Objective This is the first Malaysian machine learning model to detect and disambiguate abbreviations in clinical notes. The model has been designed to be incorporated into MyHarmony, a natural language processing system, that extracts clinical information for health care management. The model utilizes word embedding to ensure feasibility of use, not in real-time but for secondary analysis, within the constraints of low-resource settings. Methods A Malaysian clinical embedding, based on Word2Vec model, was developed using 29,895 electronic discharge summaries. The embedding was compared against conventional rule-based and FastText embedding on two tasks: abbreviation detection and abbreviation disambiguation. Machine learning classifiers were applied to assess performance. Results The Malaysian clinical word embedding contained 7 million word tokens, 24,352 unique vocabularies, and 100 dimensions. For abbreviation detection, the Decision Tree classifier augmented with the Malaysian clinical embedding showed the best performance (F-score of 0.9519). For abbreviation disambiguation, the classifier with the Malaysian clinical embedding had the best performance for most of the abbreviations (F-score of 0.9903). Conclusion Despite having a smaller vocabulary and dimension, our local clinical word embedding performed better than the larger nonclinical FastText embedding. Word embedding with simple machine learning algorithms can decipher abbreviations well. It also requires lower computational resources and is suitable for implementation in low-resource settings such as Malaysia. The integration of this model into MyHarmony will improve recognition of clinical terms, thus improving the information generated for monitoring Malaysian health care services and policymaking.

Early diagnosis of melanoma, which can save thousands of lives, relies heavily on the analysis of dermoscopic images. One crucial diagnostic criterion is the identification of unusual pigment network (PN). However, distinguishing between regular (typical) and irregular (atypical) PN is challenging. This study aims to automate the PN detection process using a directional imaging algorithm and classify PN types using machine learning classifiers. The directional imaging algorithm incorporates Principal Component Analysis (PCA), contrast enhancement, filtering, and noise reduction. Applied to the PH2 dataset, this algorithm achieved a 96% success rate, which increased to 100% after pixel intensity adjustments. We created a new dataset containing only PN images from these results. We then employed two classifiers, Convolutional Neural Network (CNN) and Bag of Features (BoF), to categorize PN into atypical and typical classes. Given the limited dataset of 200 images, a simple and effective CNN was designed, featuring two convolutional layers and two batch normalization layers. The proposed CNN achieved 90% accuracy, 90% sensitivity, and 89% specificity. When compared to state-of-the-art methods, our CNN demonstrated superior performance. Our study highlights the potential of the proposed CNN model for effective PN classification, suggesting future research should focus on expanding datasets and incorporating additional dermatological features to further enhance melanoma diagnosis.

The color of skin lesions is a crucial diagnostic feature for identifying malignant melanoma and other skin diseases. Typical colors associated with melanocytic lesions include tan, brown, black, red, white, and blue-gray. This study introduces a novel feature: the number of colors present in lesions, which can indicate the severity of skin diseases and help distinguish melanomas from benign lesions. We propose a color histogram analysis, a traditional image processing technique, to analyze the pixels of skin lesions from three publicly available datasets: PH2, ISIC2016, and Med-Node, which include dermoscopic and non-dermoscopic images. While the PH2 dataset contains ground truth about skin lesion colors, the ISIC2016 and Med-Node datasets lack such annotations; our algorithm establishes this ground truth using the color histogram analysis based on the PH2 dataset. We then design and train a 19-layer Convolutional Neural Network (CNN) with different skip connections of residual blocks to classify lesions into three categories based on the number of colors present. The DeepDream algorithm is utilized to visualize the learned features of different layers, and multiple configurations of the proposed CNN are tested, achieving the highest weighted F1-score of 75.00 % on the test set. LIME is subsequently applied to identify the most important features influencing the model's decision-making. The findings demonstrate that the number of colors in lesions is a significant feature for describing skin conditions. The proposed CNN, particularly with three skip connections, shows strong potential for clinical application in diagnosing melanoma, supporting its use alongside traditional diagnostic methods.

