional neural networks.

Current methods for cancer detection rely on tissue biopsy, chemical labeling/staining, and examination of the tissue by a pathologist. Though these methods continue to remain the gold standard, they are non-quantitative and susceptible to human error. Fourier transform infrared (FTIR) spectroscopic imaging has shown potential as a quantitative alternative to traditional histology. However, identification of histological components requires reliable classification based on molecular spectra, which are susceptible to artifacts introduced by noise and scattering. Several tissue types, particularly in heterogeneous tissue regions, tend to confound traditional classification methods. Convolutional neural networks (CNNs) are the current state-of-the-art in image classification, providing the ability to learn spatial characteristics of images. In this paper, we demonstrate that CNNs with architectures designed to process both spectral and spatial information can significantly improve classifier performance over per-pixel spectral classification. We report classification results after applying CNNs to data from tissue microarrays (TMAs) to identify six major cellular and acellular constituents of tissue, namely adipocytes, blood, collagen, epithelium, necrosis, and myofibroblasts. Experimental results show that the use of spatial information in addition to the spectral information brings significant improvements in the classifier performance and allows classification of cellular subtypes, such as adipocytes, that exhibit minimal chemical information but have distinct spatial characteristics. This work demonstrates the application and efficiency of deep learning algorithms in improving the diagnostic techniques in clinical and research activities related to cancer.

Time for a full digital approach in nephropathology: a systematic review of current artificial intelligence applications and future directions.

INTRODUCTION: Artificial intelligence (AI) integration in nephropathology has been growing rapidly in recent years, facing several challenges including the wide range of histological techniques used, the low occurrence of certain diseases, and the need for data sharing. This narrative review retraces the history of AI in nephropathology and provides insights into potential future developments. METHODS: Electronic searches in PubMed-MEDLINE and Embase were made to extract pertinent articles from the literature. Works about automated image analysis or the application of an AI algorithm on non-neoplastic kidney histological samples were included and analyzed to extract information such as publication year, AI task, and learning type. Prepublication servers and reviews were not included. RESULTS: Seventy-six (76) original research articles were selected. Most of the studies were conducted in the United States in the last 7 years. To date, research has been mainly conducted on relatively easy tasks, like single-stain glomerular segmentation. However, there is a trend towards developing more complex tasks such as glomerular multi-stain classification. CONCLUSION: Deep learning has been used to identify patterns in complex histopathology data and looks promising for the comprehensive assessment of renal biopsy, through the use of multiple stains and virtual staining techniques. Hybrid and collaborative learning approaches have also been explored to utilize large amounts of unlabeled data. A diverse team of experts, including nephropathologists, computer scientists, and clinicians, is crucial for the development of AI systems for nephropathology. Collaborative efforts among multidisciplinary experts result in clinically relevant and effective AI tools.

Open Surgery Treatment of a Rare Cause of Intestinal Obstruction: Left Paradoudenal Hernia.

An internal hernia is the protrusion of visceral contents through a congenital or acquired defect in the peritoneum or mesentery within the abdominal cavity. In approximately 0.6-5.8% of patients with small intestinal obstruction, the cause is internal hernia, with paraduodenal hernias accounting for approximately 40% of cases. Here, we present the case of a 51-year-old man diagnosed with obstruction of the small intestine caused by a hernia on the left side of the duodenum. The treatment involved returning the bowel loops to the normal position and closing the hernia pocket using Prolene 2.0 sutures. The duration of the surgery was 30 min. Five days later, the patient's condition was stable and he was discharged from the hospital; at the 32-month postoperative follow-up, he remained in stable condition with no recurrence. An abdominal computed tomography scan is valuable for early diagnosis of paraduodenal hernia in the absence of complications, and the cause can be identified and the bowel returned to the normal position by endoscopic surgery, with closure of the hernia pocket if the intestine does not stick to the pocket.

DropConnect is effective in modeling uncertainty of Bayesian deep networks.

