Machine Learning Methods for Computer-Aided Breast Cancer Diagnosis Using Histopathology: A Narrative Review

Histopathology is a method used for breast cancer diagnosis. Machine learning (ML) methods have achieved success for supervised learning tasks in the medical domain. In this article, we investigate the impact of ML for the diagnosis of breast cancer using histopathology images of conventional photomicroscopy. Cancer diagnosis is the identification of images as cancer or noncancer, and this involves image preprocessing, feature extraction, classification, and performance analysis. In this article, different approaches to perform these necessary steps are reviewed. We find that most ML research for breast cancer diagnosis has been focused on deep learning. Based on inferences from the recent research activities, we discuss how ML methods can benefit conventional microscopy-based breast cancer diagnosis. Finally, we discuss the research gaps of ML approaches for the implementation in a real pathology environment and propose future research guidelines.     

Open-source tools for data mining

With a growing volume of biomedical databases and repositories, the need to develop a set of tools to address their analysis and support knowledge discovery is becoming acute. The data mining community has developed a substantial set of techniques for computational treatment of these data. In this article, we discuss the evolution of open-source toolboxes that data mining researchers and enthusiasts have developed over the span of a few decades and review several currently available open-source data mining suites. The approaches we review are diverse in data mining methods and user interfaces and also demonstrate that the field and its tools are ready to be fully exploited in biomedical research.     

Role of artificial intelligence in the care of patients with nonsmall cell lung cancer

Background

Lung cancer is the leading cause of cancer death worldwide. In up to 57% of patients, it is diagnosed at an advanced stage and the 5‐year survival rate ranges between 10%‐16%. There has been a significant amount of research using machine learning to generate tools using patient data to improve outcomes.

Methods

This narrative review is based on research material obtained from PubMed up to Nov 2017. The search terms include “artificial intelligence,” “machine learning,” “lung cancer,” “Nonsmall Cell Lung Cancer (NSCLC),” “diagnosis” and “treatment.”

Results

Recent studies support the use of computer‐aided systems and the use of radiomic features to help diagnose lung cancer earlier. Other studies have looked at machine learning (ML) methods that offer prognostic tools to doctors and help them in choosing personalized treatment options for their patients based on molecular, genetics and histological features. Combining artificial intelligence approaches into health care may serve as a beneficial tool for patients with NSCLC, and this review outlines these benefits and current shortcomings throughout the continuum of care.

Conclusion

We present a review of the various applications of ML methods in NSCLC as it relates to improving diagnosis, treatment and outcomes.     

Data mining in genomics

This article reviews important emerging statistical concepts, data mining techniques, and applications that have been recently developed and used for genomic data analysis. First, general background and some critical issues in genomic data mining are summarized. A novel concept of statistical significance is described, the so-called "false discovery rate"-the rate of false-positives among all positive findings-which has been suggested to control the error rate of numerous false-positives in large screening biological data analysis. Two recent statistical testing methods are then introduced: significance analysis of microarray and local pooled error tests. Statistical modeling in genomic data analysis is then presented, such as analysis of variance and heterogeneous error modeling approaches that have been suggested for analyzing microarray data obtained from multiple experimental or biological conditions. Two sections then describe data exploration and discovery tools largely termed as supervised learning and unsupervised learning. The former approaches include several multivariate statistical methods to investigate coexpression patterns of multiple genes, and the latter are the classification methods to discover genomic biomarker signatures for predicting important subclasses of human diseases. The last section briefly summarizes various genomic data mining approaches in biomedical pathway analysis and patient outcome or chemotherapeutic response prediction.     

Greening the synthesis of peptide therapeutics: an industrial perspective

Solid-phase peptide synthesis (SPPS) is generally the method of choice for the chemical synthesis of peptides, allowing routine synthesis of virtually any type of peptide sequence, including complex or cyclic peptide products. Importantly, SPPS can be automated and is scalable, which has led to its widespread adoption in the pharmaceutical industry, and a variety of marketed peptide-based drugs are now manufactured using this approach. However, SPPS-based synthetic strategies suffer from a negative environmental footprint mainly due to extensive solvent use. Moreover, most of the solvents used in peptide chemistry are classified as problematic by environmental agencies around the world and will soon need to be replaced, which in recent years has spurred a movement in academia and industry to make peptide synthesis greener. These efforts have been centred around solvent substitution, recycling and reduction, as well as exploring alternative synthetic methods. In this review, we focus on methods pertaining to solvent substitution and reduction with large-scale industrial production in mind, and further outline emerging technologies for peptide synthesis. Specifically, the technical requirements for large-scale manufacturing of peptide therapeutics are addressed.

