Robust Distance Measures for kNN Classification of Cancer Data

Machine learning application in otology

This review presents a comprehensive history of Artificial Intelligence (AI) in the context of the revolutionary application of machine learning (ML) to medical research and clinical utilization, particularly for the benefit of researchers interested in the application of ML in otology. To this end, we discuss the key components of ML-input, output, and algorithms. In particular, some representation algorithms commonly used in medical research are discussed. Subsequently, we review ML applications in otology research, including diagnosis, influential identification, and surgical outcome prediction. In the context of surgical outcome prediction, specific surgical treatments, including cochlear implantation, active middle ear implantation, tympanoplasty, and vestibular schwannoma resection, are considered. Finally, we highlight the obstacles and challenges that need to be overcome in future research.

Identification of circulating miRNA as early diagnostic molecular markers in malignant glioblastoma base on decision tree joint scoring algorithm

PURPOSE

The lack of clinical markers prevents early diagnosis of glioblastoma (GBM). Many studies have found that circulating microRNAs (miRNAs) can be used as early diagnostic markers of malignant tumours. Therefore, the identification of novel circulating miRNA biomolecular markers could be beneficial to clinicians in the early diagnosis of GBM.

METHODS

We developed a decision tree joint scoring algorithm (DTSA), systematically integrating significance analysis of microarray (SAM), Pearson hierarchical clustering, T test, Decision tree and Entropy weight score algorithm, to screen out circulating miRNA molecular markers with high sensitivity and accuracy for early diagnosis of GBM.

RESULTS

DTSA was developed and applied for GBM datasets and three circulating miRNA molecular markers were identified, namely, hsa-miR-2278, hsa-miR-555 and hsa-miR-892b. We have found that hsa-miR-2278 and hsa-miR-892b regulate the GBM pathway through target genes, promoting the development of GBM and affecting the survival of patients. DTSA has better classification effect in all data sets than other classification algorithms, and identified miRNAs are better than existing markers of GBM.

CONCLUSION

These results suggest that DTSA can effectively identify circulating miRNA, thus contributing to the early diagnosis and personalised treatment of GBM.

A Single-Molecule Bioelectronic Portable Array for Early Diagnosis of Pancreatic Cancer Precursors

A cohort of 47 patients is screened for pancreatic cancer precursors with a portable 96-well bioelectronic sensing-array for single-molecule assay in cysts fluid and blood plasma, deployable at point-of-care (POC). Pancreatic cancer precursors are mucinous cysts diagnosed with a sensitivity of at most 80% by state-of-the-art cytopathological molecular analyses (e.g., KRASmut DNA). Adding the simultaneous assay of proteins related to malignant transformation (e.g., MUC1 and CD55) is deemed essential to enhance diagnostic accuracy. The bioelectronic array proposed here, based on single-molecule-with-a-large-transistor (SiMoT) technology, can assay both nucleic acids and proteins at the single-molecule limit-of-identification (LOI) (1% of false-positives and false-negatives). It comprises an enzyme-linked immunosorbent assay (ELISA)-like 8 × 12-array organic-electronics disposable cartridge with an electrolyte-gated organic transistor sensor array, and a reusable reader, integrating a custom Si-IC chip, operating via software installed on a USB-connected smart device. The cartridge is complemented by a 3D-printed sensing gate cover plate. KRASmut , MUC1, and CD55 biomarkers either in plasma or cysts-fluid from 5 to 6 patients at a time, are multiplexed at single-molecule LOI in 1.5 h. The pancreatic cancer precursors are classified via a machine-learning analysis resulting in at least 96% diagnostic-sensitivity and 100% diagnostic-specificity. This preliminary study opens the way to POC liquid-biopsy-based early diagnosis of pancreatic-cancer precursors in plasma.