Deep neural networks (DNNs) have achieved state-of-the-art performance in many important domains, including medical diagnosis, security, and autonomous driving. In domains where safety is highly critical, an erroneous decision can result in serious consequences. While a perfect prediction accuracy is not always achievable, recent work on Bayesian deep networks shows that it is possible to know when DNNs are more likely to make mistakes. Knowing what DNNs do not know is desirable to increase the safety of deep learning technology in sensitive applications; Bayesian neural networks attempt to address this challenge. Traditional approaches are computationally intractable and do not scale well to large, complex neural network architectures. In this paper, we develop a theoretical framework to approximate Bayesian inference for DNNs by imposing a Bernoulli distribution on the model weights. This method called Monte Carlo DropConnect (MC-DropConnect) gives us a tool to represent the model uncertainty with little change in the overall model structure or computational cost. We extensively validate the proposed algorithm on multiple network architectures and datasets for classification and semantic segmentation tasks. We also propose new metrics to quantify uncertainty estimates. This enables an objective comparison between MC-DropConnect and prior approaches. Our empirical results demonstrate that the proposed framework yields significant improvement in both prediction accuracy and uncertainty estimation quality compared to the state of the art.

Kidney Level Lupus Nephritis Classification Using Uncertainty Guided Bayesian Convolutional Neural Networks.

The kidney biopsy based diagnosis of Lupus Nephritis (LN) is characterized by low inter-observer agreement, with misdiagnosis being associated with increased patient morbidity and mortality. Although various Computer Aided Diagnosis (CAD) systems have been developed for other nephrohistopathological applications, little has been done to accurately classify kidneys based on their kidney level Lupus Glomerulonephritis (LGN) scores. The successful implementation of CAD systems has also been hindered by the diagnosing physician's perceived classifier strengths and weaknesses, which has been shown to have a negative effect on patient outcomes. We propose an Uncertainty-Guided Bayesian Classification (UGBC) scheme that is designed to accurately classify control, class I/II, and class III/IV LGN (3 class) at both the glomerular-level classification task (26,634 segmented glomerulus images) and the kidney-level classification task (87 MRL/lpr mouse kidney sections). Data annotation was performed using a high throughput, bulk labeling scheme that is designed to take advantage of Deep Neural Network's (or DNNs) resistance to label noise. Our augmented UGBC scheme achieved a 94.5% weighted glomerular-level accuracy while achieving a weighted kidney-level accuracy of 96.6%, improving upon the standard Convolutional Neural Network (CNN) architecture by 11.8% and 3.5% respectively.

Whole-brain tissue mapping toolkit using large-scale highly multiplexed immunofluorescence imaging and deep neural networks.

Mapping biological processes in brain tissues requires piecing together numerous histological observations of multiple tissue samples. We present a direct method that generates readouts for a comprehensive panel of biomarkers from serial whole-brain slices, characterizing all major brain cell types, at scales ranging from subcellular compartments, individual cells, local multi-cellular niches, to whole-brain regions from each slice. We use iterative cycles of optimized 10-plex immunostaining with 10-color epifluorescence imaging to accumulate highly enriched image datasets from individual whole-brain slices, from which seamless signal-corrected mosaics are reconstructed. Specific fluorescent signals of interest are isolated computationally, rejecting autofluorescence, imaging noise, cross-channel bleed-through, and cross-labeling. Reliable large-scale cell detection and segmentation are achieved using deep neural networks. Cell phenotyping is performed by analyzing unique biomarker combinations over appropriate subcellular compartments. This approach can accelerate pre-clinical drug evaluation and system-level brain histology studies by simultaneously profiling multiple biological processes in their native anatomical context.

Risk-Aware Machine Learning Classifier for Skin Lesion Diagnosis.

