Refine
Document Type
- Article (19)
Language
- English (19)
Has Fulltext
- yes (19)
Is part of the Bibliography
- no (19) (remove)
Keywords
- data science (19) (remove)
Institute
- Medizin (18)
- Biochemie, Chemie und Pharmazie (1)
Recent advances in mathematical modelling and artificial intelligence have challenged the use of traditional regression analysis in biomedical research. This study examined artificial and cancer research data using binomial and multinomial logistic regression and compared its performance with other machine learning models such as random forests, support vector machines, Bayesian classifiers, k-nearest neighbours and repeated incremental clipping (RIPPER). The alternative models often outperformed regression in accurately classifying new cases. Logistic regression had a structural problem similar to early single-layer neural networks, which limited its ability to identify variables with high statistical significance for reliable class assignment. Therefore, regression is not always the best model for class prediction in biomedical datasets. The study emphasises the importance of validating selected models and suggests that a mixture of experts approach may be a more advanced and effective strategy for analysing biomedical datasets.
Feature selection is a common step in data preprocessing that precedes machine learning to reduce data space and the computational cost of processing or obtaining the data. Filtering out uninformative variables is also important for knowledge discovery. By reducing the data space to only those components that are informative to the class structure, feature selection can simplify models so that they can be more easily interpreted by researchers in the field, reminiscent of explainable artificial intelligence. Knowledge discovery in complex data thus benefits from feature selection that aims to understand feature sets in the thematic context from which the data set originates. However, a single variable selected from a very small number of variables that are technically sufficient for AI training may make little immediate thematic sense, whereas the additional consideration of a variable discarded during feature selection could make scientific discovery very explicit. In this report, we propose an approach to explainable feature selection (XFS) based on a systematic reconsideration of unselected features. The difference between the respective classifications when training the algorithms with the selected features or with the unselected features provides a valid estimate of whether the relevant features in a data set have been selected and uninformative or trivial information was filtered out. It is shown that revisiting originally unselected variables in multivariate data sets allows for the detection of pathologies and errors in the feature selection that occasionally resulted in the failure to identify the most appropriate variables.
Background: The categorization of individuals as normosmic, hyposmic, or anosmic from test results of odor threshold, discrimination, and identification may provide a limited view of the sense of smell. The purpose of this study was to expand the clinical diagnostic repertoire by including additional tests. Methods: A random cohort of n = 135 individuals (83 women and 52 men, aged 21 to 94 years) was tested for odor threshold, discrimination, and identification, plus a distance test, in which the odor of peanut butter is perceived, a sorting task of odor dilutions for phenylethyl alcohol and eugenol, a discrimination test for odorant enantiomers, a lateralization test with eucalyptol, a threshold assessment after 10 min of exposure to phenylethyl alcohol, and a questionnaire on the importance of olfaction. Unsupervised methods were used to detect structure in the olfaction-related data, followed by supervised feature selection methods from statistics and machine learning to identify relevant variables. Results: The structure in the olfaction-related data divided the cohort into two distinct clusters with n = 80 and 55 subjects. Odor threshold, discrimination, and identification did not play a relevant role for cluster assignment, which, on the other hand, depended on performance in the two odor dilution sorting tasks, from which cluster assignment was possible with a median 100-fold cross-validated balanced accuracy of 77–88%. Conclusions: The addition of an odor sorting task with the two proposed odor dilutions to the odor test battery expands the phenotype of olfaction and fits seamlessly into the sensory focus of standard test batteries.