This review highlights the efforts made to date to promote greener peptide synthesis, from an industrial perspective.     

Systematic Review of the Effectiveness of Machine Learning Algorithms for Classifying Pain Intensity,Phenotype or Treatment Outcomes Using Electroencephalogram Data

Recent attempts to utilize machine learning (ML) to predict pain-related outcomes from Electroencephalogram (EEG) data demonstrate promising results. The primary aim of this review was to evaluate the effectiveness of ML algorithms for predicting pain intensity, phenotypes or treatment response from EEG. Electronic databases MEDLINE, EMBASE, Web of Science, PsycINFO and The Cochrane Library were searched. A total of 44 eligible studies were identified, with 22 presenting attempts to predict pain intensity, 15 investigating the prediction of pain phenotypes and seven assessing the prediction of treatment response. A meta-analysis was not considered appropriate for this review due to heterogeneous methods and reporting. Consequently, data were narratively synthesized. The results demonstrate that the best performing model of the individual studies allows for the prediction of pain intensity, phenotypes and treatment response with accuracies ranging between 62 to 100%, 57 to 99% and 65 to 95.24%, respectively. The results suggest that ML has the potential to effectively predict pain outcomes, which may eventually be used to assist clinical care. However, inadequate reporting and potential bias reduce confidence in the results. Future research should improve reporting standards and externally validate models to decrease bias, which would increase the feasibility of clinical translation.PerspectiveThis systematic review explores the state-of-the-art machine learning methods for predicting pain intensity, phenotype or treatment response from EEG data. Results suggest that machine learning may demonstrate clinical utility, pending further research and development. Areas for improvement, including standardized processing, reporting and the need for better methodological assessment tools, are discussed.     

A gold mine for drug discovery: Strategies to develop cyclic peptides into therapies

As a versatile therapeutic modality, peptides attract much attention because of their great binding affinity, low toxicity, and the capability of targeting traditionally “undruggable” protein surfaces. However, the deficiency of cell permeability and metabolic stability always limits the success of in vitro bioactive peptides as drug candidates. Peptide macrocyclization is one of the most established strategies to overcome these limitations. Over the past decades, more than 40 cyclic peptide drugs have been clinically approved, the vast majority of which are derived from natural products. The de novo discovered cyclic peptides on the basis of rational design and in vitro evolution, have also enabled the binding with targets for which nature provides no solutions. The current review summarizes different classes of cyclic peptides with diverse biological activities, and presents an overview of various approaches to develop cyclic peptide-based drug candidates, drawing upon series of examples to illustrate each strategy.     

Modern lipid‐, carbohydrate‐, and peptide‐based delivery systems for peptide,vaccine, and gene products

Research related to peptide, vaccine, and gene delivery has grown exponentially over the last decade. In this review, we discuss the development of delivery systems for peptides, gene and vaccine products. Special focus is given to different lipidation and glycosylation strategies to improve the metabolic stability and membrane permeability of therapeutics, and their targeting to specific sites. The synthetic methods for preparation of the systems are also described. © 2009 Wiley Periodicals, Inc. Med Res Rev, 31, No. 4, 520–547, 2011     

QSAR modeling of the toxicity classification of superparamagnetic iron oxide nanoparticles (SPIONs) in stem-cell monitoring applications: an integrated study from data curation to model development

The use of in silico approaches for the prediction of biomedical properties of nano-biomaterials (NBMs) can play a significant role in guiding and reducing wetlab experiments. Computational methods, such as data mining and machine learning techniques, can increase the efficiency and reduce the time and cost required for hazard and risk assesment and for designing new safer NBMs. A major obstacle in developing accurate and well-validated in silico models such as Nano Quantitative Structure–Activity Relationships (Nano-QSARs) is that although the volume of data published in the literature is increasing, the data are fragmented in many different publications and are not sufficiently curated for modelling purposes. Moreover, NBMs exhibit high complexity and heterogeneity in their structures, making data collection and curation and QSAR model development more challenging compared to traditional small molecules. The aim of this study was to construct and fully validate a Nano-QSAR model for the prediction of toxicological properties of superparamagnetic iron oxide nanoparticles (SPIONs), focusing on their application as Magnetic Resonance Imaging (MRI) contrast agents for non-invasive stem cell labelling and tracking. To achieve this goal, we first performed an extensive search through the literature for collecting and curating relevant data and we developed a dataset containing both physicochemical and toxicological properties of SPIONs. The data were analysed next, using Automated machine learning (Auto-ML) approaches for optimising the development and validation of nanotoxicity classification QSAR models of SPIONs. Further analysis of relative attribute importances revealed that physicochemical properties such as the size and the magnetic core are the dominant attributes correlated to the toxicity of SPIONs. Our results suggest that as more systematic information from NBM experimental tests becomes available, computational tools could play an important role in supporting the safety-by-design (SbD) concept in regenerative medicine and disease therapeutics.