Defining and using immune archetypes to classify and treat cancer

Tumours are surrounded by a host immune system that can suppress or promote tumour growth. The tumour microenvironment (TME) has often been framed as a singular entity, suggesting a single type of immune state that is defective and in need of therapeutic intervention. By contrast, the past few years have highlighted a plurality of immune states that can surround tumours. In this Perspective, we suggest that different TMEs have 'archetypal' qualities across all cancers - characteristic and repeating collections of cells and gene-expression profiles at the level of the bulk tumour. We discuss many studies that together support a view that tumours typically draw from a finite number (around 12) of 'dominant' immune archetypes. In considering the likely evolutionary origin and roles of these archetypes, their associated TMEs can be predicted to have specific vulnerabilities that can be leveraged as targets for cancer treatment with expected and addressable adverse effects for patients.

Effect of distance measures on confidences of t-SNE embeddings and its implications on clustering for scRNA-seq data

Arguably one of the most famous dimensionality reduction algorithms of today is t-distributed stochastic neighbor embedding (t-SNE). Although being widely used for the visualization of scRNA-seq data, it is prone to errors as any algorithm and may lead to inaccurate interpretations of the visualized data. A reasonable way to avoid misinterpretations is to quantify the reliability of the visualizations. The focus of this work is first to find the best possible way to predict sample-based confidence scores for t-SNE embeddings and next, to use these confidence scores to improve the clustering algorithms. We adopt an RF regression algorithm using seven distance measures as features for having the sample-based confidence scores with a variety of different distance measures. The best configuration is used to assess the clustering improvement using K-means and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) based on Adjusted Rank Index (ARI), Normalized Mutual Information (NMI), and accuracy (ACC) scores. The experimental results show that distance measures have a considerable effect on the precision of confidence scores and clustering performance can be improved substantially if these confidence scores are incorporated before the clustering algorithm. Our findings reveal the usefulness of these confidence scores on downstream analyses for scRNA-seq data.

Distribution of the existence forms of condensable particulate matter during condensation: The surface collection and the space suspension forms

Condensable particulate matter (CPM), as an air pollutant that has received wide attention in recent years, has a high emission concentration compared to filterable particulate matter (FPM), yet there is not a well-developed removal method. Air pollution control devices (APCDs) with a condensation process have a certain effect on CPM removal, which inspired us to study the condensation behavior of CPM. During the condensation process, the condensed CPM may exist in two final forms: one was collected by the cold surface that caused the condensation; the other was converted to fine particles and suspended in the space of the flue. In a sense, the surface collection form can reflect the removal of CPM, while the CPM in the space suspension form should be further separated with the aim of removal. In this work, we adopted a CPM sampling system based on EPA Method 202 to reveal the distribution of the condensation behavior of CPM. In this sampling system, the CPM collected by all the cooling surfaces, including the cooling coil and impingers, can be counted as the surface collection form, while those collected by the terminal CPM filter can be regarded as the space suspension form. It was found that about 75 % of CPM was collected by the cooling surfaces, which suggested that CPM preferred to be in the surface collection form than the space suspension form. This preference characteristic also could be observed in the inorganic (CPM_i) and organic components of the CPM (CPM_o). Among the CPM_i, almost all NH₄⁺ and SO₄^2- condensed in the form of surface collection. The preference characteristics in CPM's (and its components') condensation behavior are similar under every temperature reduction condition. In this work, the interference of CPM measurement error was resolved by the statistical method of ANOVA.

Modified Euclidean-Canberra blend distance metric for kNN classifier

In today’s world different data sets are available on which regression or classification algorithms of machine learning are applied. One of the classification algorithms is k-nearest neighbor (kNN) which computes distance amongst various rows in a dataset. The performance of kNN is evaluated based on K-value and distance metric used where K is the total count of neighboring elements. Many different distance metrics have been used by researchers in literature, one of them is Canberra distance metric. In this paper the performance of kNN based on Canberra distance metric is measured on different datasets, further the proposed Canberra distance metric, namely, Modified Euclidean-Canberra Blend Distance (MECBD) metric has been applied to the kNN algorithm which led to improvement of class prediction efficiency on the same datasets measured in terms of accuracy, precision, recall, F1-score for different values of k. Further, this study depicts that MECBD metric use led to improvement in accuracy value 80.4% to 90.3%, 80.6% to 85.4% and 70.0% to 77.0% for various data sets used. Also, implementation of ROC curves and auc for k= 5 is done to show the improvement is kNN model prediction which showed increase in auc values for different data sets, for instance increase in auc values form 0.873 to 0.958 for Spine (2 Classes) dataset, 0.857 to 0.940, 0.983 to 0.983 (no change), 0.910 to 0.957 for DH, SL and NO class for Spine (3 Classes) data set and 0.651 to 0.742 for Haberman’s data set.