Knowledge discovery in biomedical data using supervised methods assumes that the data contain structure relevant to the class structure if a classifier can be trained to assign a case to the correct class better than by guessing. In this setting, acceptance or rejection of a scientific hypothesis may depend critically on the ability to classify cases better than randomly, without high classification performance being the primary goal. Random forests are often chosen for knowledge-discovery tasks because they are considered a powerful classifier that does not require sophisticated data transformation or hyperparameter tuning and can be regarded as a reference classifier for tabular numerical data. Here, we report a case where the failure of random forests using the default hyperparameter settings in the standard implementations of R and Python would have led to the rejection of the hypothesis that the data contained structure relevant to the class structure. After tuning the hyperparameters, classification performance increased from 56% to 65% balanced accuracy in R, and from 55% to 67% balanced accuracy in Python. More importantly, the 95% confidence intervals in the tuned versions were to the right of the value of 50% that characterizes guessing-level classification. Thus, tuning provided the desired evidence that the data structure supported the class structure of the data set. In this case, the tuning made more than a quantitative difference in the form of slightly better classification accuracy, but significantly changed the interpretation of the data set. This is especially true when classification performance is low and a small improvement increases the balanced accuracy to over 50% when guessing.
Background: Persistent postsurgical neuropathic pain (PPSNP) can occur after intraoperative damage to somatosensory nerves, with a prevalence of 29–57% in breast cancer surgery. Proteomics is an active research field in neuropathic pain and the first results support its utility for establishing diagnoses or finding therapy strategies. Methods: 57 women (30 non-PPSNP/27 PPSNP) who had experienced a surgeon-verified intercostobrachial nerve injury during breast cancer surgery, were examined for patterns in 74 serum proteomic markers that allowed discrimination between subgroups with or without PPSNP. Serum samples were obtained both before and after surgery. Results: Unsupervised data analyses, including principal component analysis and self-organizing maps of artificial neurons, revealed patterns that supported a data structure consistent with pain-related subgroup (non-PPSPN vs. PPSNP) separation. Subsequent supervised machine learning-based analyses revealed 19 proteins (CD244, SIRT2, CCL28, CXCL9, CCL20, CCL3, IL.10RA, MCP.1, TRAIL, CCL25, IL10, uPA, CCL4, DNER, STAMPB, CCL23, CST5, CCL11, FGF.23) that were informative for subgroup separation. In cross-validated training and testing of six different machine-learned algorithms, subgroup assignment was significantly better than chance, whereas this was not possible when training the algorithms with randomly permuted data or with the protein markers not selected. In particular, sirtuin 2 emerged as a key protein, presenting both before and after breast cancer treatments in the PPSNP compared with the non-PPSNP subgroup. Conclusions: The identified proteins play important roles in immune processes such as cell migration, chemotaxis, and cytokine-signaling. They also have considerable overlap with currently known targets of approved or investigational drugs. Taken together, several lines of unsupervised and supervised analyses pointed to structures in serum proteomics data, obtained before and after breast cancer surgery, that relate to neuroinflammatory processes associated with the development of neuropathic pain after an intraoperative nerve lesion.
Bayesian inference is ubiquitous in science and widely used in biomedical research such as cell sorting or “omics” approaches, as well as in machine learning (ML), artificial neural networks, and “big data” applications. However, the calculation is not robust in regions of low evidence. In cases where one group has a lower mean but a higher variance than another group, new cases with larger values are implausibly assigned to the group with typically smaller values. An approach for a robust extension of Bayesian inference is proposed that proceeds in two main steps starting from the Bayesian posterior probabilities. First, cases with low evidence are labeled as “uncertain” class membership. The boundary for low probabilities of class assignment (threshold 𝜀
) is calculated using a computed ABC analysis as a data-based technique for item categorization. This leaves a number of cases with uncertain classification (p < 𝜀
). Second, cases with uncertain class membership are relabeled based on the distance to neighboring classified cases based on Voronoi cells. The approach is demonstrated on biomedical data typically analyzed with Bayesian statistics, such as flow cytometric data sets or biomarkers used in medical diagnostics, where it increased the class assignment accuracy by 1–10% depending on the data set. The proposed extension of the Bayesian inference of class membership can be used to obtain robust and plausible class assignments even for data at the extremes of the distribution and/or for which evidence is weak.