Development of a novel QSAR model for the prediction of toxicity of superparamagnetic iron oxide nanoparticles (SPIONs) in stem-cell monitoring applications.     

Disulfide-constrained peptides that bind to the extracellular portion of the prostate-specific membrane antigen

The prostate-specific membrane antigen (PSMA) is a well-characterized surface antigen, overexpressed in the most advanced, androgen-resistant human prostate cancer cells. We sought to exploit PSMA cell surface properties as a target for short peptides that will potentially guide protein-based therapeutics, such as viral vectors, to prostate cancer cells. Two separate phage display peptide strategies were applied, in parallel, to purified PSMA protein bound to two separate substrates. We reasoned that peptide sequences common to both substrate selections would be specific binders of PSMA. Additionally, the design allowed for stringent cross-selections, where phage populations from one selection condition could be applied to the alternative substrate. These strategies resulted in a series of phage displayed peptides able to bind to PSMA by ELISA and direct binding assays, both with purified protein and in prostate cancer cells. Cell binding is competitively inhibited by purified PSMA. The synthesized peptides are capable of enhancing PSMA carboxypeptidase enzymatic activity, suggesting protein folding stabilization. The discovery of these peptides provides the foundation for subsequent development of peptide targeted therapeutics against prostate cancer.     

Venoms as a platform for human drugs: translating toxins into therapeutics

Introduction: An extraordinarily diverse range of animals have evolved venoms for predation, defence, or competitor deterrence. The major components of most venoms are peptides and proteins that are often protease-resistant due to their disulfide-rich architectures. Some of these toxins have become valuable as pharmacological tools and/or therapeutics due to their extremely high specificity and potency for particular molecular targets. There are currently six FDA-approved drugs derived from venom peptides or proteins.

Areas covered: This article surveys the current pipeline of venom-derived therapeutics and critically examines the potential of peptide and protein drugs derived from venoms. Emerging trends are identified, including an increasing industry focus on disulfide-rich venom peptides and the use of a broader array of molecular targets in order to develop venom-based therapeutics for treating a wider range of clinical conditions.

Expert opinion: Key technical advances in combination with a renewed industry-wide focus on biologics have converged to provide a larger than ever pipeline of venom-derived therapeutics. Disulfide-rich venom peptides obviate some of the traditional disadvantages of therapeutic peptides and some may be suitable for oral administration. Moreover, some venom peptides can breach the blood brain barrier and translocate across cell membranes, which opens up the possibility of exploiting molecular targets not previously accessible to peptide drugs.     

Predicting drug susceptibility of non–small cell lung cancers based on genetic lesions

Somatic genetic alterations in cancers have been linked with response to targeted therapeutics by creation of specific dependency on activated oncogenic signaling pathways. However, no tools currently exist to systematically connect such genetic lesions to therapeutic vulnerability. We have therefore developed a genomics approach to identify lesions associated with therapeutically relevant oncogene dependency. Using integrated genomic profiling, we have demonstrated that the genomes of a large panel of human non–small cell lung cancer (NSCLC) cell lines are highly representative of those of primary NSCLC tumors. Using cell-based compound screening coupled with diverse computational approaches to integrate orthogonal genomic and biochemical data sets, we identified molecular and genomic predictors of therapeutic response to clinically relevant compounds. Using this approach, we showed that v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations confer enhanced Hsp90 dependency and validated this finding in mice with KRAS-driven lung adenocarcinoma, as these mice exhibited dramatic tumor regression when treated with an Hsp90 inhibitor. In addition, we found that cells with copy number enhancement of v-abl Abelson murine leukemia viral oncogene homolog 2 (ABL2) and ephrin receptor kinase and v-src sarcoma (Schmidt-Ruppin A-2) viral oncogene homolog (avian) (SRC) kinase family genes were exquisitely sensitive to treatment with the SRC/ABL inhibitor dasatinib, both in vitro and when it xenografted into mice. Thus, genomically annotated cell-line collections may help translate cancer genomics information into clinical practice by defining critical pathway dependencies amenable to therapeutic inhibition.     

