Diagnostic Pattern Recognition Research Articles

Emerging Role of Artificial Intelligence and Big Data in Spine Care.

The pace of US Food and Drug Administration-approved medical devices that incorporate artificial intelligence (AI) or machine learning as part of the device is accelerating. As of September 2021, 350 such devices have been approved for commercial sale in the United States. As much as AI has become ubiquitous in our lives-keeping our cars between the lines on the highway, converting speech to text on the fly, recommending movies, books, or restaurants, and so much more, AI also appears destined to become a routine aspect in daily spine surgery. Neural network types of AI programs have achieved extraordinary pattern recognition and predictive abilities-far surpassing human capabilities-and thus appears well suited to back pain and spine surgery diagnostic and treatment pattern recognition and prediction tasks. These AI programs are also data hungry. As luck would have it, surgery generates an estimated 80 MB per patient per day collected in a variety of datasets. When aggregated, this represents a 200+ billion patient record data ocean of diagnostic and treatment patterns. Such Big Data, when combined with a new generation of convolutional neural network (CNN) AI, set the stage for a cognitive revolution in spine surgery. However, there are important issues and concerns. Spine surgery is a mission-critical task. Because AI programs lack explainability and are absolutely reliant on correlative, not causative, data relationships, the emerging role of AI and Big Data in spine surgery will likely come first in productivity tools and later in narrowly defined spine surgery tasks. The purpose of this article is to review the emergence of AI in spine surgery applications and examine spine surgery heuristics and "expert" decision models within the context of AI and Big Data.

Fault Diagnosis of Hydraulic Systems Based on Deep Learning Model With Multirate Data Samples.

Hydraulic systems are a class of typical complex nonlinear systems, which have been widely used in manufacturing, metallurgy, energy, and other industries. Nowadays, the intelligent fault diagnosis problem of hydraulic systems has received increasing attention for it can increase operational safety and reliability, reduce maintenance cost, and improve productivity. However, because of the high nonlinear and strong fault concealment, the fault diagnosis of hydraulic systems is still a challenging task. Besides, the data samples collected from the hydraulic system are always in different sampling rates, and the coupling relationship between the components brings difficulties to accurate data acquisition. To solve the above issues, a deep learning model with multirate data samples is proposed in this article, which can extract features from the multirate sampling data automatically without expertise, thus it is more suitable in the industrial situation. Experiment results demonstrate that the proposed method achieves high diagnostic and fault pattern recognition accuracy even when the imbalance degree of sample data is as large as 1:100. Moreover, the proposed method can increase about 10% diagnosis accuracy when compared with some state-of-the-art methods.

Some applications of intuitionistic fuzzy sets using new similarity measure

Intuitionistic fuzzy set proposed by Atanassov, have gained attention from researchers for their applications in various fields. It is very useful in providing a flexible model to elaborate uncertainty and vagueness involved in decision making. In this paper, firstly some existing measures of distance and similarity are reviewed. We proposed its application in real life situations such as in diagnostic medicine and pattern recognition using new similarity measure. Moreover it is demonstrated that the proposed new similarity measure is capable of discriminating difference between patterns.

Characterization of hidden rules linking symptoms and selection of acupoint using an artificial neural network model.

Comprehension of the medical diagnoses of doctors and treatment of diseases is important to understand the underlying principle in selecting appropriate acupoints. The pattern recognition process that pertains to symptoms and diseases and informs acupuncture treatment in a clinical setting was explored. A total of 232 clinical records were collected using a Charting Language program. The relationship between symptom information and selected acupoints was trained using an artificial neural network (ANN). A total of 11 hidden nodes with the highest average precision score were selected through a tenfold cross-validation. Our ANN model could predict the selected acupoints based on symptom and disease information with an average precision score of 0.865 (precision, 0.911; recall, 0.811). This model is a useful tool for diagnostic classification or pattern recognition and for the prediction and modeling of acupuncture treatment based on clinical data obtained in a real-world setting. The relationship between symptoms and selected acupoints could be systematically characterized through knowledge discovery processes, such as pattern identification.

The Role of Muscle Imaging in the Diagnosis and Assessment of Children with Genetic Muscle Disease.