Olfactory self-assessments have been analyzed with often negative but also positive conclusions about their usefulness as a surrogate for sensory olfactory testing. Patients with nasal polyposis have been highlighted as a well-predisposed group for reliable self-assessment. In a prospective cohort of n = 156 nasal polyposis patients, olfactory threshold, odor discrimination, and odor identification were tested using the “Sniffin’ Sticks” test battery, along with self-assessments of olfactory acuity on a numerical rating scale with seven named items or on a 10-point scale with only the extremes named. Apparent highly significant correlations in the complete cohort proved to reflect the group differences in olfactory diagnoses of anosmia (n = 65), hyposmia (n = 74), and normosmia (n = 17), more than the true correlations of self-ratings with olfactory test results, which were mostly very weak. The olfactory self-ratings correlated with a quality of life score, however, only weakly. By contrast, olfactory self-ratings proved as informative in assigning the categorical olfactory diagnosis. Using an olfactory diagnostic instrument, which consists of a mapping rule of two numerical rating scales of one’s olfactory function to the olfactory functional diagnosis based on the “Sniffin’ Sticks” clinical test battery, the diagnoses of anosmia, hyposmia, or normosmia could be derived from the self-ratings at a satisfactorily balanced accuracy of about 80%. It remains to be seen whether this approach of translating self-assessments into olfactory diagnoses of anosmia, hyposmia, and normosmia can be generalized to other clinical cohorts in which olfaction plays a role.
The genetic background of pain is becoming increasingly well understood, which opens up possibilities for predicting the individual risk of persistent pain and the use of tailored therapies adapted to the variant pattern of the patient’s pain-relevant genes. The individual variant pattern of pain-relevant genes is accessible via next-generation sequencing, although the analysis of all “pain genes” would be expensive. Here, we report on the development of a cost-effective next generation sequencing-based pain-genotyping assay comprising the development of a customized AmpliSeq™ panel and bioinformatics approaches that condensate the genetic information of pain by identifying the most representative genes. The panel includes 29 key genes that have been shown to cover 70% of the biological functions exerted by a list of 540 so-called “pain genes” derived from transgenic mice experiments. These were supplemented by 43 additional genes that had been independently proposed as relevant for persistent pain. The functional genomics covered by the resulting 72 genes is particularly represented by mitogen-activated protein kinase of extracellular signal-regulated kinase and cytokine production and secretion. The present genotyping assay was established in 61 subjects of Caucasian ethnicity and investigates the functional role of the selected genes in the context of the known genetic architecture of pain without seeking functional associations for pain. The assay identified a total of 691 genetic variants, of which many have reports for a clinical relevance for pain or in another context. The assay is applicable for small to large-scale experimental setups at contemporary genotyping costs.
The use of artificial intelligence (AI) systems in biomedical and clinical settings can disrupt the traditional doctor–patient relationship, which is based on trust and transparency in medical advice and therapeutic decisions. When the diagnosis or selection of a therapy is no longer made solely by the physician, but to a significant extent by a machine using algorithms, decisions become nontransparent. Skill learning is the most common application of machine learning algorithms in clinical decision making. These are a class of very general algorithms (artificial neural networks, classifiers, etc.), which are tuned based on examples to optimize the classification of new, unseen cases. It is pointless to ask for an explanation for a decision. A detailed understanding of the mathematical details of an AI algorithm may be possible for experts in statistics or computer science. However, when it comes to the fate of human beings, this “developer’s explanation” is not sufficient. The concept of explainable AI (XAI) as a solution to this problem is attracting increasing scientific and regulatory interest. This review focuses on the requirement that XAIs must be able to explain in detail the decisions made by the AI to the experts in the field.