Autistic people typically need methodical support as they explore and interact with their immediate surroundings and the objects associated with them, emphasising the importance of spatial knowledge and cognitive skills in improving and understanding their surroundings. The objective of this research paper is to present a conceptual and technical framework that could be of significant assistance in developing spatial ability and cognitive skills in autistic people. This framework is constructed based on aspects of autistic behaviour, spatial cognition, and technical aspects of haptic technology and virtual environments (VEs). The framework uses the Model, View, Controller (MVC) pattern as its structural foundation. This paper highlights the important elements required for scholars and designers to create and develop a haptic-based virtual environment (HBVE) for autistic people in order to improve spatial knowledge and cognitive skills. Furthermore, this research also aimed to create an HBVE application based on the proposed framework in order to assess its effectiveness. This research paper evaluates the proposed framework based on end-user perspectives. The findings indicate that the experimental study conducted with autistic people has demonstrated notable progress in enhancing spatial knowledge and cognitive abilities among this population. The experimental implementation of these findings supports the developed framework's efficacy.

Melanoma, one of the deadliest types of skin cancer, accounts for thousands of fatalities globally. The bluish, blue-whitish, or blue-white veil (BWV) is a critical feature for diagnosing melanoma, yet research into detecting BWV in dermatological images is limited. This study utilizes a non-annotated skin lesion dataset, which is converted into an annotated dataset using a proposed imaging algorithm (color threshold techniques) on lesion patches based on color palettes. A Deep Convolutional Neural Network (DCNN) is designed and trained separately on three individual and combined dermoscopic datasets, using custom layers instead of standard activation function layers. The model is developed to categorize skin lesions based on the presence of BWV. The proposed DCNN demonstrates superior performance compared to the conventional BWV detection models across different datasets. The model achieves a testing accuracy of 85.71 % on the augmented PH2 dataset, 95.00 % on the augmented ISIC archive dataset, 95.05 % on the combined augmented (PH2+ISIC archive) dataset, and 90.00 % on the Derm7pt dataset. An explainable artificial intelligence (XAI) algorithm is subsequently applied to interpret the DCNN's decision-making process about the BWV detection. The proposed approach, coupled with XAI, significantly improves the detection of BWV in skin lesions, outperforming existing models and providing a robust tool for early melanoma diagnosis.

In dermoscopic images, which allow visualization of surface skin structures not visible to the naked eye, lesion shape offers vital insights into skin diseases. In clinically practiced methods, asymmetric lesion shape is one of the criteria for diagnosing Melanoma. Initially, we labeled data for a non-annotated dataset with symmetrical information based on clinical assessments. Subsequently, we propose a supporting technique-a supervised learning image processing algorithm-to analyze the geometrical pattern of lesion shape, aiding non-experts in understanding the criteria of an asymmetric lesion. We then utilize a pre-trained convolutional neural network (CNN) to extract shape, color, and texture features from dermoscopic images for training a multiclass support vector machine (SVM) classifier, outperforming state-of-the-art methods from the literature. In the geometry-based experiment, we achieved a 99.00 % detection rate for dermatological asymmetric lesions. In the CNN-based experiment, the best performance is found 94 % Kappa Score, 95 % Macro F1-score, and 97 % weighted F1-score for classifying lesion shapes (Asymmetric, Half-Symmetric, and Symmetric).

Background: Many neurodegenerative diseases affect human gait. Gait analysis is an example of a non-invasive manner to diagnose these diseases. Nevertheless, gait analysis is difficult to do because patients with different neurodegenerative diseases may have similar human gaits. Machine learning algorithms may improve the correct identification of these pathologies. However, the problem with many classification algorithms is a lack of transparency and interpretability for the final user. Methods: In this study, we implemented the PS-Merge operator for the classification, employing gait biomarkers of a public dataset. Results: The highest classification percentage was 83.77%, which means an acceptable degree of reliability. Conclusions: Our results show that PS-Merge has the ability to explain how the algorithm chooses an option, i.e., the operator can be seen as a first step to obtaining an eXplainable Artificial Intelligence (XAI).