AbstractMuscle magnetic resonance imaging (MRI) and ultrasound (US) are emerging tools to assist in the diagnosis of children with genetic muscle disease. Increasing number of studies demonstrate that these imaging techniques can identify selective patterns of muscle atrophy, fatty degeneration, and muscle edema that help to distinguish between different early-onset genetic myopathies and muscular dystrophies. Recognizing patterns of pathology by muscle imaging can help to guide genetic testing and avoid the more invasive procedure of a muscle biopsy. Conversely, since massive parallel sequencing is now more commonly used as the initial step in diagnostic testing, imaging techniques can help to confirm or exclude if a variant of uncertain significance is indeed disease causing and compatible with a pattern of pathology as detected by muscle imaging. Whereas for diagnostic purposes and pattern recognition, muscle pathology does not need to be quantified, measuring disease progression is increasingly supported by quantitative muscle imaging, which is critical given the recent increment in rare disease therapeutic trials. Here, we discuss the value of muscle imaging techniques in pediatric muscle disease and summarize data identifying specific patterns of involvement in muscle MRI and US in some of the more common genetic myopathies and muscular dystrophies.

Pairwise diversity ranking of polychotomous features for ensemble physiological signal classifiers

It is well known that fusion classifiers for physiological signal classification with diverse components (classifiers or data sets) outperform those with less diverse components. Determining component diversity, therefore, is of the utmost importance in the design of fusion classifiers that are often employed in clinical diagnostic and numerous other pattern recognition problems. In this article, a new pairwise diversity-based ranking strategy is introduced to select a subset of ensemble components, which when combined will be more diverse than any other component subset of the same size. The strategy is unified in the sense that the components can be classifiers or data sets. Moreover, the classifiers and data sets can be polychotomous. Classifier-fusion and data-fusion systems are formulated based on the diversity-based selection strategy, and the application of the two fusion strategies are demonstrated through the classification of multichannel event-related potentials. It is observed that for both classifier and data fusion, the classification accuracy tends to increase/decrease when the diversity of the component ensemble increases/decreases. For the four sets of 14-channel event-related potentials considered, it is shown that data fusion outperforms classifier fusion. Furthermore, it is demonstrated that the combination of data components that yield the best performance, in a relative sense, can be determined through the diversity-based selection strategy.

Tracking Development of Clinical Reasoning Ability Across Five Medical Schools Using a Progress Test

Little is known about the acquisition of clinical reasoning skills in medical school, the development of clinical reasoning over the medical curriculum as a whole, and the impact of various curricular methodologies on these skills. This study investigated (1) whether there are differences in clinical reasoning skills between learners at different years of medical school, and (2) whether there are differences in performance between students at schools with various curricular methodologies. Students (n = 2,394) who had completed zero to three years of medical school at five U.S. medical schools participated in a cross-sectional study in 2008. Students took the same diagnostic pattern recognition (DPR) and clinical data interpretation (CDI) tests. Percent correct scores were used to determine performance differences. Data from all schools and students at all levels were aggregated for further analysis. Student performance increased substantially as a result of each year of training. Gains in DPR and CDI performance during the third year of medical school were not as great as in previous years across the five schools. CDI performance and performance gains were lower than DPR performance and gains. Performance gains attributable to training at each of the participating medical schools were more similar than different. Years of training accounted for most of the variation in DPR and CDI performance. As a rule, students at higher training levels performed better on both tests, though the expected larger gains during the third year of medical school did not materialize.

Commentary: Measurement and Interpretation Challenges in Comparing Student Performance Outcomes From Different Medical Schools

This commentary points to several measurement issues that arise in assessing medical student performance outcomes and then discusses the challenge of interpreting between-school differences. A problem often encountered in assessing student learning is creating an instrument that is at the right "pay grade." If it is too easy, ceiling effects compress scores. If it is too difficult, examinee performance can compress about chance values, and morale problems can occur. These issues are discussed in the context of a report by Williams and colleagues that measures medical student performance across five institutions on instruments assessing diagnostic pattern recognition and clinical data interpretation. The author of this commentary observes that, when interpreting between-school differences in assessing student learning, what can seem like small differences can have important consequences.

Multi-comparative systems biology analysis reveals time-course biosignatures of in vivo bovine pathway responses to B. melitensis, S. enterica Typhimurium and M. avium paratuberculosis