Because it is associated with central nervous changes, and olfactory dysfunction has been reported with increased prevalence among persons with diabetes, this study addressed the question of whether the risk of developing diabetes in the next 10 years is reflected in olfactory symptoms. In a cross-sectional study, in 164 individuals seeking medical consulting for possible diabetes, olfactory function was evaluated using a standardized clinical test assessing olfactory threshold, odor discrimination, and odor identification. Metabolomics parameters were assessed via blood concentrations. The individual diabetes risk was quantified according to the validated German version of the “FINDRISK” diabetes risk score. Machine learning algorithms trained with metabolomics patterns predicted low or high diabetes risk with a balanced accuracy of 63–75%. Similarly, olfactory subtest results predicted the olfactory dysfunction category with a balanced accuracy of 85–94%, occasionally reaching 100%. However, olfactory subtest results failed to improve the prediction of diabetes risk based on metabolomics data, and metabolomics data did not improve the prediction of the olfactory dysfunction category based on olfactory subtest results. Results of the present study suggest that olfactory function is not a useful predictor of diabetes.
Genetic association studies have shown their usefulness in assessing the role of ion channels in human thermal pain perception. We used machine learning to construct a complex phenotype from pain thresholds to thermal stimuli and associate it with the genetic information derived from the next-generation sequencing (NGS) of 15 ion channel genes which are involved in thermal perception, including ASIC1, ASIC2, ASIC3, ASIC4, TRPA1, TRPC1, TRPM2, TRPM3, TRPM4, TRPM5, TRPM8, TRPV1, TRPV2, TRPV3, and TRPV4. Phenotypic information was complete in 82 subjects and NGS genotypes were available in 67 subjects. A network of artificial neurons, implemented as emergent self-organizing maps, discovered two clusters characterized by high or low pain thresholds for heat and cold pain. A total of 1071 variants were discovered in the 15 ion channel genes. After feature selection, 80 genetic variants were retained for an association analysis based on machine learning. The measured performance of machine learning-mediated phenotype assignment based on this genetic information resulted in an area under the receiver operating characteristic curve of 77.2%, justifying a phenotype classification based on the genetic information. A further item categorization finally resulted in 38 genetic variants that contributed most to the phenotype assignment. Most of them (10) belonged to the TRPV3 gene, followed by TRPM3 (6). Therefore, the analysis successfully identified the particular importance of TRPV3 and TRPM3 for an average pain phenotype defined by the sensitivity to moderate thermal stimuli.
An important measure in pain research is the intensity of nociceptive stimuli and their cortical representation. However, there is evidence of different cerebral representations of nociceptive stimuli, including the fact that cortical areas recruited during processing of intranasal nociceptive chemical stimuli included those outside the traditional trigeminal areas. Therefore, the aim of this study was to investigate the major cerebral representations of stimulus intensity associated with intranasal chemical trigeminal stimulation. Trigeminal stimulation was achieved with carbon dioxide presented to the nasal mucosa. Using a single‐blinded, randomized crossover design, 24 subjects received nociceptive stimuli with two different stimulation paradigms, depending on the just noticeable differences in the stimulus strengths applied. Stimulus‐related brain activations were recorded using functional magnetic resonance imaging with event‐related design. Brain activations increased significantly with increasing stimulus intensity, with the largest cluster at the right Rolandic operculum and a global maximum in a smaller cluster at the left lower frontal orbital lobe. Region of interest analyses additionally supported an activation pattern correlated with the stimulus intensity at the piriform cortex as an area of special interest with the trigeminal input. The results support the piriform cortex, in addition to the secondary somatosensory cortex, as a major area of interest for stimulus strength‐related brain activation in pain models using trigeminal stimuli. This makes both areas a primary objective to be observed in human experimental pain settings where trigeminal input is used to study effects of analgesics.
Advances in flow cytometry enable the acquisition of large and high-dimensional data sets per patient. Novel computational techniques allow the visualization of structures in these data and, finally, the identification of relevant subgroups. Correct data visualizations and projections from the high-dimensional space to the visualization plane require the correct representation of the structures in the data. This work shows that frequently used techniques are unreliable in this respect. One of the most important methods for data projection in this area is the t-distributed stochastic neighbor embedding (t-SNE). We analyzed its performance on artificial and real biomedical data sets. t-SNE introduced a cluster structure for homogeneously distributed data that did not contain any subgroupstructure. Inotherdatasets,t-SNEoccasionallysuggestedthewrongnumberofsubgroups or projected data points belonging to different subgroups, as if belonging to the same subgroup. As an alternative approach, emergent self-organizing maps (ESOM) were used in combination with U-matrix methods. This approach allowed the correct identification of homogeneous data while in sets containing distance or density-based subgroups structures; the number of subgroups and data point assignments were correctly displayed. The results highlight possible pitfalls in the use of a currently widely applied algorithmic technique for the detection of subgroups in high dimensional cytometric data and suggest a robust alternative.