Knowing when a machine learning system is not confident about its prediction is crucial in medical domains where safety is critical. Ideally, a machine learning algorithm should make a prediction only when it is highly certain about its competency, and refer the case to physicians otherwise. In this paper, we investigate how Bayesian deep learning can improve the performance of the machine-physician team in the skin lesion classification task. We used the publicly available HAM10000 dataset, which includes samples from seven common skin lesion categories: Melanoma (MEL), Melanocytic Nevi (NV), Basal Cell Carcinoma (BCC), Actinic Keratoses and Intraepithelial Carcinoma (AKIEC), Benign Keratosis (BKL), Dermatofibroma (DF), and Vascular (VASC) lesions. Our experimental results show that Bayesian deep networks can boost the diagnostic performance of the standard DenseNet-169 model from 81.35% to 83.59% without incurring additional parameters or heavy computation. More importantly, a hybrid physician-machine workflow reaches a classification accuracy of 90 % while only referring 35 % of the cases to physicians. The findings are expected to generalize to other medical diagnosis applications. We believe that the availability of risk-aware machine learning methods will enable a wider adoption of machine learning technology in clinical settings.

Phasetime: Deep Learning Approach to Detect Nuclei in Time Lapse Phase Images.

Time lapse microscopy is essential for quantifying the dynamics of cells, subcellular organelles and biomolecules. Biologists use different fluorescent tags to label and track the subcellular structures and biomolecules within cells. However, not all of them are compatible with time lapse imaging, and the labeling itself can perturb the cells in undesirable ways. We hypothesized that phase image has the requisite information to identify and track nuclei within cells. By utilizing both traditional blob detection to generate binary mask labels from the stained channel images and the deep learning Mask RCNN model to train a detection and segmentation model, we managed to segment nuclei based only on phase images. The detection average precision is 0.82 when the IoU threshold is to be set 0.5. And the mean IoU for masks generated from phase images and ground truth masks from experts is 0.735. Without any ground truth mask labels during the training time, this is good enough to prove our hypothesis. This result enables the ability to detect nuclei without the need for exogenous labeling.

Enhanced gastric stability of esomeprazole by molecular interaction and modulation of microenvironmental pH with alkalizers in solid dispersion.

Due to the instability of esomeprazole magnesium dihydrate (EPM), a proton pump inhibitor, in gastric fluid, enteric-coated dosage form is commonly used for therapeutic application. In this study, we prepared new gastric fluid resistant solid dispersions (SDs) containing alkalizers. Then, new mechanistic evidence regarding the effects of pharmaceutical alkalizers on the aqueous stability of EPM in simulated gastric fluid was investigated. The alkalizer-loaded SD were prepared by dissolving or dispersing EPM, hydroxypropyl methylcellulose (HPMC) 6 cps, and an alkalizer, in ethanol 50% (v/v) followed by spray drying. Nine different alkalizers were assessed for in vitro stability in two media, simulated gastric fluid (pH 1.2 buffer) and simulated intestinal fluid (pH 6.8 buffer). The microenvironmental pH (pHM) was measured to evaluate the effect of the alkalizer on the pHM of SDs. Drug crystallinity and morphology of the SDs were also examined by differential scanning calorimetry (DSC), powder X-ray diffraction (PXRD), and scanning electron microscopy (SEM). The interactions among EPM, the polymer, and the alkalizer were elucidated by Fourier transform infrared (FTIR) spectroscopy. The in vivo absorption studies of the optimized alkalizer-containing SD and the enteric-coated reference tablet Nexium® were then conducted in beagle dogs. Among alkalizers, MgO loaded in SDs proved to be the best alkalizer to stabilize EPM in simulated gastric fluid. pHM values of the alkalizer-containing SDs were significantly higher than that of the SD without alkalizer. The pHM values decreased in the following order: MgO, Na2CO3, Ca(OH)2, and no alkalizer. DSC and PXRD data exhibited a change in the drug crystallinity of the SDs from crystalline to amorphous form. SEM data showed a relatively spherical shape of the MgO-loaded SD compared to the less-defined shape of pure drug. FTIR indicated a strong molecular interaction among EPM, alkalizer and polymer; in particular, MgO showed the strongest interaction with EPM. It was evident that alkalizer interacts with benzimidazole ring and/or sulfonyl group of EPM for enhancing EPM stability in gastric fluid. Regarding the in vivo absorption studies in beagle dogs, the optimized SD (C16) was bioequivalent to the reference Nexium® and had a considerable greater absorption at the early stages. The current alkalizer-containing SD could provide a promising approach for aqueous stabilization of acid-labile drugs without using enteric coating method.