BackgroundTo decipher the complexity and improve the understanding of host-pathogen interactions, biologists must adopt new system level approaches in which the hierarchy of biological interactions and dynamics can be studied. This paper presents the application of systems biology for the cross-comparative analysis and interactome modeling of three different infectious agents, leading to the identification of novel, unique and common molecular host responses (biosignatures).MethodsA computational systems biology method was utilized to create interactome models of the host responses to Brucella melitensis (BMEL), Salmonella enterica Typhimurium (STM) and Mycobacterium avium paratuberculosis (MAP). A bovine ligated ileal loop biological model was employed to capture the host gene expression response at four time points post infection. New methods based on Dynamic Bayesian Network (DBN) machine learning were employed to conduct a systematic comparative analysis of pathway and Gene Ontology category perturbations.ResultsA cross-comparative assessment of 219 pathways and 1620 gene ontology (GO) categories was performed on each pathogen-host condition. Both unique and common pathway and GO perturbations indicated remarkable temporal differences in pathogen-host response profiles. Highly discriminatory pathways were selected from each pathogen condition to create a common system level interactome model comprised of 622 genes. This model was trained with data from each pathogen condition to capture unique and common gene expression features and relationships leading to the identification of candidate host-pathogen points of interactions and discriminatory biosignatures.ConclusionsOur results provide deeper understanding of the overall complexity of host defensive and pathogen invasion processes as well as the identification of novel host-pathogen interactions. The application of advanced computational methods for developing interactome models based on DBN has proven to be instrumental in conducting multi-conditional cross-comparative analyses. Further, this approach generates a fully simulateable model with capabilities for predictive analysis as well as for diagnostic pattern recognition. The resulting biosignatures may represent future targets for identification of emerging pathogens as well as for development of antimicrobial drugs, immunotherapeutics, or vaccines for prevention and treatment of diseases caused by known, emerging/re-emerging infectious agents.

The Efficacy of a Targeted Remediation Process for Students Who Fail Standardized Patient Examinations

Background: Current remediation strategies for students failing standardized patient examinations represent poorly targeted approaches since the specific nature of clinical performance weaknesses has not been defined. Purpose: The purpose is to determine the impact of a specifically targeted clinical performance course required of students who failed a clinical performance examination. Methods: A month-long clinical performance course, targeted to treat specific types of clinical performance deficiencies, was designed to remediate students failing standardized patient examinations in 2007 (n = 8) and 2008 (n = 5). Participating students were assessed on pre- and postperformance measures, including multiple-choice tests that measured diagnostic pattern recognition and clinical data interpretation and clinical performance measures using standardized clinical encounters. Comparisons between average pre- and postintervention performance scores were computed using paired sample t tests. Results were adjusted for regression toward the mean. Results: In both 2007 and 2008, the mean preintervention clinical data interpretation and standardized patient examination scores were below the criterion referenced passing standard set for the clinical competency exam. In both years the mean postintervention scores for the participants were above the passing standard for these two examinations. Pre- and postintervention differences were statistically significant in both cases. Conclusions: This study provides insight into the reasons that students fail clinical performance examinations and elucidates one method by which such students may be successfully remediated.

Semantic Sparse Representation of Disease Patterns

Semantic Sparse Representation of Disease PatternsSparse data representation is discussed in a context of useful fundamentals led to semantic content description and extraction of information. Disease patterns as semantic information extracted from medical images were underlined because of discussed application of computer-aided diagnosis. Compressive sensing rules were adjusted to the requirements of diagnostic pattern recognition. Proposed methodology of sparse disease patterns considers accuracy of sparse representation to estimate target content for detailed analysis. Semantics of sparse representation were modeled by morphological content analysis. Subtle or hidden components were extracted and displayed to increase information completeness. Usefulness of sparsity was verified for computer-aided diagnosis of stroke based on brain CT scans. Implemented method was based on selective and sparse representation of subtle hypodensity to improve diagnosis. Visual expression of disease signatures was fixed to radiologist requirements, domain knowledge and experimental analysis issues. Diagnosis assistance suitability was proven by experimental subjective rating and automatic recognition.

Medical Student Acquisition of Clinical Working Knowledge

Background: Working knowledge of physicians manifests as a combination of diagnostic pattern recognition and clinical data interpretation (analytic fact checking). Purpose: The purpose was to study medical student acquisition of these abilities as a function of years of medical training/experience. Methods: A cross-sectional study involving students who had completed 0, 1, 2, and 3 years of medical school. All students at all levels of training took the same tests of diagnostic pattern recognition and clinical data interpretation. Percent correct scores were calculated and used to estimate learning curves. A cohort of family physicians also took the test to provide a benchmark. Results: Student diagnostic pattern recognition and clinical data interpretation ability demonstrated a steady upward growth curve but leveled off in Year 3. Diagnostic pattern recognition performance was consistently higher than clinical data interpretation performance. The rate of diagnostic performance gain with training and experience was also higher. Conclusions: Medical students acquired diagnostic pattern recognition ability and all years of medical training contributed. The rate of clinical data interpretation performance improvement was slower, and the absolute performance level was lower. What was surprising was the lower rate of improvement in diagnostic pattern recognition and clinical data interpretation performance for students during their 1st year of clinical training. Students' understanding of findings and their relationships to disease processes may be affected by their limited patient experience.