One of the most prevalent behavioral impairments in autistic people is difficulty processing sensory information. People commonly observe this phenomenon as either hypersensitivity or hyposensitivity to tactile stimuli. To rectify this irregularity, numerous researchers have suggested wearable sensor-based systems and applications within the realm of virtual environments. However, they have neglected to carry out an adequate evaluation and proof of its feasibility for autistic people. Hence, this study compares three methods to identify the most effective approach to understanding tactile sensory processing in autistic people using haptic technology. The evaluation included behavioral response analysis, which involves observing autistic people; statistical analysis on tactile sensory patterns (TSP), which analyzes data from 9-axis IMU sensors and EMG sensors; and machine learning models, such as recurrent neural networks (RNNs), trained on tactile sensory sensitivity data. The study demonstrates that behavioral response analysis is limited by subjectivity and variability in responses, despite its capacity to provide useful qualitative perspectives. Meanwhile, statistical analysis reveals limitations in its ability to predict sensory outcomes, despite its capacity to provide quantitative measurements of variations in tactile sensory processing. Comparative analysis using machine learning, on the other hand, outperforms both behavioral response analysis and statistical analysis in tactile sensory processing classification and prediction. In particular, the RNN model exhibits remarkable accuracy and correctness in detecting tactile sensory processing among autistic people. This study demonstrated that machine learning can be advantageous for autistic people to analyze tactile sensory processing, explore, and develop touch sensitivity to improve their quality of life.

Medical abbreviations can be misinterpreted and endanger patients' lives. This research is the first to investigate the prevalence of abbreviations in Malaysian electronic discharge summaries, where English is widely used, and elicit the risk factors associated with dangerous abbreviations. We randomly sampled and manually annotated 1102 electronic discharge summaries for abbreviations and their senses. Three medical doctors assigned a danger level to ambiguous abbreviations based on their potential to cause patient harm if misinterpreted. The predictors for dangerous abbreviations were determined using binary logistic regression. Abbreviations accounted for 19% (33,824) of total words; 22.6% (7640) of those abbreviations were ambiguous; and 52.3% (115) of the ambiguous abbreviations were labelled dangerous. Increased risk of danger occurs when abbreviations have more than two senses (OR = 2.991; 95% CI 1.586, 5.641), they are medication-related (OR = 6.240; 95% CI 2.674, 14.558), they are disorders (OR = 7.771; 95% CI 2.054, 29.409) and procedures (OR = 3.492; 95% CI 1.376, 8.860). Reduced risk of danger occurs when abbreviations are confined to a single discipline (OR = 0.519; 95% CI 0.278, 0.967). Managing abbreviations through awareness and implementing automated abbreviation detection and expansion would improve the quality of clinical documentation, patient safety, and the information extracted for secondary purposes.

Machine learning (ML)-based detection of diseases using sequence-based gut microbiome data has been of great interest within the artificial intelligence in medicine (AIM) community.The approach offers a non-invasive alternative for colorectal cancer detection, which is based on stool samples.Considering limitations of existing methods in CRC detection, medical research has shown interest in the use of high throughput data to identify the disease.Owing to several limitations of conventional ML algorithms, deep learning (DL) methods are becoming more popular due to their outstanding performance in related fields.However, the performance of DL methods is affected by limitations such as dimensionality, sparsity, and feature dominance inherent in microbiome data.This research proposes stacking and chaining of normalization methods to address the limitations.While the stacking technique offers a robust, easy to use, and interpretable alternative for augmenting microbiome and other tabular data, the chaining technique is an alternative to data normalization that dynamically adjusts the underlying properties of data towards the normal distribution.The proposed techniques are combined with rank transformation and feature selection to further improve the performance of the model, with area under the curve (AUC) values between 0.857 to 0.987 using publicly available datasets.