Physics-aware learning and domain-specific loss design in ophthalmology

The human cataract, a developing opacification of the human eye lens, currently constitutes the world’s most frequent cause for blindness. As a result, cataract surgery has become the most frequently performed ophthalmic surgery in the world. By removing the human lens and replacing it with an artificial intraocular lens (IOL), the optical system of the eye is restored. In order to receive a good refractive result, the IOL specifications, especially the refractive power, have to be determined precisely prior to surgery. In the last years, there has been a body of work to perform this prediction by using biometric information extracted from OCT imaging data, recently also by machine learning (ML) methods. Approaches so far consider only biometric information or physical modelling, but provide no effective combination, while often also neglecting IOL geometry. Additionally, ML on small data sets without sufficient domain coverage can be challenging. To solve these issues, we propose OpticNet, a novel optical refraction network based on an unsupervised, domain-specific loss function that explicitly incorporates physical information into the network. By providing a precise and differentiable light propagation eye model, physical gradients following the eye optics are backpropagated into the network. We further propose a new transfer learning procedure, which allows the unsupervised pre-training on the optical model and fine-tuning of the network on small amounts of surgical patient data. We show that our method outperforms the current state of the art on five OCT-image based data sets, provides better domain coverage within its predictions, and achieves better physical consistency.     

The quantitative evaluation of functional neuroimaging experiments: the NPAIRS data analysis framework

We introduce a data-analysis framework and performance metrics for evaluating and optimizing the interaction between activation tasks, experimental designs, and the methodological choices and tools for data acquisition, preprocessing, data analysis, and extraction of statistical parametric maps (SPMs). Our NPAIRS (nonparametric prediction, activation, influence, and reproducibility resampling) framework provides an alternative to simulations and ROC curves by using real PET and fMRI data sets to examine the relationship between prediction accuracy and the signal-to-noise ratios (SNRs) associated with reproducible SPMs. Using cross-validation resampling we plot training-test set predictions of the experimental design variables (e.g., brain-state labels) versus reproducibility SNR metrics for the associated SPMs. We demonstrate the utility of this framework across the wide range of performance metrics obtained from [(15)O]water PET studies of 12 age- and sex-matched data sets performing different motor tasks (8 subjects/set). For the 12 data sets we apply NPAIRS with both univariate and multivariate data-analysis approaches to: (1) demonstrate that this framework may be used to obtain reproducible SPMs from any data-analysis approach on a common Z-score scale (rSPM[Z]); (2) demonstrate that the histogram of a rSPM[Z] image may be modeled as the sum of a data-analysis-dependent noise distribution and a task-dependent, Gaussian signal distribution that scales monotonically with our reproducibility performance metric; (3) explore the relation between prediction and reproducibility performance metrics with an emphasis on bias-variance tradeoffs for flexible, multivariate models; and (4) measure the broad range of reproducibility SNRs and the significant influence of individual subjects. A companion paper describes learning curves for four of these 12 data sets, which describe an alternative mutual-information prediction metric and NPAIRS reproducibility as a function of training-set sizes from 2 to 18 subjects. We propose the NPAIRS framework as a validation tool for testing and optimizing methodological choices and tools in functional neuroimaging.     

Identifying Goals of Care Conversations in the Electronic Health Record Using Natural Language Processing and Machine Learning

ContextGoals-of-care discussions are an important quality metric in palliative care. However, goals-of-care discussions are often documented as free text in diverse locations. It is difficult to identify these discussions in the electronic health record (EHR) efficiently.ObjectivesTo develop, train, and test an automated approach to identifying goals-of-care discussions in the EHR, using natural language processing (NLP) and machine learning (ML).MethodsFrom the electronic health records of an academic health system, we collected a purposive sample of 3183 EHR notes (1435 inpatient notes and 1748 outpatient notes) from 1426 patients with serious illness over 2008–2016, and manually reviewed each note for documentation of goals-of-care discussions. Separately, we developed a program to identify notes containing documentation of goals-of-care discussions using NLP and supervised ML. We estimated the performance characteristics of the NLP/ML program across 100 pairs of randomly partitioned training and test sets. We repeated these methods for inpatient-only and outpatient-only subsets.ResultsOf 3183 notes, 689 contained documentation of goals-of-care discussions. The mean sensitivity of the NLP/ML program was 82.3% (SD 3.2%), and the mean specificity was 97.4% (SD 0.7%). NLP/ML results had a median positive likelihood ratio of 32.2 (IQR 27.5–39.2) and a median negative likelihood ratio of 0.18 (IQR 0.16–0.20). Performance was better in inpatient-only samples than outpatient-only samples.ConclusionUsing NLP and ML techniques, we developed a novel approach to identifying goals-of-care discussions in the EHR. NLP and ML represent a potential approach toward measuring goals-of-care discussions as a research outcome and quality metric.     