Artificial intelligence-driven prediction of multiple drug interactions

When a drug is administered to exert its efficacy, it will encounter multiple barriers and go through multiple interactions. Predicting the drug-related multiple interactions is critical for drug development and safety monitoring because it provides foundations for practical, safe compatibility and rational use of multiple drugs. With the progress of artificial intelligence (AI) technology, a variety of novel prediction methods for single interaction have emerged and shown great advantages compared to the traditional, expensive and time-consuming laboratory research. To promote the comprehensive and simultaneous predictions of multiple interactions, we systematically reviewed the application of AI in drug-drug, drug-food (excipients) and drug-microbiome interactions. We began by outlining the model methods, evaluation indicators, algorithms and databases commonly used to build models for three types of drug interactions. The models based on the metabolic enzyme P450, drug similarity and drug targets have empathized among the machine learning models of drug-drug interactions. In particular, we discussed the limitations of current approaches and identified potential areas for future research. It is anticipated the in-depth review will be helpful for the development of the next-generation of systematic prediction models for simultaneous multiple interactions.

Selection of Soybean Genotypes under Drought and Saline Stress Conditions Using Manhattan Distance and TOPSIS

The search for soybean genotypes more adapted to abiotic stress conditions is essential to boost the development and yield of the crop in Brazil and worldwide. In this research, we propose a new approach using the concept of distance (or similarity) in a vector space that can quantify changes in the morphological traits of soybean seedlings exposed to stressful environments. Thus, this study was conducted to select soybean genotypes exposed to stressful environments (saline or drought) using similarity based on Manhattan distance and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method. TOPSIS is a multi-criteria decision method for selecting the best alternative using the concept of distance. The use of TOPSIS is essential because the genotypes are not absolutely similar in both treatments. That is, just the distance measure is not enough to select the best genotype simultaneously in the two stress environments. Drought and saline stresses were induced by exposing seeds of 70 soybean genotypes to -0.20 MPa iso-osmotic solutions with polyethylene glycol-PEG 6000 (119.6 g L^-1) or NaCl (2.36 g L^-1) for 14 days at 25 °C. The germination rate, seedling length, and seedling dry matter were measured. We showed here how the genotypic stability of soybean plants could be quantified by TOPSIS when comparing drought and salinity conditions to a non-stressful environment (control) and how this method can be employed under different conditions. Based on the TOPSIS method, we can select the best soybean genotypes for environments with multiple abiotic stresses. Among the 70 tested soybean genotypes, RK 6813 RR, ST 777 IPRO, RK 7214 IPRO, TMG 2165 IPRO, 5G 830 RR, 98R35 IPRO, 98R31 IPRO, RK 8317 IPRO, CG 7464 RR, and LG 60177 IPRO are the 10 most stable genotypes under drought and saline stress conditions. Owing to high stability and gains with selection verified for these genotypes under salinity and drought conditions, they can be used as genitors in breeding programs to obtain offspring with higher resistance to antibiotic stresses.

Can You Do That Again? Time Series Consolidation as a Robust Method of Tailoring Gesture Recognition to Individual Users

Robust inter-session modeling of gestures is still an open learning challenge. A sleeve equipped with capacitive strap sensors was used to capture two gesture data sets from a convenience sample of eight subjects. Two pipelines were explored. In FILT a novel two-stage algorithm was introduced which uses an unsupervised learning algorithm to find samples representing gesture transitions and discards them prior to training and validating conventional models. In TSC a confusion matrix was used to automatically consolidate commonly confused class labels, resulting in a set of gestures tailored to an individual subject's abilities. The inter-session testing accuracy using the Time Series Consolidation (TSC) method increased from a baseline inter-session average of 42.47 ± 3.83% to 93.02% ± 4.97% while retaining an average of 5.29 ± 0.46 out of the 11 possible gesture categories. These pipelines used classic machine learning algorithms which require relatively small amounts of data and computational power compared to deep learning solutions. These methods may also offer more flexibility in interface design for users suffering from handicaps limiting their manual dexterity or ability to reliably make gestures, and be possible to implement on edge devices with low computational power.