Based on accumulating evidence of a role of lipid signaling in many physiological and pathophysiological processes including psychiatric diseases, the present data driven analysis was designed to gather information needed to develop a prospective biomarker, using a targeted lipidomics approach covering different lipid mediators. Using unsupervised methods of data structure detection, implemented as hierarchal clustering, emergent self-organizing maps of neuronal networks, and principal component analysis, a cluster structure was found in the input data space comprising plasma concentrations of d = 35 different lipid-markers of various classes acquired in n = 94 subjects with the clinical diagnoses depression, bipolar disorder, ADHD, dementia, or in healthy controls. The structure separated patients with dementia from the other clinical groups, indicating that dementia is associated with a distinct lipid mediator plasma concentrations pattern possibly providing a basis for a future biomarker. This hypothesis was subsequently assessed using supervised machine-learning methods, implemented as random forests or principal component analysis followed by computed ABC analysis used for feature selection, and as random forests, k-nearest neighbors, support vector machines, multilayer perceptron, and naïve Bayesian classifiers to estimate whether the selected lipid mediators provide sufficient information that the diagnosis of dementia can be established at a higher accuracy than by guessing. This succeeded using a set of d = 7 markers comprising GluCerC16:0, Cer24:0, Cer20:0, Cer16:0, Cer24:1, C16 sphinganine, and LacCerC16:0, at an accuracy of 77%. By contrast, using random lipid markers reduced the diagnostic accuracy to values of 65% or less, whereas training the algorithms with randomly permuted data was followed by complete failure to diagnose dementia, emphasizing that the selected lipid mediators were display a particular pattern in this disease possibly qualifying as biomarkers.
Persistent and, in particular, neuropathic pain is a major healthcare problem with still insufficient pharmacological treatment options. This triggered research activities aimed at finding analgesics with a novel mechanism of action. Results of these efforts will need to pass through the phases of drug development, in which experimental human pain models are established components e.g. implemented as chemical hyperalgesia induced by capsaicin. We aimed at ranking the various readouts of a human capsaicin–based pain model with respect to the most relevant information about the effects of a potential reference analgesic. In a placebo‐controlled, randomized cross‐over study, seven different pain‐related readouts were acquired in 16 healthy individuals before and after oral administration of 300 mg pregabalin. The sizes of the effect on pain induced by intradermal injection of capsaicin were quantified by calculating Cohen's d. While in four of the seven pain‐related parameters, pregabalin provided a small effect judged by values of Cohen's d exceeding 0.2, an item categorization technique implemented as computed ABC analysis identified the pain intensities in the area of secondary hyperalgesia and of allodynia as the most suitable parameters to quantify the analgesic effects of pregabalin. Results of this study provide further support for the ability of the intradermal capsaicin pain model to show analgesic effects of pregabalin. Results can serve as a basis for the designs of studies where the inclusion of this particular pain model and pregabalin is planned.
Background: Many gene variants modulate the individual perception of pain and possibly also its persistence. The limited selection of single functional variants is increasingly being replaced by analyses of the full coding and regulatory sequences of pain-relevant genes accessible by means of next generation sequencing (NGS).
Methods: An NGS panel was created for a set of 77 human genes selected following different lines of evidence supporting their role in persisting pain. To address the role of these candidate genes, we established a sequencing assay based on a custom AmpliSeqTM panel to assess the exomic sequences in 72 subjects of Caucasian ethnicity. To identify the systems biology of the genes, the biological functions associated with these genes were assessed by means of a computational over-representation analysis.