Blastocystis sp . is an enteric protistan parasite that affects individuals worldwide with gastrointestinal symptoms such as abdominal discomfort, diarrhea, and flatulence. However, its pathogenicity is controversial due to its presence among asymptomatic individuals. Blastocystis sp. subtype 3 (ST3) is the most prevalent subtype among humans that have been associated with irritable bowel syndrome (IBS), Crohn’s disease, ulcerative colitis, and colorectal cancer. Axenization of the parasite has been shown to impede its growth thus revealing the importance of accompanying bacteria in ensuring Blastocystis sp. survival. This study aims to identify the influence of accompanying bacteria on the growth of Blastocystis sp. ST3. Blastocystis sp. cultures were treated with Meropenem, Vancomycin, and Amoxicillin-Clavulanic acid (Augmentin). Bacteria-containing supernatant of antibiotic-treated and control cultures were isolated and identified through 16 s rRNA sequencing. Morphological changes of antibiotic-treated Blastocystis sp. ST3 were also observed. The cultures treated with meropenem and augmentin exhibited opposing effects with reduced growth of isolates from symptomatic patients and a significant increase in asymptomatic isolates. Whereas, vancomycin-treated cultures had no difference in the growth of Blastocystis sp. ST3 isolates from symptomatic and asymptomatic patients. Isolates from symptomatic and asymtomatic patients had 6 and 2 distinct bacterial species identified with Proteus mirabilis as the common bacteria among both types of isolates. Morphologically, Blastocystis sp. ST3 cultures exposed to meropenem and augmentin demonstrated an increase in pre-cystic forms. These findings demonstrate the effects of accompanying bacteria on the growth of Blastocystis sp. ST3 that could translate into clinical manifestations observed among Blastocystis sp.-infected patients.

Location metonymy resolution is a study that deals with locations being used in a non-literal way that create problems in several natural language processing tasks such as Named entity recognition and Geographical parsing. Many studies were conducted attempting to accurately classify whether the location is used literally or metonymically, however, most of the approaches that performed well had to employ a considerable amount of resources along with complex machine learning models; those that reduced the resources experienced a decline in performance due to data sparseness. This study proposes a novel feature selection approach that uses bag-of-words and augments it with GloVe embeddings to obtain features that can be recognized based on the context of the sentence. We then implement a minimalist deep learning model making the entire classification task as light as possible. The study found that relying solely on the given datasets to identify features without depending on other external resources can achieve remarkable results despite the small size of the datasets. The results obtained from evaluating our method compared to the state-of-the-art methods show that eliminating noise based on the context notwithstanding the usage of low-cost resources has outperformed all of the previous methods with an accuracy of 99.2% on the WIMCOR dataset.

One of the worst forms of skin cancer is melanoma which can be curable if it is diagnosed at an early stage.The earlier the cancer is diagnosed, the better is the outcome.The risk of death from melanoma is directly related to the delay in identifying a lesion.A Deep Learning-based computer diagnosing system can be an automated solution in clinical assessments to overcome this problem.Convolutional Neural Network (CNN) can help to improve the classification rate of skin lesions from dermoscopic images without the need for any human assistance.The linear and nonlinear activation functions act as a node placed at hidden layers or output layers of a Neural Network to play a role in deciding whether that node should pass information to the following layer or not.This critical mathematical mechanism influences the accuracy rate of the CNN.To obtain acceptable performance, CNN requires a large amount of training data.This research shows how fast and effective different types of nonlinear activation functions work on a CNN with limited image datasets.Experimental analysis reveals that the proposed CNN model with parameterized Leaky ReLU function outperforms (97.50% accuracy, 98.00% precision, and 98.00% sensitivity) the same network with distinct nonlinear activation functions for the problem of melanoma recognition by classifying skin lesion into three classes.All experimental studies are carried out using images from PH2 (a dermoscopic image database obtained at the Hospital Pedro Hispano Dermatology Service in Matosinhos, Portugal) and International Skin Imaging Collaboration (ISIC) archive datasets.

