Analysis of neuropsychological and neuroradiological features for diagnosis of Alzheimer's disease and mild cognitive impairment.

BACKGROUND: Age-related neurodegenerative diseases are constantly increasing with prediction that in 2050 over 60 % of population will suffer from some level of cognitive impairment. A cure for the Alzheimer's disease (AD) does not exist, so early diagnosis is of a great importance. Machine learning techniques can help in early diagnosis with deep medical data processing, disease understanding, intervention analysis and knowledge discovery for achieving better medical decision making. METHODS: In this paper, we analyze the dataset consisting of 90 individuals and 482 input features. We investigate the achieved AD prediction performances using seven classifiers and five feature selection algorithms. We pay special focus on analyzing performance by utilizing only a subset of best ranked attributes to establish the minimum amount of input features that ensure acceptable performance. We also investigate the significance of neuropsychological (NP) and neuroradiological (NR) attributes for the AD diagnosis. RESULTS: The accuracy for the whole set of attributes ranged between 66.22 % and 81.00 %, and the weighted average AUROC was between 76.3 % and 95.0 %. The best results were achieved by the naive Bayes classifier and the Relief feature selection algorithm. Additionally, Support Vector Machines classifier shows the most stable results since it depends the least on the feature selection algorithm which is used. As the main result of this paper, we compare the performance of models trained with automatically selected features to models trained with hand-selected features performed by medical experts (NP and NR features). CONCLUSIONS: The results reveal that unlike the NR attributes, the NP attributes achieve a good performance that is comparable to the full set of attributes, which suggests that they possess a high predictive power for AD diagnosis.

SuperFormer: Continual learning superposition method for text classification.

One of the biggest challenges in continual learning domains is the tendency of machine learning models to forget previously learned information over time. While overcoming this issue, the existing approaches often exploit large amounts of additional memory and apply model forgetting mitigation mechanisms which substantially prolong the training process. Therefore, we propose a novel SuperFormer method that alleviates model forgetting, while spending negligible additional memory and time. We tackle the continual learning challenges in a learning scenario, where we learn different tasks in a sequential order. We compare our method against several prominent continual learning methods, i.e., EWC, SI, MAS, GEM, PSP, etc. on a set of text classification tasks. We achieve the best average performance in terms of AUROC and AUPRC (0.7% and 0.9% gain on average, respectively) and the lowest training time among all the methods of comparison. On average, our method reduces the total training time by a factor of 5.4-8.5 in comparison to similarly performing methods. In terms of the additional memory, our method is on par with the most memory-efficient approaches.

The General Explanation Method with NMR Spectroscopy Enables the Identification of Metabolite Profiles Specific for Normal and Tumor Cell Lines.

Machine learning models in metabolomics, despite their great prediction accuracy, are still not widely adopted owing to the lack of an efficient explanation for their predictions. In this study, we propose the use of the general explanation method to explain the predictions of a machine learning model to gain detailed insight into metabolic differences between biological systems. The method was tested on a dataset of 1 H NMR spectra acquired on normal lung and mesothelial cell lines and their tumor counterparts. Initially, the random forests and artificial neural network models were applied to the dataset, and excellent prediction accuracy was achieved. The predictions of the models were explained with the general explanation method, which enabled identification of discriminating metabolic concentration differences between individual cell lines and enabled the construction of their specific metabolic concentration profiles. This intuitive and robust method holds great promise for in-depth understanding of the mechanisms that underline phenotypes as well as for biomarker discovery in complex diseases.

Mining data from hemodynamic simulations for generating prediction and explanation models.

One of the most common causes of human death is stroke, which can be caused by carotid bifurcation stenosis. In our work, we aim at proposing a prototype of a medical expert system that could significantly aid medical experts to detect hemodynamic abnormalities (increased artery wall shear stress). Based on the acquired simulated data, we apply several methodologies for1) predicting magnitudes and locations of maximum wall shear stress in the artery, 2) estimating reliability of computed predictions, and 3) providing user-friendly explanation of the model's decision. The obtained results indicate that the evaluated methodologies can provide a useful tool for the given problem domain.

Modern parameterization and explanation techniques in diagnostic decision support system: a case study in diagnostics of coronary artery disease.