The Impact of Artificial Intelligence and Machine Learning in Radiation Therapy: Considerations for Future Curriculum Enhancement

Artificial intelligence (AI) and machine learning (ML) approaches have caught the attention of many in health care. Current literature suggests there are many potential benefits that could transform future clinical workflows and decision making. Embedding AI and ML concepts in radiation therapy education could be a fundamental step in equipping radiation therapists (RTs) to engage in competent and safe practice as they utilise clinical technologies. In this discussion paper, the authors provide a brief review of some applications of AI and ML in radiation therapy and discuss pertinent considerations for radiation therapy curriculum enhancement. As the current literature suggests, AI and ML approaches will impose changes to routine clinical radiation therapy tasks. The emphasis in RT education could be on critical evaluation of AI and ML application in routine clinical workflows and gaining an understanding of the impact on quality assurance, provision of quality of care and safety in radiation therapy as well as research. It is also imperative RTs have a broader understanding of AI/ML impact on health care, including ethical and legal considerations. The paper concludes with recommendations and suggestions to deliberately embed AI and ML aspects in RT education to empower future RT practitioners.     

Rethinking PICO in the Machine Learning Era: ML-PICO

Background  Machine learning (ML) has captured the attention of many clinicians who may not have formal training in this area but are otherwise increasingly exposed to ML literature that may be relevant to their clinical specialties. ML papers that follow an outcomes-based research format can be assessed using clinical research appraisal frameworks such as PICO (Population, Intervention, Comparison, Outcome). However, the PICO frameworks strain when applied to ML papers that create new ML models, which are akin to diagnostic tests. There is a need for a new framework to help assess such papers. Objective  We propose a new framework to help clinicians systematically read and evaluate medical ML papers whose aim is to create a new ML model: ML-PICO (Machine Learning, Population, Identification, Crosscheck, Outcomes). We describe how the ML-PICO framework can be applied toward appraising literature describing ML models for health care. Conclusion  The relevance of ML to practitioners of clinical medicine is steadily increasing with a growing body of literature. Therefore, it is increasingly important for clinicians to be familiar with how to assess and best utilize these tools. In this paper we have described a practical framework on how to read ML papers that create a new ML model (or diagnostic test): ML-PICO. We hope that this can be used by clinicians to better evaluate the quality and utility of ML papers.     

Comparing methods of analyzing fMRI statistical parametric maps

Approaches for the analysis of statistical parametric maps (SPMs) can be crudely grouped into three main categories in which different philosophies are applied to delineate activated regions. These being type I error control thresholding, false discovery rate (FDR) control thresholding and posterior probability thresholding. To better understand the properties of these main approaches, we carried out a simulation study to compare the approaches as they would be used on real data sets. Using default settings, we find that posterior probability thresholding is the most powerful approach, and type I error control thresholding provides the lowest levels of type I error. False discovery rate control thresholding performs in between the other approaches for both these criteria, although for some parameter settings this approach can approximate the performance of posterior probability thresholding. Based on these results, we discuss the relative merits of the three approaches in an attempt to decide upon an optimal approach. We conclude that viewing the problem of delineating areas of activation as a classification problem provides a highly interpretable framework for comparing the methods. Within this framework, we highlight the role of the loss function, which explicitly penalizes the types of errors that may occur in a given analysis.     

Cell-penetrating peptides can confer biological function: Regulation of inflammatory cytokines in human monocytes by MK2 inhibitor peptides

Cell-penetrating peptides have been used as a method of delivering biologically active peptide for over two decades. In this paper, we covalently attached four different cell-penetrating peptides to a peptide that inhibits a kinase important in inflammation, mitogen-activated protein kinase activated protein kinase 2 (MAPKAP2 or MK2). We evaluated the specificity, toxicity, and functionality of these therapeutics in an in vitro model of inflammation using THP-1 monocytes. When treated with the MK2 peptide inhibitors, activated THP-1 human monocytes challenged with lipopolysaccharide (LPS) showed a decrease in TNF-α and IL-6 excretion without apparent toxicity. In addition, western blot analysis revealed decreases in the phosphorylation of heat shock protein 27 (HSP27), a downstream substrate of MK2. These results suggested that our peptides inhibited MK2 activity in vitro and should be investigated further as a potential therapeutic for applications involving inflammation. Furthermore, our results suggested that cell-penetrating peptides can be bioactive.