Anthony B. Costa

Cholesterol Sulfate Imaging in Human Prostate Cancer Tissue by Desorption Electrospray Ionization Mass Spectrometry

Development of methods for rapid distinction between cancerous and non-neoplastic tissues is an important goal in disease diagnosis. To this end, desorption electrospray ionization mass spectrometry (DESI-MS) imaging was applied to analyze the lipid profiles of thin tissue sections of 68 samples of human prostate cancer and normal tissue. The disease state of the tissue sections was determined by independent histopathological examination. Cholesterol sulfate was identified as a differentiating compound, found almost exclusively in cancerous tissues including tissue containing precancerous lesions. The presence of cholesterol sulfate in prostate tissues might serve as a tool for prostate cancer diagnosis although confirmation through larger and more diverse cohorts and correlations with clinical outcome data is needed.

Multivariate statistical differentiation of renal cell carcinomas based on lipidomic analysis by ambient ionization imaging mass spectrometry.

AbstractDesorption electrospray ionization (DESI) mass spectrometry (MS) was used in an imaging mode to interrogate the lipid profiles of thin tissue sections of 11 sample pairs of human papillary renal cell carcinoma (RCC) and adjacent normal tissue and nine sample pairs of clear cell RCC and adjacent normal tissue. DESI-MS images showing the spatial distributions of particular glycerophospholipids (GPs) and free fatty acids in the negative ion mode were compared to serial tissue sections stained with hematoxylin and eosin (H&E). Increased absolute intensities as well as changes in relative abundance were seen for particular compounds in the tumor regions of the samples. Multivariate statistical analysis using orthogonal projection to latent structures treated partial least square discriminate analysis (PLS-DA) was used for visualization and classification of the tissue pairs using the full mass spectra as predictors. PLS-DA successfully distinguished tumor from normal tissue for both papillary and clear cell RCC with misclassification rates obtained from the validation set of 14.3% and 7.8%, respectively. It was also used to distinguish papillary and clear cell RCC from each other and from the combined normal tissues with a reasonable misclassification rate of 23%, as determined from the validation set. Overall DESI-MS imaging combined with multivariate statistical analysis shows promise as a molecular pathology technique for diagnosing cancerous and normal tissue on the basis of GP profiles. FigureMolecular disease diagnostics by DESI without sample preparation. a Good information is obtained by mapping the distribution of individual compounds in the tissue (e.g., PI(18:0/20:4). b Even better discrimination between tumor and healthy tissue is achieved using PLS-DA to consider all the data after having established through a training set of samples the features that correlate with disease as recognized by standard H&E stain pathological examination

Lipid profiles of canine invasive transitional cell carcinoma of the urinary bladder and adjacent normal tissue by desorption electrospray ionization imaging mass spectrometry.

Desorption electrospray ionization mass spectrometry (DESI-MS) was used in an imaging mode to interrogate the lipid profiles of thin tissue sections of canine spontaneous invasive transitional cell carcinoma of the urinary bladder (a model of human invasive bladder cancer) as well as adjacent normal tissue from four different dogs. The glycerophospholipids and sphingolipids that appear as intense signals in both the negative ion and positive ion modes were identified by tandem mass spectrometry product ion scans using collision-induced dissociation. Differences in the relative distributions of the lipid species were present between the tumor and adjacent normal tissue in both the negative and positive ion modes. DESI-MS images showing the spatial distributions of particular glycerophospholipids, sphinoglipids, and free fatty acids in both the negative and positive ion modes were compared to serial tissue sections that were stained with hematoxylin and eosin (H&E). Increased absolute and relative intensities for at least five different glycerophospholipids and three free fatty acids in the negative ion mode and at least four different lipid species in the positive ion mode were seen in the tumor region of the samples in all four dogs. In addition, one sphingolipid species exhibited increased signal intensity in the positive ion mode in normal tissue relative to the diseased tissue. Principal component analysis was also used to generate unsupervised statistical images from the negative ion mode data, and these images are in excellent agreement with the DESI images obtained from the selected ions and also the H&E-stained tissue.

New ionization methods and miniature mass spectrometers for biomedicine: DESI imaging for cancer diagnostics and paper spray ionization for therapeutic drug monitoring