Expert therapists use specific clinical reasoning processes in the assessment and management of patients with shoulder pain: a qualitative study

What are the key items in the clinical reasoning process which expert clinicians identify as being relevant to the assessment and management of patients with shoulder pain? Qualitative study using a three-round Delphi procedure. Twenty-six experts in the UK consented to be involved and were contactable, of whom 20 contributed, with 12, 15, and 15 contributing to the different rounds. Clinical reasoning was mostly about diagnostic reasoning, but also involved narrative reasoning. Diagnostic reasoning involved both pattern recognition and hypothetico-deductive reasoning. Diagnostic reasoning emphasised general history items, a constellation of signs and symptoms to identify specific diagnostic categories, and standard physical examination procedures. Narrative reasoning was highlighted by the communication involved in expert history taking, seeing patients in their functional and psychological context, and collaborative reasoning with the patient regarding management. These expert clinicians demonstrated the use of diagnostic pattern recognition, and hypothetico-deductive and narrative clinical reasoning processes. The emphasis was on the history and basic physical examination procedures to make clinical decisions.

Diagnostic Pattern Recognition on Gene-Expression Profile Data by Using One-Class Classification

In this paper, we perform diagnostic pattern recognition on a gene-expression profile data set by using one-class classification. Unlike conventional multiclass classifiers, the one-class (OC) classifier is built on one class only. For optimal performance, it accepts samples coming from the class used for training and rejects all samples from other classes. We evaluate six OC classifiers: the Gaussian model, Parzen windows, support vector data description (with two types of kernels: inner product and Gaussian), nearest neighbor data description, K-means, and PCA on three gene-expression profile data sets, those being an SRBCT data set, a Colon data set, and a Leukemia data set. Providing there is a good splitting of training and test samples and feature selection, most OC classifiers can produce high quality results. Parzen windows and support vector data description are "over-strict" in most cases, while nearest neighbor data description is "over-loose". Other classifiers are intermediate between these two extremes. The main difficulty for the OC classifier is it is difficult to obtain an optimum decision threshold if there are a limited number of training samples.

Is JPEG Compression of Videomicroscopic Images Compatible with Telediagnosis? Comparison between Diagnostic Performance and Pattern Recognition on Uncompressed TIFF Images and JPEG Compressed Ones

Early melanoma diagnosis is an important goal for dermatologists. Polarized light systems are increasingly employed for dermatoscopic diagnosis of melanocytic lesions. For the purpose of teledermoscopy, whose importance is increasingly growing for consultation and teaching purposes, it is of utmost importance to establish whether, after compression, polarized light images maintain their informativeness. The aim of our study was to check the effects of compression on melanocytic lesion images acquired by means of a digital videomicroscope on the identification of morphological details of the image and on diagnostic accuracy. A total of 170 50-fold-magnified pigmented skin lesion images, acquired in Tagged Image File Format (TIFF) by a digital videomicroscope, were compressed using Joint Photographic Experts Group (JPEG) algorithms (compression factor 30). Two experts in videomicroscopy evaluated both original and compressed images twice by describing single lesion features and expressing a diagnosis. Reproducibility in the assessment of dermoscopic parameters and observer performance were studied by kappa statistics and Receiver Operating Characteristic (ROC) analysis. Both intra- and interobserver reproducibility in the assessment of morphological details were higher when TIFF images were considered, indicating a better image quality. Nonetheless, there was no significant difference in the diagnostic accuracy between uncompressed images and compressed ones, although the intraobserver reproducibility in the diagnostic judgement was higher for uncompressed images. Despite loss in image details, factor 30 compressed videomicroscopic images enable a good diagnostic accuracy.