Dual release and molecular mechanism of bilayered aceclofenac tablet using polymer mixture.

The objectives of the present study were to develop a controlled-release bilayered tablet of aceclofenac (AFN) 200mg with dual release and to gain a mechanistic understanding of the enhanced sustained release capability achieved by utilizing a binary mixture of the sustained release materials. Different formulations of the sustained-release layer were formulated by employing hydroxypropyl methylcellulose (HPMC) and hydroxypropyl cellulose (HPC) as the major retarding polymers. The in vitro dissolution studies of AFN bilayered tablets were carried out in intestinal fluid (pH 6.8 buffer). The mechanism of the synergistic rate-retarding effect of the polymer mixture containing HPC and carbomer was elucidated by the rate of swelling and erosion in intestinal fluid and the molecular interactions in the polymer network. The optimized bilayered tablets had similar in vitro dissolution profiles to the marketed tablet Clanza®CR based on the similarity factor (f2) in combination with their satisfactory micromeritic, physicochemical properties, and stability profiles. Drug release from HPMC-based matrix was controlled by non-Fickian transport, while drug release from HPC-based matrix was solely governed by drug diffusion. The swelling and erosion data exhibited a dramatic increase of water uptake and a reduction of weight loss in the polymer mixture-loaded tablet. Fourier transform infrared (FTIR) spectra revealed strong hydrogen bonding between HPC and carbomer in the polymer mixture. Regarding spatial distribution of polymers in the polymer mixture-loaded tablet, carbomer was found to be the main component of the gel layer during the first 2h of the hydration process, which was responsible for retarding drug release at initial stage. This process was then followed by a gradual transition of HPC from the glassy core to the gel layer for further increasing gel strength.

Coupled projections for adaptation of dictionaries.

Data-driven dictionaries have produced the state-of-the-art results in various classification tasks. However, when the target data has a different distribution than the source data, the learned sparse representation may not be optimal. In this paper, we investigate if it is possible to optimally represent both source and target by a common dictionary. In particular, we describe a technique which jointly learns projections of data in the two domains, and a latent dictionary which can succinctly represent both the domains in the projected low-dimensional space. The algorithm is modified to learn a common discriminative dictionary, which can further improve the classification performance. The algorithm is also effective for adaptation across multiple domains and is extensible to nonlinear feature spaces. The proposed approach does not require any explicit correspondences between the source and target domains, and yields good results even when there are only a few labels available in the target domain. We also extend it to unsupervised adaptation in cases where the same feature is extracted across all domains. Further, it can also be used for heterogeneous domain adaptation, where different features are extracted for different domains. Various recognition experiments show that the proposed method performs on par or better than competitive state-of-the-art methods.

Support vector shape: a classifier-based shape representation.

We introduce a novel implicit representation for 2D and 3D shapes based on Support Vector Machine (SVM) theory. Each shape is represented by an analytic decision function obtained by training SVM, with a Radial Basis Function (RBF) kernel so that the interior shape points are given higher values. This empowers support vector shape (SVS) with multifold advantages. First, the representation uses a sparse subset of feature points determined by the support vectors, which significantly improves the discriminative power against noise, fragmentation, and other artifacts that often come with the data. Second, the use of the RBF kernel provides scale, rotation, and translation invariant features, and allows any shape to be represented accurately regardless of its complexity. Finally, the decision function can be used to select reliable feature points. These features are described using gradients computed from highly consistent decision functions instead from conventional edges. Our experiments demonstrate promising results.

Design of non-linear kernel dictionaries for object recognition.

In this paper, we present dictionary learning methods for sparse signal representations in a high dimensional feature space. Using the kernel method, we describe how the well known dictionary learning approaches, such as the method of optimal directions and KSVD, can be made nonlinear. We analyze their kernel constructions and demonstrate their effectiveness through several experiments on classification problems. It is shown that nonlinear dictionary learning approaches can provide significantly better performance compared with their linear counterparts and kernel principal component analysis, especially when the data is corrupted by different types of degradations.