The state-of-the-art in two new ambient ionization methods for mass spectrometry, desorption electrospray ionization (DESI) and paper spray (PS), is described and their utility is illustrated with new studies on tissue imaging and biofluid analysis. DESI is an ambient ionization method that can be performed on untreated histological sections of biological tissue in the open lab environment to image lipids, fatty acids, hormones and other compounds. Paper spray is performed in the open lab too; it involves electrospraying dry blood spots or biofluid deposits from a porous medium. PS is characterized by extreme simplicity and speed: a spot of whole blood or other biofluid is analyzed directly from paper, simply by applying a high voltage to the moist paper. Both methods are being developed for use in diagnostics as a means to inform therapy. DESI imaging is applied to create molecular maps of tissue sections without prior labeling or other sample preparation. Like other methods of mass spectrometry imaging (MSI), it combines the chemical speciation of multiple analytes with information on spatial distributions. DESI imaging provides valuable information which correlates with the disease state of tissue as determined by standard histochemical methods. Positive-ion data are presented which complement previously reported negative-ion data on paired human bladder cancerous and adjacent normal tissue sections from 20 patients. These data add to the evidence already in the literature demonstrating that differences in the distributions of particular lipids contain disease-diagnostic information. Multivariate statistical analysis using principal component analysis (PCA) is used to analyze the imaging MS data, and so confirm differences between the lipid profiles of diseased and healthy tissue types. As more such data is acquired, DESI imaging has the potential to be a diagnostic tool for future cancer detection in situ; this suggests a potential role in guiding therapy in parallel with standard histochemical and immunohistological methods. The PS methodology is aimed at high-throughput clinical measurement of quantitative levels of particular therapeutic agents in blood and other biofluids. The experiment allows individual drugs to be quantified at therapeutic levels and data is presented showing quantitative drug analysis from mixtures of therapeutic drugs in whole blood. Data on cholesterol sulfate, a new possible prostate biomarker seen at elevated levels in diseased prostate tissue, but not in healthy prostate tissue in serum are reported using paper spray ionization.

Multivariate Statistical Identification of Human Bladder Carcinomas Using Ambient Ionization Imaging Mass Spectrometry

Diagnosis of human bladder cancer in untreated tissue sections is achieved by using imaging data from desorption electrospray ionization mass spectrometry (DESI-MS) combined with multivariate statistical analysis. We use the distinctive DESI-MS glycerophospholipid (GP) mass spectral profiles to visually characterize and formally classify twenty pairs (40 tissue samples) of human cancerous and adjacent normal bladder tissue samples. The individual ion images derived from the acquired profiles correlate with standard histological hematoxylin and eosin (H&E)-stained serial sections. The profiles allow us to classify the disease status of the tissue samples with high accuracy as judged by reference histological data. To achieve this, the data from the twenty pairs were divided into a training set and a validation set. Spectra from the tumor and normal regions of each of the tissue sections in the training set were used for orthogonal projection to latent structures (O-PLS) treated partial least-square discriminate analysis (PLS-DA). This predictive model was then validated by using the validation set and showed a 5% error rate for classification and a misclassification rate of 12%. It was also used to create synthetic images of the tissue sections showing pixel-by-pixel disease classification of the tissue and these data agreed well with the independent classification that uses histological data by a certified pathologist. This represents the first application of multivariate statistical methods for classification by ambient ionization although these methods have been applied previously to other MS imaging methods. The results are encouraging in terms of the development of a method that could be utilized in a clinical setting through visualization and diagnosis of intact tissue.

Relationship between dynamical entropy and energy dissipation far from thermodynamic equilibrium

Significance The dissipation of energy occurs naturally in systems as diverse as those in biology and as common as those responsible for the weather. Systems that dissipate energy while self-assembling also show promise as a synthetic route to responsive nanoscale materials. Optimizing the energy efficiency of these syntheses requires reducing the theoretical description to only those variables that are essential to an accurate prediction of the energy lost as heat. Here, with computer simulations and a model of assembly, we establish a methodology for identifying those essential variables. With this minimal description of the system we also find a linear relationship between the dynamical entropy of the self-assembling system and the energy dissipated as heat during assembly. Connections between microscopic dynamical observables and macroscopic nonequilibrium (NE) properties have been pursued in statistical physics since Boltzmann, Gibbs, and Maxwell. The simulations we describe here establish a relationship between the Kolmogorov–Sinai entropy and the energy dissipated as heat from a NE system to its environment. First, we show that the Kolmogorov–Sinai or dynamical entropy can be separated into system and bath components and that the entropy of the system characterizes the dynamics of energy dissipation. Second, we find that the average change in the system dynamical entropy is linearly related to the average change in the energy dissipated to the bath. The constant energy and time scales of the bath fix the dynamical relationship between these two quantities. These results provide a link between microscopic dynamical variables and the macroscopic energetics of NE processes.

Patient-Specific Cranial Nerve Identification Using a Discrete Deformable Contour Model for Skull Base Neurosurgery Planning and Simulation

In this paper, we present a minimally supervised method for the identification of the intra-cranial portion of cranial nerves, using a novel, discrete 1-Simplex 3D active contour model. The clinical applications include planning and personalized simulation of skull base neurosurgery. The centerline of a cranial nerve is initialized from two user-supplied end points by computing a Minimal Path. The 1-Simplex is a Newtonian model for vertex motion, where every non-endpoint vertex has 2-connectivity with neighboring vertices, with which it is linked by edges. The segmentation behavior of the model is governed by the equilibrium between internal and external forces. The external forces include an image force that favors a centered path within high-vesselness points. The method is validated quantitatively using synthetic and real MRI datasets.