Stacked ensemble, which formulates an ensemble by using a meta-learner to combine (stack) the predictions of multiple base classifiers, suffers from the problem of suboptimal performance on imbalanced classification. To improve the classification performance of stacked ensemble on imbalanced datasets, we proposed a method named Neighborhood Undersampling Stacked Ensemble (NUS-SE) in this paper. In general, the NUS-SE can be broken down into two proposed components, an undersampling based stacked ensemble framework (US-SE) component and an undersampling technique component. In the metadata generation step of stacked ensemble, a cross-validation-like procedure (CV-prediction) is commonly used. Unfortunately, incomplete metadata with missing prediction values is generated when undersampling is performed within a stacked ensemble which utilized CV-prediction as the metadata generation procedure. Therefore, in the proposed US-SE component, we replaced the standard CV-prediction procedure with our proposed method coined as Subset and Out-of-Subset (S-OOS) prediction procedure as the metadata generation method. S-OOS prediction procedure will generate metadata without missing prediction values and thus enabling the integration of undersampling within stacked ensemble. By integrating undersampling within stacked ensemble, multiple undersampled-data-subsets are used in the training of US-SE's base learners. While in the undersampling component, we further proposed a novel undersampling technique — Neighborhood Undersampling (NUS) which selects majority instances based on their local neighborhood information. The performance of the NUS-SE is evaluated against those non-resampling based stacked ensemble as baseline methods. The experiment demonstrates that the proposed NUS-SE, which is an undersampling based stacked ensemble, is capable of achieving a better performance when compared to the non-resampling based stacked ensemble.

The selection of a meta-learner determines the success of a stacked ensemble as the meta-learner is responsible for the final predictions of the stacked ensemble. Unfortunately, in imbalanced classification, selecting an appropriate and well-performing meta-learner of stacked ensemble is not straightforward as different meta-learners are advocated by different researchers. To investigate and identify a well-performing type of meta-learner in stacked ensemble for imbalanced classification, an experiment consisting of 19 meta-learners was conducted, detailed in this paper. Among the 19 meta-learners of stacked ensembles, a new weighted combination-based meta-learner that maximizes the H-measure during the training of stacked ensemble was first introduced and implemented in the empirical evaluation of this paper. The classification performances of stacked ensembles with 19 different meta-learners were recorded using both the area under the receiver operating characteristic curve (AUC) and H-measure (a metric that overcomes the deficiencies of the AUC). The weighted combination-based meta-learners of stacked ensembles have better classification performances on imbalanced datasets when compared to bagging-based, boosting-based, Decision Trees, Support Vector Machines, Naive Bayes, and Feedforward Neural Network meta-learners. Thus, the adoption of weighted combination-based meta-learners in stacked ensembles is recommended for their better performance on imbalanced datasets. Also, based on the empirical results, we identified better-performing meta-learners (such as the AUC maximizing meta-learner and the H-measure maximizing meta-learner) than the widely adopted meta-learner - Logistic Regression - in imbalanced classification.

Colorectal cancer (CRC) is the third most deadly cancer worldwide. The use of gut microbiome in early detection of the disease has attracted much attention from the research community, mainly because of its noninvasive nature. Recent achievements in next generation sequencing technology have led to increased availability of sequence data and enabled an environment for the growth of gut microbiome research. The use of conventional machine learning algorithms for automatic detection of CRC based on the microbiome is limited by factors such as low accuracy and the need for manual selection of features. Despite their success in other fields, Deep Neural Network (DNN) algorithms have limitations in microbiome-based CRC classification. These limitations include high dimensionality of microbiome data and other characteristics associated with sequence data such as feature dominance. In this paper, we propose a feature augmentation approach that aggregates data normalization methods to extend existing features of a dataset. The proposed method combines feature extension with data augmentation to improve CRC classification performance of a DNN model. The proposed model obtained area under the curve (AUC) scores of 0.96 and 0.89 on two publicly available microbiome datasets.