Is JPEG Compression of Videomicroscopic Images Compatible with Telediagnosis? Comparison between Diagnostic Performance and Pattern Recognition on Uncompressed TIFF Images and JPEG Compressed Ones

Early melanoma diagnosis is an important goal for dermatologists. Polarized light systems are increasingly employed for dermatoscopic diagnosis of melanocytic lesions. For the purpose of teledermoscopy, whose importance is increasingly growing for consultation and teaching purposes, it is of utmost importance to establish whether, after compression, polarized light images maintain their informativeness. The aim of our study was to check the effects of compression on melanocytic lesion images acquired by means of a digital videomicroscope on the identification of morphological details of the image and on diagnostic accuracy. A total of 170 50-fold-magnified pigmented skin lesion images, acquired in Tagged Image File Format (TIFF) by a digital videomicroscope, were compressed using Joint Photographic Experts Group (JPEG) algorithms (compression factor 30). Two experts in videomicroscopy evaluated both original and compressed images twice by describing single lesion features and expressing a diagnosis. Reproducibility in the assessment of dermoscopic parameters and observer performance were studied by kappa statistics and Receiver Operating Characteristic (ROC) analysis. Both intra- and interobserver reproducibility in the assessment of morphological details were higher when TIFF images were considered, indicating a better image quality. Nonetheless, there was no significant difference in the diagnostic accuracy between uncompressed images and compressed ones, although the intraobserver reproducibility in the diagnostic judgement was higher for uncompressed images. Despite loss in image details, factor 30 compressed videomicroscopic images enable a good diagnostic accuracy.

Classification of signatures of Bovine Spongiform Encephalopathy in serum using infrared spectroscopy.

Signatures of Bovine Spongiform Encephalopathy (BSE) have been identified in serum by means of "Diagnostic Pattern Recognition (DPR)". For DPR-analysis, mid-infrared spectroscopy of dried films of 641 serum samples was performed using disposable silicon sample carriers and a semi-automated DPR research system operating at room temperature. The combination of four mathematical classification approaches (principal component analysis plus linear discriminant analysis, robust linear discriminant analysis, artificial neural network, support vector machine) allowed for a reliable assignment of spectra to the class "BSE-positive" or "BSE-negative". An independent, blinded validation study was carried out on a second DPR research system at the Veterinary Laboratory Agency, Weybridge, UK. Out of 84 serum samples originating from terminally-ill, BSE-positive cattle, 78 were classified correctly. Similarly, 73 out of 76 BSE-negative samples were correctly identified by DPR such that, numerically, an accuracy of 94.4 % can be calculated. At a confidence level of 0.95 (alpha = 0.05) these results correspond to a sensitivity > 85% and a specificity > 90%. Identical class assignment by all four classifiers occurred in 75% of the cases while ambiguous results were obtained in only 8 of the 160 cases. With an area under the ROC (receiver operating charateristics) curve of 0.991, DPR may potentially supply a valuable surrogate marker for BSE even in cases in which a deliberate bias towards improved sensitivity or specificity is desired. To the best of our knowledge, DPR is the first and--up to now--only method which has demonstrated its capability of detecting BSE-related signatures in serum.

Traceability from weathered oil spills in the marine environment to the original contamination source. A case study.

An algorithm for stepwise approximation to the most reliable source of contamination is proposed. The concept of the algorithm is based on discrimination rather than on similarity. Gradually, by means of more complicated and higher distinguishing power approaches, the differences between the spilled sample and the suspected source are checked. Capillary gas chromatography with flame-ionization and mass spectrometric detection was applied to gather the necessary data. Fingerprint, diagnostic parameter approaches and pattern recognition methods were used for reliably distinguishing the analytical results. If the compared samples are not discriminated by any means, they are considered as identical.

Breast tissue classification using diagnostic ultrasound and pattern recognition techniques: II. Experimental results

The methods of statistical pattern recognition have been applied to the problem of in vivo ultrasonic characterization of breast disease in humans. Backscattered A-mode signals obtained from a commercial pulse imaging system were used to generate a large set of potentially useful features. Using statistical tests, a small subset of discriminatory features was selected to design a Bayes decision rule for each of two tissue classification schemes: malignant disease vs. benign disease, and malignant disease vs. (benign disease + normal tissue). Classification results obtained by the rotation method included sensitivities of 88 percent and 76 percent for the two schemes, based on data obtained from 32 women. These results are encouraging, though a definitive statement concerning the extrapolation of these numbers to the general population should only be made after obtaining results with a large data base.

Breast tissue classification using diagnostic ultrasound and pattern recognition techniques: I. Methods of pattern recognition

This paper discusses the application of statistical pattern recognition techniques to problems in diagnostic ultrasound. Using our own system as an example, we describe the concepts and specific methods that we have applied to a problem involving the computer-aided classification of breast tissue in vivo. Topics include feature generation, feature selection and classification, as well as a method which estimates the probability of error on classifying future data. An accompanying paper applies these methods to the classification of backscattered RF signals from normal and diseased breast tissue.