Natural Language–based Machine Learning Models for the Annotation of Clinical Radiology Reports

Purpose To compare different methods for generating features from radiology reports and to develop a method to automatically identify findings in these reports. Materials and Methods In this study, 96 303 head computed tomography (CT) reports were obtained. The linguistic complexity of these reports was compared with that of alternative corpora. Head CT reports were preprocessed, and machine-analyzable features were constructed by using bag-of-words (BOW), word embedding, and Latent Dirichlet allocation-based approaches. Ultimately, 1004 head CT reports were manually labeled for findings of interest by physicians, and a subset of these were deemed critical findings. Lasso logistic regression was used to train models for physician-assigned labels on 602 of 1004 head CT reports (60%) using the constructed features, and the performance of these models was validated on a held-out 402 of 1004 reports (40%). Models were scored by area under the receiver operating characteristic curve (AUC), and aggregate AUC statistics were reported for (a) all labels, (b) critical labels, and (c) the presence of any critical finding in a report. Sensitivity, specificity, accuracy, and F1 score were reported for the best performing models (a) predictions of all labels and (b) identification of reports containing critical findings. Results The best-performing model (BOW with unigrams, bigrams, and trigrams plus average word embeddings vector) had a held-out AUC of 0.966 for identifying the presence of any critical head CT finding and an average 0.957 AUC across all head CT findings. Sensitivity and specificity for identifying the presence of any critical finding were 92.59% (175 of 189) and 89.67% (191 of 213), respectively. Average sensitivity and specificity across all findings were 90.25% (1898 of 2103) and 91.72% (18 351 of 20 007), respectively. Simpler BOW methods achieved results competitive with those of more sophisticated approaches, with an average AUC for presence of any critical finding of 0.951 for unigram BOW versus 0.966 for the best-performing model. The Yule I of the head CT corpus was 34, markedly lower than that of the Reuters corpus (at 103) or I2B2 discharge summaries (at 271), indicating lower linguistic complexity. Conclusion Automated methods can be used to identify findings in radiology reports. The success of this approach benefits from the standardized language of these reports. With this method, a large labeled corpus can be generated for applications such as deep learning.

Navigation-Linked Heads-Up Display in Intracranial Surgery: Early Experience

Abstract BACKGROUND The use of intraoperative navigation during microscope cases can be limited when attention needs to be divided between the operative field and the navigation screens. Heads-up display (HUD), also referred to as augmented reality, permits visualization of navigation information during surgery workflow. OBJECTIVE To detail our initial experience with HUD. METHODS We retrospectively reviewed patients who underwent HUD-assisted surgery from April 2016 through April 2017. All lesions were assessed for accuracy and those from the latter half of the study were assessed for utility. RESULTS Seventy-nine patients with 84 pathologies were included. Pathologies included aneurysms (14), arteriovenous malformations (6), cavernous malformations (5), intracranial stenosis (3), meningiomas (27), metastasis (4), craniopharygniomas (4), gliomas (4), schwannomas (3), epidermoid/dermoids (3), pituitary adenomas (2) hemangioblastoma (2), choroid plexus papilloma (1), lymphoma (1), osteoblastoma (1), clival chordoma (1), cerebrospinal fluid leak (1), abscess (1), and a cerebellopontine angle Teflon granuloma (1). Fifty-nine lesions were deep and 25 were superficial. Structures identified included the lesion (81), vessels (48), and nerves/brain tissue (31). Accuracy was deemed excellent (71.4%), good (20.2%), or poor (8.3%). Deep lesions were less likely to have excellent accuracy (P = .029). HUD was used during bed/head positioning (50.0%), skin incision (17.3%), craniotomy (23.1%), dural opening (26.9%), corticectomy (13.5%), arachnoid opening (36.5%), and intracranial drilling (13.5%). HUD was deactivated at some point during the surgery in 59.6% of cases. There were no complications related to HUD use. CONCLUSION HUD can be safely used for a wide variety of vascular and oncologic intracranial pathologies and can be utilized during multiple stages of surgery.

Automated deep-neural-network surveillance of cranial images for acute neurologic events

Rapid diagnosis and treatment of acute neurological illnesses such as stroke, hemorrhage, and hydrocephalus are critical to achieving positive outcomes and preserving neurologic function—‘time is brain’1–5. Although these disorders are often recognizable by their symptoms, the critical means of their diagnosis is rapid imaging6–10. Computer-aided surveillance of acute neurologic events in cranial imaging has the potential to triage radiology workflow, thus decreasing time to treatment and improving outcomes. Substantial clinical work has focused on computer-assisted diagnosis (CAD), whereas technical work in volumetric image analysis has focused primarily on segmentation. 3D convolutional neural networks (3D-CNNs) have primarily been used for supervised classification on 3D modeling and light detection and ranging (LiDAR) data11–15. Here, we demonstrate a 3D-CNN architecture that performs weakly supervised classification to screen head CT images for acute neurologic events. Features were automatically learned from a clinical radiology dataset comprising 37,236 head CTs and were annotated with a semisupervised natural-language processing (NLP) framework16. We demonstrate the effectiveness of our approach to triage radiology workflow and accelerate the time to diagnosis from minutes to seconds through a randomized, double-blinded, prospective trial in a simulated clinical environment.A deep-learning algorithm is developed to provide rapid and accurate diagnosis of clinical 3D head CT-scan images to triage and prioritize urgent neurological events, thus potentially accelerating time to diagnosis and care in clinical settings.