Confidence estimation for t-SNE embeddings using random forest

Dimensionality reduction algorithms are commonly used for reducing the dimension of multi-dimensional data to visualize them on a standard display. Although many dimensionality reduction algorithms such as the t-distributed Stochastic Neighborhood Embedding aim to preserve close neighborhoods in low-dimensional space, they might not accomplish that for every sample of the data and eventually produce erroneous representations. In this study, we developed a supervised confidence estimation algorithm for detecting erroneous samples in embeddings. Our algorithm generates a confidence score for each sample in an embedding based on a distance-oriented score and a random forest regressor. We evaluate its performance on both intra- and inter-domain data and compare it with the neighborhood preservation ratio as our baseline. Our results showed that the resulting confidence score provides distinctive information about the correctness of any sample in an embedding compared to the baseline. The source code is available at https://github.com/gsaygili/dimred.

A Machine Learning Algorithm for Predicting the Risk of Developing to M1b Stage of Patients With Germ Cell Testicular Cancer

Objective

Distant metastasis other than non-regional lymph nodes and lung (i.e., M1b stage) significantly contributes to the poor survival prognosis of patients with germ cell testicular cancer (GCTC). The aim of this study was to develop a machine learning (ML) algorithm model to predict the risk of patients with GCTC developing the M1b stage, which can be used to assist in early intervention of patients.

Methods

The clinical and pathological data of patients with GCTC were obtained from the Surveillance, Epidemiology, and End Results (SEER) database. Combing the patient's characteristic variables, we applied six machine learning (ML) algorithms to develop the predictive models, including logistic regression(LR), eXtreme Gradient Boosting (XGBoost), light Gradient Boosting Machine (lightGBM), random forest (RF), multilayer perceptron (MLP), and k-nearest neighbor (kNN). Model performances were evaluated by 10-fold cross-receiver operating characteristic (ROC) curves, which calculated the area under the curve (AUC) of models for predictive accuracy. A total of 54 patients from our own center (October 2006 to June 2021) were collected as the external validation cohort.

Results

A total of 4,323 patients eligible for inclusion were screened for enrollment from the SEER database, of which 178 (4.12%) developing M1b stage. Multivariate logistic regression showed that lymph node dissection (LND), T stage, N stage, lung metastases, and distant lymph node metastases were the independent predictors of developing M1b stage risk. The models based on both the XGBoost and RF algorithms showed stable and efficient prediction performance in the training and external validation groups.

Conclusion

S-stage is not an independent factor for predicting the risk of developing the M1b stage of patients with GCTC. The ML models based on both XGBoost and RF algorithms have high predictive effectiveness and may be used to predict the risk of developing the M1b stage of patients with GCTC, which is of promising value in clinical decision-making. Models still need to be tested with a larger sample of real-world data.

Determination of Reactivity Ratios from Binary Copolymerization Using the k-Nearest Neighbor Non-Parametric Regression

This paper proposes a new method for calculating the monomer reactivity ratios for binary copolymerization based on the terminal model. The original optimization method involves a numerical integration algorithm and an optimization algorithm based on k-nearest neighbour non-parametric regression. The calculation method has been tested on simulated and experimental data sets, at low (<10%), medium (10–35%) and high conversions (>40%), yielding reactivity ratios in a good agreement with the usual methods such as intersection, Fineman–Ross, reverse Fineman–Ross, Kelen–Tüdös, extended Kelen–Tüdös and the error in variable method. The experimental data sets used in this comparative analysis are copolymerization of 2-(N-phthalimido) ethyl acrylate with 1-vinyl-2-pyrolidone for low conversion, copolymerization of isoprene with glycidyl methacrylate for medium conversion and copolymerization of N-isopropylacrylamide with N,N-dimethylacrylamide for high conversion. Also, the possibility to estimate experimental errors from a single experimental data set formed by n experimental data is shown.