Results: Sequencing generated a median of 2.85 ⋅ 106 reads per run with a mean depth close to 200 reads, mean read length of 205 called bases and an average chip loading of 71%. A total of 3,185 genetic variants were called. A computational functional genomics analysis indicated that the proposed NGS gene panel covers biological processes identified previously as characterizing the functional genomics of persisting pain.
Conclusion: Results of the NGS assay suggested that the produced nucleotide sequences are comparable to those earned with the classical Sanger sequencing technique. The assay is applicable for small to large-scale experimental setups to target the accessing of information about any nucleotide within the addressed genes in a study cohort.
Background: Prevention of persistent pain after breast cancer surgery, via early identification of patients at high risk, is a clinical need. Psychological factors are among the most consistently proposed predictive parameters for the development of persistent pain. However, repeated use of long psychological questionnaires in this context may be exhaustive for a patient and inconvenient in everyday clinical practice.
Methods: Supervised machine learning was used to create a short form of questionnaires that would provide the same predictive performance of pain persistence as the full questionnaires in a cohort of 1000 women followed up for 3 yr after breast cancer surgery. Machine-learned predictors were first trained with the full-item set of Beck's Depression Inventory (BDI), Spielberger's State–Trait Anxiety Inventory (STAI), and the State–Trait Anger Expression Inventory (STAXI-2). Subsequently, features were selected from the questionnaires to create predictors having a reduced set of items.
Results: A combined seven-item set of 10% of the original psychological questions from STAI and BDI, provided the same predictive performance parameters as the full questionnaires for the development of persistent postsurgical pain. The seven-item version offers a shorter and at least as accurate identification of women in whom pain persistence is unlikely (almost 95% negative predictive value).
Conclusions: Using a data-driven machine-learning approach, a short list of seven items from BDI and STAI is proposed as a basis for a predictive tool for the persistence of pain after breast cancer surgery.
The measurement of concentrations of drugs and endogenous substances is widely used in basic and clinical pharmacology research and service tasks. Using data science‐derived visualizations of laboratory data, it is demonstrated on a real‐life example that basic statistical exploration of laboratory assay results or advised standard visual methods of data inspection may fall short in detecting systematic laboratory errors. For example, data pathologies such as generating always the same value in all probes of a particular assay run may pass undetected when using standard methods of data quality check. It is shown that the use of different data visualizations that emphasize different views of the data may enhance the detection of systematic laboratory errors. A dotplot of single data in the order of assay is proposed that provides an overview on the data range, outliers and a particular type of systematic errors where similar values are wrongly measured in all probes.
Background: Modulation of cortical excitability by transcranial magnetic stimulation (TMS) is used for investigating human brain functions. A common observation is the high variability of long-term depression (LTD)-like changes in human (motor) cortex excitability. This study aimed at analyzing the response subgroup distribution after paired continuous theta burst stimulation (cTBS) as a basis for subject selection.
Methods: The effects of paired cTBS using 80% active motor threshold (AMT) in 31 healthy volunteers were assessed at the primary motor cortex (M1) corresponding to the representation of the first dorsal interosseous (FDI) muscle of the left hand, before and up to 50 min after plasticity induction. The changes in motor evoked potentials (MEPs) were analyzed using machine-learning derived methods implemented as Gaussian mixture modeling (GMM) and computed ABC analysis.
Results: The probability density distribution of the MEP changes from baseline was tri-modal, showing a clear separation at 80.9%. Subjects displaying at least this degree of LTD-like changes were n = 6 responders. By contrast, n = 7 subjects displayed a paradox response with increase in MEP. Reassessment using ABC analysis as alternative approach led to the same n = 6 subjects as a distinct category.
Conclusion: Depressive effects of paired cTBS using 80% AMT endure at least 50 min, however, only in a small subgroup of healthy subjects. Hence, plasticity induction by paired cTBS might not reflect a general mechanism in human motor cortex excitability. A mathematically supported criterion is proposed to select responders for enrolment in assessments of human brain functional networks using virtual brain lesions.