OBJECTIVE: Coronary artery disease has been described as one of the curses of the western world, as it is one of its most important causes of mortality. Therefore, clinicians seek to improve diagnostic procedures, especially those that allow them to reach reliable early diagnoses. In the clinical setting, coronary artery disease diagnostics are typically performed in a sequential manner. The four diagnostic levels consist of evaluation of (1) signs and symptoms of the disease and electrocardiogram at rest, (2) sequential electrocardiogram testing during the controlled exercise, (3) myocardial perfusion scintigraphy, and (4) finally coronary angiography, that is considered as the "gold standard" reference method. Our study focuses on improving diagnostic performance of the third, virtually non-invasive, diagnostic level. METHODS AND MATERIALS: Myocardial scintigraphy results in a series of medical images that are obtained by relatively inexpensive means. In clinical practice, these images are manually described (parameterized) by expert physicians. In the paper we present an innovative alternative to manual image evaluation-an automatic image parameterization on multiple resolutions, based on texture description with specialized association rules. Extracted image parameters are combined into more informative composite parameters by means of principal component analysis, and finally used to build automatic classifiers with machine learning methods. RESULTS: Our experiments with synthetic datasets show that association-rule-based multi-resolution image parameterization works very well for scintigraphic images of the heart. In coronary artery disease diagnostics we confirm these results as our approach significantly improves on clinical results in terms of diagnostic performance. We improve diagnostic accuracy by 17%, specificity by 12% and sensitivity by 22%. We also significantly improve the number of reliably diagnosed patients by 19% for positive diagnoses, and 16% for negative diagnoses, so that no costly further tests are necessary for them. CONCLUSIONS: Multi-resolution image parameterization equals or even betters that of the physicians in terms of the diagnostic quality of image parameters. By using these parameters for building machine learning classifiers, we can significantly improve diagnostic performance with respect to the results of clinical practice, affect process rationalization, as well as possibly provide novel insights into the diagnostic problems, features and/or processes.

Computerized segmentation and diagnostics of whole-body bone scintigrams.

Bone scintigraphy or whole-body bone scan is one of the most common diagnostic procedures in nuclear medicine. Since expert physicians evaluate images manually some automated procedure for pathology detection is desired. A robust knowledge based methodology for segmenting body scans into the main skeletal regions is presented. The algorithm is simultaneously applied on anterior and posterior whole-body bone scintigrams. Expert knowledge is represented as a set of parameterized rules, used to support standard image processing algorithms. The segmented bone regions are parameterized with algorithms for classifying patterns so the pathologies can be classified with machine learning algorithms. This approach enables automatic scintigraphy evaluation of pathological changes, thus in addition to detection of point-like high-uptake lesions also other types can be discovered. Our study includes 467 consecutive, non-selected scintigrams. Automatic analysis of whole-body bone scans using our segmentation algorithm gives more accurate and reliable results than previous studies. Preliminary experiments show that our expert system based on machine learning closely mimics the results of expert physicians.

Computerized segmentation of whole-body bone scintigrams and its use in automated diagnostics.

Bone scintigraphy or whole-body bone scan is one of the most common diagnostic procedures in nuclear medicine used in the last 25 years. Pathological conditions, technically poor image resolution and artefacts necessitate that algorithms use sufficient background knowledge of anatomy and spatial relations of bones in order to work satisfactorily. A robust knowledge based methodology for detecting reference points of the main skeletal regions that is simultaneously applied on anterior and posterior whole-body bone scintigrams is presented. Expert knowledge is represented as a set of parameterized rules which are used to support standard image-processing algorithms. Our study includes 467 consecutive, non-selected scintigrams, which is, to our knowledge the largest number of images ever used in such studies. Automatic analysis of whole-body bone scans using our segmentation algorithm gives more accurate and reliable results than previous studies. Obtained reference points are used for automatic segmentation of the skeleton, which is applied to automatic (machine learning) or manual (expert physicians) diagnostics. Preliminary experiments show that an expert system based on machine learning closely mimics the results of expert physicians.

Comprehensible evaluation of prognostic factors and prediction of wound healing.

We analyzed the data of a controlled clinical study of the chronic wound healing acceleration as a result of electrical stimulation. The study involved a conventional conservative treatment, sham treatment, biphasic pulsed current, and direct current electrical stimulation. Data was collected over 10 years and suffices for an analysis with machine learning methods. So far, only a limited number of studies have investigated the wound and patient attributes which affect the chronic wound healing. There is none to our knowledge to include treatment attributes. The aims of our study are to determine effects of the wound, patient and treatment attributes on the wound healing process and to propose a system for prediction of the wound healing rate. First we analyzed which wound and patient attributes play a predominant role in the wound healing process and investigated a possibility to predict the wound healing rate at the beginning of the treatment based on the initial wound, patient and treatment attributes. Later we tried to enhance the wound healing rate prediction accuracy by predicting it after a few weeks of the wound healing follow-up. Using the attribute estimation algorithms ReliefF and RReliefF we obtained a ranking of the prognostic factors which was comprehensible to experts. We used regression and classification trees to build models for prediction of the wound healing rate. The obtained results are encouraging and may form a basis for an expert system for the chronic wound healing rate prediction. If the wound healing rate is known, then the provided information can help to formulate the appropriate treatment decisions and orient resources towards individuals with poor prognosis.